Merging win probabilities

[TravisVOX]

Let's say you have a set of win probabilities for a race...

HORSE PUBLIC USER OVER%
V. E. DAY 0.0410 0.072 76%
WICKED STRONG 0.2290 0.170 -26%
TONALIST 0.2290 0.210 -8%
KID CRUZ 0.0630 0.063 0%
MR SPEAKER 0.0970 0.114 18%
VIVA MAJORCA 0.0270 0.020 -26%
CHARGE NOW 0.0250 0.083 232%
ULANBATOR 0.0100 0.073 630%
COMMAND. CURVE 0.0400 0.036 -10%
BAYERN 0.2420 0.160 -34%

Let's say the user column was your probabilities of winning the race. The right column would be "overlay percentage" of each horse relative to the public's win percentage.

My question is this...

Obviously, overlays can carry a positive expectation if you account for odds. However, I'm not worried about that in this part of the process. What I'm trying to do is balance the difference between the users odds and the users odds overlay percentage.

Said differently, the user thinks that Ulanbator is a tremendous overlay. However, he/she still thinks that Bayern is a more likely winner, despite being an underlay.

What mathematical and/or statistical techniques can be utilized to balance/merge/combine the user and over% columns to reach a logical win percentage or ordinal representation of the analysis.

One way I could do it is to rank each column then sum the rankings and order them. Then I thought, however, I might want to weight each of those columns differently... at which point I figured I would come on here and see if anyone had thoughts or insight.

[davew]

How good is your line? Many people feel that even if your line is very good, the public line on the average is better.

How about the third column instead of a percentage overlay/underlay make it a difference of your probability percentage and the publics.

Then add a 4th column that moves your line a certain percentage towards the publics. I read somewhere a guy moving it 15% although up towards 25% seems reasonable.

This will give you a blended line that lessens the outliers (which may or may not be correct). You should be able to check this to see which line is closer with a few thousand races. (public, yours, blended)

Then use the best probabilities for your exactas and straight pools to form profitable 'dutch bets' that include all overlays and very minor underlays that are higher frequency hits. The goal is to try and stay away from massive underlays (which pay the track takeout when they miss and give you profit on your overlay bets when they hit).

[sjk]

I would bet in proportion to edge x win probability as per fractional Kelly. It is likely that a linear combination of your line and the public's line is a more accurate estimate of win probability than your line on its own so you might use a 50-50 blend of the two as your win probability.

For the edge part you want to establish a base over% which is the lowest you will bet. If you think the 50-50 blend is reasonable, a 100% threshold is probably OK. Pick a ceiling (say 300%) since the further you are from the public's line the more suspect your overlay tends to be.

So here you bet the two who are over 100%:

Charge now - 2.32 x 1/2(.025+.083)
Ulanbator - 3.00 x 1/2(.01+.073) since the 630 exceeds the cap.

Oddly both work out to around .125. Choose a base amount to multiply by to scale the bet, say $200. In this case you bet $25 on each.

[classhandicapper]

I never really understood combining your own line with the public's line.

I understand why it might be more predictive. But IMO all you need to do is give yourself a margin of safety based on your confidence level in that particular race before you bet.

Let's say you give a horse a 33% chance of winning and he's 2.1 - 1.

You know none of your lines are going to be perfect, so you don't bet. There's not enough margin of safety in 2.1 -1 .

If he's 5-2 and you are very confident in your analysis you might consider it.

If you aren't as confident, maybe 3-1 is not enough.

[sjk]

Incorporating the public odds into your line is arithmetically equivalent to manipulating the acceptable overlay percent. It doesn't really matter why you are doing it.

[GameTheory]

When the idea of blending your own oddsline with the public's to make a "better" one comes up, this is always the response. What is the point? You're just averaging them and you'll end up somewhere in the middle, right? Just demand a bigger overlay. And that's all true.

But a simple weighted average is not the proper way to make a blended line -- the idea is to end up with an unbiased line relative to the public's. In other words, if you take your own oddsline and look at (for instance) all the horses that it has assigned 3-1 odds to (looking over many races), then we all know even if that sample wins at the overall average rate of 25% as it should that the subset of those where the public odds are below 3-1 are going to win more than those where the public odds are higher than 3-1. An unbiased line doesn't have that property, the 3-1 horses will win at 25% regardless of their public odds. A simple constant weighted linear average of the two lines will never achieve this (unless you set the weights to 100-0 in favor of the public).

So I would suggest instead of simple averaging that you actual make a new model of your line and the public line's. This is exactly what Benter originally suggested and presumably just what he did, I don't know why that always gets lost in the shuffle when talk of blending lines comes up. He said make a new logistic regression model to create a new output line. (But there are various other forms it could take, and you're not strictly limited to only those two inputs either.)

In other words build a new model using a training sample of your lines and the public's from actual races and using actual results. The output from this model for each race would then then need to be re-normalized (so the probabilities sum to 1.0 again) and/or further calibrated (sometimes spreading the probabilities is needed).

Now instead of just making some assumption about how much weight each line should have, you are using a training sample of both lines and using actual race results to predict a better number which will in some cases be higher or lower than EITHER of the input numbers (an amplification effect if you will, rather than just a blending to the middle), and then once re-normalized you'll have a (possibly) completely different line (an unbiased one hopefully). This is a much more powerful transformation. That is, it will be if the user line actually contains value-added information not contained in the public line. If it does not, the model will end up spitting out a copy of the public line. (And that's what often happens -- this will show you just what your line is really worth.)

OR, depending on how you came up with your line to begin with, you might be able to just throw the public probability in your base model as a handicapping factor, and it would have a similar effect. But in order to get an unbiased probability, you've got to deal with the public probability SOMEHOW, and a simple averaging at the end isn't it as it is just too crude...

[TrifectaMike]

When the idea of blending your own oddsline with the public's to make a "better" one comes up, this is always the response. What is the point? You're just averaging them and you'll end up somewhere in the middle, right? Just demand a bigger overlay. And that's all true.

But a simple weighted average is not the proper way to make a blended line -- the idea is to end up with an unbiased line relative to the public's. In other words, if you take your own oddsline and look at (for instance) all the horses that it has assigned 3-1 odds to (looking over many races), then we all know even if that sample wins at the overall average rate of 25% as it should that the subset of those where the public odds are below 3-1 are going to win more than those where the public odds are higher than 3-1. An unbiased line doesn't have that property, the 3-1 horses will win at 25% regardless of their public odds. A simple constant weighted linear average of the two lines will never achieve this (unless you set the weights to 100-0 in favor of the public).

So I would suggest instead of simple averaging that you actual make a new model of your line and the public line's. This is exactly what Benter originally suggested and presumably just what he did, I don't know why that always gets lost in the shuffle when talk of blending lines comes up. He said make a new logistic regression model to create a new output line. (But there are various other forms it could take, and you're not strictly limited to only those two inputs either.)

In other words build a new model using a training sample of your lines and the public's from actual races and using actual results. The output from this model for each race would then then need to be re-normalized (so the probabilities sum to 1.0 again) and/or further calibrated (sometimes spreading the probabilities is needed).

Now instead of just making some assumption about how much weight each line should have, you are using a training sample of both lines and using actual race results to predict a better number which will in some cases be higher or lower than EITHER of the input numbers (an amplification effect if you will, rather than just a blending to the middle), and then once re-normalized you'll have a (possibly) completely different line (an unbiased one hopefully). This is a much more powerful transformation. That is, it will be if the user line actually contains value-added information not contained in the public line. If it does not, the model will end up spitting out a copy of the public line. (And that's what often happens -- this will show you just what your line is really worth.)

OR, depending on how you came up with your line to begin with, you might be able to just throw the public probability in your base model as a handicapping factor, and it would have a similar effect. But in order to get an unbiased probability, you've got to deal with the public probability SOMEHOW, and a simple averaging at the end isn't it as it is just too crude...

Nice post GT....well stated.

Mike

[sjk]

I make a line for the race with no reference to the public line. At the end I determine whether the bet is an to be made by comparing it with the public line and demanding an overlay percentage.

I have not found it necessary to do any other transformation involving the two lines.

I don't see how you can say it is not good enough. I have bet 63,000+ races over a period of years with good results. In my mind it is not broken so I am not going to try to fix it.

I never actually make a linear combination of lines but in the context of scaling the bet I think it is a reasonable way to go.

[DLigett]

So I would suggest instead of simple averaging that you actual make a new model of your line and the public line's. This is exactly what Benter originally suggested and presumably just what he did, I don't know why that always gets lost in the shuffle when talk of blending lines comes up. He said make a new logistic regression model to create a new output line. (But there are various other forms it could take, and you're not strictly limited to only those two inputs either.)


GameTheory, thanks for the pointer. I just looked at Benter's "Computer Based Horse Race Handicapping and Wager Systems: A Report" and I'm confused on one point... are his alpha and beta the result of a logistic regression or some other approach?

Thanks for any insight.

[thaskalos]

I make a line for the race with no reference to the public line. At the end I determine whether the bet is an to be made by comparing it with the public line and demanding an overlay percentage.

I have not found it necessary to do any other transformation involving the two lines.

I don't see how you can say it is not good enough. I have bet 63,000+ races over a period of years with good results. In my mind it is not broken so I am not going to try to fix it.

I never actually make a linear combination of lines but in the context of scaling the bet I think it is a reasonable way to go.
Have you noticed a difference in your results in recent years over the years prior?

[sjk]

Have you noticed a difference in your results in recent years over the years prior?

Absolutely. There are fewer plays in general and far fewer among the low and medium priced horses.

[sjk]

Here is what I don't get about the step that I transform the two lines into one. Suppose I make a horse 4-1 and the public makes him 10-1. I don't see why there should be any useful information in those two numbers that would get me to a better line.

It appears to be an overlay. Sometimes I am right and sometimes I am wrong but I don't think there is a single reason or set of reasons why I would be wrong. Maybe I have made a poor speed figure for the horses last race; maybe I have totally misjudged some other horse; maybe the guys in the paddock have observed something and bet big.

There are a thousand possible reasons. Of course I could look at all of the horses I made 4-1 and the public made 10-1 over the last year and see how often they win. This strikes me as backfitting and I can't imagine there is a sustainable relationship.

Perhaps I will take a look at it when time permits and see.

[GameTheory]

I don't see how you can say it is not good enough. I have bet 63,000+ races over a period of years with good results. In my mind it is not broken so I am not going to try to fix it.That's completely fair, I was not suggesting that if you do not blend the lines using a new model you are doomed to failure -- just addressing the "what is the point of blending the lines?" question because the assumption always seems to be you'd just be averaging and ending up in the middle somewhere (which is rather pointless, or equivalent to a simple overlay threshold percentage anyway, just as you describe). However, don't knock it until you try it, the results are more interesting and profound than you might think. (And since you've had good results for so long, we know your line has some powerful non-public information in it, and this may help leverage that even more than you are. And presumably you've got a large sample of your own lines you could use to make a training sample for such a model. So something you may want to toy with if you're bored.)

Of course you need to be able to recalculate the percentages as the odds change (or very close to post time anyway) to recalc the line before betting. (Or possibly even make a probability distribution of where the final line might end up to optimally compute your bet size.) So implementation is much trickier than just setting a threshold...

[sjk]

That's completely fair, I was not suggesting that if you do not blend the lines using a new model you are doomed to failure -- just addressing the "what is the point of blending the lines?" question because the assumption always seems to be you'd just be averaging and ending up in the middle somewhere (which is rather pointless, or equivalent to a simple overlay threshold percentage anyway, just as you describe). However, don't knock it until you try it, the results are more interesting and profound than you might think. (And since you've had good results for so long, we know your line has some powerful non-public information in it, and this may help leverage that even more than you are. And presumably you've got a large sample of your own lines you could use to make a training sample for such a model. So something you may want to toy with if you're bored.)

Of course you need to be able to recalculate the percentages as the odds change (or very close to post time anyway) to recalc the line before betting. (Or possibly even make a probability distribution of where the final line might end up to optimally compute your bet size.) So implementation is much trickier than just setting a threshold...

Thanks for the suggestion. Recalculating the odds is not an issue at all. I will take a look at it.

[GameTheory]

Here is what I don't get about the step that I transform the two lines into one. Suppose I make a horse 4-1 and the public makes him 10-1. I don't see why there should be any useful information in those two numbers that would get me to a better line.Well, remember you'd be doing this on a whole race basis, not just for one horse. Some horses go up, some go down, but not in equal proportions like a simple weighted average would. So you must renormalize (and possibly an additional "spreading" calibration depending on how you make that model) and THEN you'll know how it all shakes out -- your 4-1 horse could end up as 3-1 (which cannot happen with linear averaging) and now maybe you have a bet you didn't have before (because it is a bigger overlay) or you can bet more on him, etc.

It appears to be an overlay. Sometimes I am right and sometimes I am wrong but I don't think there is a single reason or set of reasons why I would be wrong. Maybe I have made a poor speed figure for the horses last race; maybe I have totally misjudged some other horse; maybe the guys in the paddock have observed something and bet big.Since we are going to be modeling your line vs the public, it is assumed that your line is made in some consistent fashion, you can't model randomness. (For instance if you make lines for all races, but your handicapping procedure for maidens or turf races or [whatever category] is radically different than how you do things for other races, then it might be appropriate to separate those into different models.) But generally the specific reasons don't really matter.

There are a thousand possible reasons. Of course I could look at all of the horses I made 4-1 and the public made 10-1 over the last year and see how often they win. This strikes me as backfitting and I can't imagine there is a sustainable relationship.Well of course there is SOME sustainable relationship or else you'd be a perpetual loser with your current method. And again, think whole race, not one horse. I will just repeat that the results are more interesting than you'd think at first, downright magical in some cases.

It would seem the number of people that have actually tried this instead of just deciding it is not helpful beforehand is startlingly small. But of course you DO have to have a method to implement it in practice -- i.e. the resulting model has to be hooked up to a tote feed and you must bet as late as possible.

[GameTheory]

GameTheory, thanks for the pointer. I just looked at Benter's "Computer Based Horse Race Handicapping and Wager Systems: A Report" and I'm confused on one point... are his alpha and beta the result of a logistic regression or some other approach?

Thanks for any insight.
Without actually going to look at it again, from what I remember I believe all the probability models he was discussing in that paper were done with logistic regression, yes. I know later on he started using probit models just modeling the winner instead of trying to fit for all horses. I don't think I've ever heard him talk about using any non-parametric methods (which is my personal preference) -- just standard logit models...

[davew]

Here is what I don't get about the step that I transform the two lines into one. Suppose I make a horse 4-1 and the public makes him 10-1. I don't see why there should be any useful information in those two numbers that would get me to a better line.

It appears to be an overlay. Sometimes I am right and sometimes I am wrong but I don't think there is a single reason or set of reasons why I would be wrong. Maybe I have made a poor speed figure for the horses last race; maybe I have totally misjudged some other horse; maybe the guys in the paddock have observed something and bet big.

There are a thousand possible reasons. Of course I could look at all of the horses I made 4-1 and the public made 10-1 over the last year and see how often they win. This strikes me as backfitting and I can't imagine there is a sustainable relationship.

Perhaps I will take a look at it when time permits and see.

If you are just betting win pool, then probably does not matter much
but what if you are betting exactas and calculating probabilities for fair pay-offs?

[sjk]

In practice 90% of my bets are exactas. The way I size bets is close to the same.

I have out of town business for a few days but I will try to look at some data on the weekend.

[Capper Al]

Good discussion and a bit over my head. I need to research the mentioned models. I am working on changing my play from value only to value and tote-board analysis. High odds alone might mean that I am missing something on a horse or two in the race. With the public being so astute today at about 38% winners, there is little room left to strike on value. One has to be more accurate when going for the hit.

[TravisVOX]

Good discussion going here and I appreciate the responses both here and via PM. I'm going to kick the tires on the various suggestions and see what surfaces. This particular project is different from the other modeling I'm involved in where I have utilized Benter's approach of merging model with public. Thanks all!

[classhandicapper]

Imagine 2 races.

In the first race I am dealing with high quality older horses that have been running steadily on the same surface and at a similar distance to today's race. I have seen replays of all their races, have a good set of figures, I am very confident I know these horses well, and I have a record of being very good at this kind of race etc...

In the second I am dealing with a field of lightly raced horses whose form has tended to be less stable. A couple have been moving back and forth between sprints/routes and dirt/turf clouding their form. A couple are returning off a layoff. One is taking a suspicious drop in class that can only be explained by insiders. One is shipping in from another circuit.

In the first instance if I give some horse a 25% chance of winning, I might take 7-2 and definitely 4-1. I don't care what the board says.

In the second instance, if I give some horse a 25% chance of winning, I might not take 5-1. I will also be more interested in what the board says because some people might have insights about the layoffs, distance, and surface preferences of some of these horses as insiders than I have from the public record.

IMO, knowing when your line is probably very accurate and when it might not be is a key to profits. As a gambler, you are not required to play every race or even produce a line for every race. You can turn the page.

If I understand what is being suggested (and I agree it could be enlightening) whatever new formula you create for your line vs. the public line would be applied equally in both these races. That might be right long term covering 1000s of races, but it's not right for these 2 races.

IMO, anything you can do to improve your ability to estimate the chances of each horse in the race and the confidence level you have in that line is a great idea, but once you are there the board is just a market and should only rarely be used to adjust your own thinking.

[DeltaLover]

I do not believe that building a model to estimate the probabilities of each horse to win, is going to work in the real word.

In theory, it seems simple and straightforward: Compare the odds against your “correct” line, bet the overlays and watch your bankroll growing to the sky!

In the real world though, things are a bit more complicated..

First of all you have no way of knowing what value is offered by the pools until they are closed; in a situation where a 5.5-1 might be an overlay but a 4-1 a miserable underlay, it easy to realize that even if we assume that our odds are always correct, still this is not enough... By the way, I did not select these odds randomly; in the game of dice, the proposition bet for SEVEN pays exactly 4-1, a value that converts it to one of the worst bets you can find in a casino, since the fair odds are 5-1...

Secondly, in the vast majority when you will be comparing your line to the crowd, whatever will be looking like an “overlay” will just be quite the opposite, a case that can be partially explained due to some handicapping anomalies, like a sick or over matched horse, factors that probably cannot be estimated properly by the model.

Without having the ability to prove it, I am still convinced, that the very best models will behave closely to the public, pretty much resembling its opinions... It might even be possible to build a model slightly better than the public, but even in this case chances are that it will still be losing, since the takeout almost impossible to overcome.

[classhandicapper]

I do not believe that building a model to estimate the probabilities of each horse to win, is going to work in the real word.

In theory, it seems simple and straightforward: Compare the odds against your “correct” line, bet the overlays and watch your bankroll growing to the sky!

In the real world though, things are a bit more complicated..

Secondly, in the vast majority when you will be comparing your line to the crowd, whatever will be looking like an “overlay” will just be quite the opposite, a case that can be partially explained due to some handicapping anomalies, like a sick or over matched horse, factors that probably cannot be estimated properly by the model.



My way around this kind of thing is to try to insist that I can identify the reason this horse is an overlay also.

In other words, it's often not enough for me to think a horse has a 25% chance of winning and then take 4-1.

I have to be able say:

"this horse is 4-1 because the crowd does not realize he raced on a dead rail in his last 2 starts and he's actually sharper than he looks".

"this horse is 4-1 because the crowd does not realize that his last speed figure is highly suspect and I have concluded through other means that the race as stronger than it looks".

"this horse is 4-1 because the favorite has been over hyped in the media and is nowhere near as good as his reputation".

[DeltaLover]

My way around this kind of thing is to try to insist that I can identify the reason this horse is an overlay also.

In other words, it's often not enough for me to think a horse has a 25% chance of winning and then take 4-1.

I have to be able say:

"this horse is 4-1 because the crowd does not realize he raced on a dead rail in his last 2 starts and he's actually sharper than he looks".

"this horse is 4-1 because the crowd does not realize that his last speed figure is highly suspect and I have concluded through other means that the race as stronger than it looks".

"this horse is 4-1 because the favorite has been over hyped in the media and is nowhere near as good as his reputation".

Great post from someone who seems to understand the game in depth! :ThmbUp: :ThmbUp:

This is exactly how the advance handicapper should think about the game, this is a step towards meta-handicapping which IMHO has the potential of solving the game.

[TrifectaMike]

This is exactly how the advance handicapper should think about the game, this is a step towards meta-handicapping which IMHO has the potential of solving the game.

Interesting DL, but isn't meta-handicapping just a route to a different type of model? Logistic or Probit are not the only models available.

Mike

[TrifectaMike]

Without actually going to look at it again, from what I remember I believe all the probability models he was discussing in that paper were done with logistic regression, yes. I know later on he started using probit models just modeling the winner instead of trying to fit for all horses. I don't think I've ever heard him talk about using any non-parametric methods (which is my personal preference) -- just standard logit models...

GT, just a guess. Do your prefer stacking?

Mike

[DeltaLover]

Interesting DL, but isn't meta-handicapping just a route to a different type of model? Logistic or Probit are not the only models available.

Mike

Sure.

What differs is the output of the model, which does not reflect the probability of each starter to win, but the probability of the crowd making a significant error in its assessments.

This approach, assumes the need of a metric to measure the normality of the crowd's prediction and consists on the development of an algorithm to predict how possible is for the crowd to deviate from its average precision, given a specific event as it is described by all its known factors.

The objective is not to try to detect discrepancies based on a comparison of a 'correct' line vs what is provided from the crowd, but to begin with the assumption that the crowd has a p probability of committing an error in its final forming of the pools...

I am not sure that I explain clearly what I have in mind though... Please let me know if my wording seems confusing.

[TrifectaMike]

Sure.

What differs is the output of the model, which does not reflect the probability of each starter to win, but the probability of the crowd making a significant error in its assessments.

This approach, assumes the need of a metric to measure the normality of the crowd's prediction and consists on the development of an algorithm to predict how possible is for the crowd to deviate from its average precision, given a specific event as it is described by all its known factors.

The objective is not to try to detect discrepancies based on a comparison of a 'correct' line vs what is provided from the crowd, but to begin with the assumption that the crowd has a p probability of committing an error in its final forming of the pools...

I am not sure that I explain clearly what I have in mind though... Please let me know if my wording seems confusing.

DL, sounds a lot like Bayesian to me.

Mike

[raybo]

DL, sounds a lot like Bayesian to me.

Mike

Sounds much more like a huge, very fast processing database program, using a huge number of factors and combinations of factors as inputs, that are all run through the database, with the output being the percentage that the public is wrong for each factor and factor combination that were used as inputs.

The key, of course, being the knowledge of which individual race circumstances the public is most likely to be wrong, and exploiting only those races.

[GameTheory]

GT, just a guess. Do your prefer stacking?
If modeling methods were food, stacking would be a condiment, not a meal, so that wouldn't really make sense, no...

[classhandicapper]

The key, of course, being the knowledge of which individual race circumstances the public is most likely to be wrong, and exploiting only those races.

Exactly. You slowly build a portfolio of "potential plays". Then you handicap the race and see what the odds are.

I don't do this through computer research or a model. I do this manually by trial and error. But this is an area where I wish I had the data and a good research tool available. I'd find more potential plays. I'd find them faster and I'd make fewer mistakes during the learning process.

[DeltaLover]

Interesting DL, but isn't meta-handicapping just a route to a different type of model? Logistic or Probit are not the only models available.

Mike


Of course, it is bayssian although at the current stage of my development. I am not applying a formal application of Bayesian Statistics. I probably need to spend more time understanding how I can apply the theory to the practice, since at this point I have to admit that I do not know exactly how to do it...

Some related bibliography would certainly be appreciated!

[raybo]

Exactly. You slowly build a portfolio of "potential plays". Then you handicap the race and see what the odds are.

I don't do this through computer research or a model. I do this manually by trial and error. But this is an area where I wish I had the data and a good research tool available. I'd find more potential plays. I'd find them faster and I'd make fewer mistakes during the learning process.

I am only guessing at what DL is actually doing with his "meta-handicapping" approach, but if you are looking at a race type that the public's favorite almost never wins, wouldn't that be an excellent opportunity to bet against the favorite, with both fists

[DeltaLover]

Sounds much more like a huge, very fast processing database program, using a huge number of factors and combinations of factors as inputs, that are all run through the database, with the output being the percentage that the public is wrong for each factor and factor combination that were used as inputs.

The key, of course, being the knowledge of which individual race circumstances the public is most likely to be wrong, and exploiting only those races.

Not really Raybo, not really...

I never believed that have a huge number of factors and combinations is the way and in my systems I always try to keep things as simple as possible. The main agent in the model is based on pace handicapping, while there are some additional signals generated by some very basic handicapping factors, like layoff days, class movement, connections changes and weight.

Of course it is true that a very fast program can never hurt but in my case this is easily accomplished by using the proper hardware and been careful about the design of your algorithms..

Using my new box (https://system76.com/desktops/model/leox4), I am now able to load in memory a complete universe of 20,000 races including full BRIS DRF past performances and results and process the race in R/T; the good thing is, that I had to go through very limited modification of my original code that original was written for much weaker hardware... The main change, had to do with parallel processing, so I can take full advantage of the multiple cores, something that took me a few weeks to accomplish...

[classhandicapper]

I am only guessing at what DL is actually doing with his "meta-handicapping" approach, but if you are looking at a race type that the public's favorite almost never wins, wouldn't that be an excellent opportunity to bet against the favorite, with both fists

Yes. I approach it from both directions. In some cases I find horses I think the public will underestimate and in others horses that the public is likely to overestimate. In the latter case, I then sometimes have the perplexing problem of determining who to actually play against the false favorite if no one is especially appealing. I am not shy about betting 2 horses and occasionally even 3 horses in those situations. The ideal situation is when you have 2 positives for the same horse or a positive on one horse and a negative on another inside the same race.

[TrifectaMike]

Of course, it is bayssian although at the current stage of my development. I am not applying a formal application of Bayesian Statistics. I probably need to spend more time understanding how I can apply the theory to the practice, since at this point I have to admit that I do not know exactly how to do it...

Some related bibliography would certainly be appreciated!

DL, recall the heirarchical model I sent to you (the actual code) vary the observations for that model and observe how the posterior probabilities change. Don't concern yourself at this stage with priors or likelihoods for your particular problem. Get a good understanding of that particular model, it will go a long way in the search for your particle structure.

I'll help you, when I can.

Mike

[DeltaLover]

DL, recall the heirarchical model I sent to you (the actual code) vary the observations for that model and observe how the posterior probabilities change. Don't concern yourself at this stage with priors or likelihoods for your particular problem. Get a good understanding of that particular model, it will go a long way in the search for your particle structure.

I'll help you, when I can.

Mike


Thx Mike, I will revisit the code by next week and let you know about questions!

[sjk]

Thanks GT. I think I am seeing something interesting in the data. The real chance of winning is always between my odds and the public odds but the shorter the price, the closer it is to the public odds and the longer the price the more likely it is to be closer to mine.

Should be easy enough to put into practice if further testing bears it out.

[GameTheory]

Thanks GT. I think I am seeing something interesting in the data. The real chance of winning is always between my odds and the public odds but the shorter the price, the closer it is to the public odds and the longer the price the more likely it is to be closer to mine.

Should be easy enough to put into practice if further testing bears it out.Yes of course it IS almost always in-between those values (especially if they have much spread), but sometimes (usually at the tails) you get something that is higher or lower than either (like if both lines have a horse as a heavy favorite it may be even better than that). Or if your line is highly uncorrelated with the public weird things can shake out.

But still, even if it remains in the middle you are seeing the weight of the effect is not constant, which is expected also...

[sjk]

I expected it to be constant which would be in keeping with using a constant overlay requirement so it is useful to see that it is not.

The relationship is clearly apparent and I am using a large data set so I think it is real and I can put it to profitable use.

It looks like I should be using something close to 60-40 on the short end and 45-55 on the long end,

[raybo]

I expected it to be constant which would be in keeping with using a constant overlay requirement so it is useful to see that it is not.

The relationship is clearly apparent and I am using a large data set so I think it is real and I can put it to profitable use.

It looks like I should be using something close to 60-40 on the short end and 45-55 on the long end,

Can you explain "60-40 on the short end" and "45-55 on the long end". I think I know what you mean but not sure.

Thanks!

[sjk]

I will think about using 60% of the public odds and 40% of my odds line on the one end and 45% public plus 55% my line on the other end. Looks like that will be helpful but I have not got far enough to see what the expected benefit will be.

[Cratos]

I will think about using 60% of the public odds and 40% of my odds line on the one end and 45% public plus 55% my line on the other end. Looks like that will be helpful but I have not got far enough to see what the expected benefit will be.

Aren’t you mixing “apples and oranges” when you mix the public odds with your odds? To you the public odds will be random odds and your odds to you will be conditional odds.

Given that understanding I don’t see what you expect to achieve.

I personally don’t buy the “merging probability philosophy” although it might be academic intriguing I feel better by setting my own odds and forcing the public to beat me.

For example, my minimum wagering odds are 3-1 and if me and the public like the same horse and they have pushed odds below my 3-1 odds betting threshold I will pass the race. On the other hand, if the public let my choice go at 3-1 or higher I am making my bet.

[Jeff P]

Using my own (large) data samples:

If I take probability arrived at by merging 1. Public (after the odds are known or as the odds are known close to post time) probability with 2. Private (based on factors in my own private model) probability and plot the result graphically:

The result isn't linear.

It's actually an S shaped curve.

I'm not saying you can't use a simple (linear) weighted average to merge the two. Obviously you can if you want to.

But that's hardly the best way to do it.

A simple linear weighted average (60/40, 70/30, 80,20, what have you, etc.) might produce a good fit in one area of your curve only. But it won't (or can't) produce anything close to an optimal fit across the entire length of your curve.

If you are going to use linear math - and you might want to or have to for sake of simplicity - then I think you have to look at the shape of your curve - especially the way the slope changes as you get closer to the edges - and apply different ratios for the different regions of your curve.

My experience is that if you do it that way you'll get better results than if you apply a single ratio across the entire length of your curve.

But again, that's hardly an optimal way to merge probability estimates. (Like Gametheory mentioned back in post #6 of this thread, the optimal way involves modeling the resulting probability estimate.)

Several years ago, I went looking for a close to optimal way of merging multiple probability estimates, and came across a pdf at the Univ of Washington website:
https://www.stat.washington.edu/research/reports/2008/tr543.pdf

The math at the above link is a bit involved. But the author does present a methodology for merging distinct probability estimates in way that (in my opinion) produces results that are much better than using linear math.



-jp

.

[Overlay]

I realize that there will always be information and knowledge reflected in the public odds that I don't/can't know or have access to. But I've found that if I can develop my own estimate of a horse's winning probability using statistical/mathematical models or factor combinations that have proven their long-term predictive reliability over time (even if the data that they incorporate is generally available), I can make confident handicapping/wagering judgments based solely on a face-value comparison of that probability with the horse's odds, without having to address the issue of how the odds themselves or their possible underlying rationales should influence or modify my decision-making in any individual race.

[GameTheory]

Why the extreme resistance to actually empirically trying this? Although it can be tricky to implement in real time, if you make oddslines (and don't immediately throw them away forever), then it is pretty easy to study with a sample of past races. If you make an oddsline without reference to the public odds, this ALWAYS improves it -- always (over a sample). (And is always better than "just demand X% overlay method".) To "not buy it" without trying it is just silly.

Now the claim, "I don't need it to win" may well be valid, but that's different. Just because you're not using a piece of useful value-adding information and winning anyway is no indictment of that piece of information.

[Cratos]

Why the extreme resistance to actually empirically trying this? Although it can be tricky to implement in real time, if you make oddslines (and don't immediately throw them away forever), then it is pretty easy to study with a sample of past races. If you make an oddsline without reference to the public odds, this ALWAYS improves it -- always (over a sample). (And is always better than "just demand X% overlay method".) To "not buy it" without trying it is just silly.

Now the claim, "I don't need it to win" may well be valid, but that's different. Just because you're not using a piece of useful value-adding information and winning anyway is no indictment of that piece of information.

GT, I went back and re-read your post (#6) in this thread and I read the PDF supplied by JP and still have the same problem: " Where is the value added?"

This is not to say "I don't need it to win," but it is to say unless I am missing something I fail to see additional odds value; subsequently better payoffs

[GameTheory]

GT, I went back and re-read your post (#6) in this thread and I read the PDF supplied by JP and still have the same problem: " Where is the value added?"

This is not to say "I don't need it to win," but it is to say unless I am missing something I fail to see additional odds value; subsequently better payoffsYou make an oddsline, you find it useful. I can make it better. That is not useful? What does a better oddline get you?

-- It is unbiased -- the horses win at the rates they are predicted to by the probabilities regardless of the public odds. (I absolutely guarantee any line that doesn't account for the public line does not do this.)

-- Optimal bet-sizing. Does the amount of "edge" you perceive to have based on your line affect how much you bet for that race? If so, than that is reason enough for an improved line.

-- It will put you off bad bets. Some of your overlays are not really.

-- It will put you on good bets. Some of your perceived underlays or break-even horses may be overlays after all.

What's not to like? The resistance to this is almost pathological, I don't get it. Presumably those of you that make full race lines (that work) worked hard to do so you're making it as good as you can. This is GUARANTEED to improve it (assuming you have not already accounted for the public probabilities otherwise), but the response is almost always, "No thanks, why would I want to do that?" or "The public line?! Boo! What kind of handicapper cares what the *public* thinks?"

Is it just ego or what? I'll have to check, but if memory serves in Benter's paper he said something along the lines that adding this final step was THE thing that put him over the top so they could really exploit the pools. But he only made millions, so he could be full of it...

[Cratos]

You make an oddsline, you find it useful. I can make it better. That is not useful? What does a better oddline get you?

-- It is unbiased -- the horses win at the rates they are predicted to by the probabilities regardless of the public odds. (I absolutely guarantee any line that doesn't account for the public line does not do this.)

-- Optimal bet-sizing. Does the amount of "edge" you perceive to have based on your line affect how much you bet for that race? If so, than that is reason enough for an improved line.

-- It will put you off bad bets. Some of your overlays are not really.

-- It will put you on good bets. Some of your perceived underlays or break-even horses may be overlays after all.

What's not to like? The resistance to this is almost pathological, I don't get it. Presumably those of you that make full race lines (that work) worked hard to do so you're making it as good as you can. This is GUARANTEED to improve it (assuming you have not already accounted for the public probabilities otherwise), but the response is almost always, "No thanks, why would I want to do that?" or "The public line?! Boo! What kind of handicapper cares what the *public* thinks?"

Is it just ego or what? I'll have to check, but if memory serves in Benter's paper he said something along the lines that adding this final step was THE thing that put him over the top so they could really exploit the pools. But he only made millions, so he could be full of it...

Excellent synopsis and I appreciate your diligence in responding.

[GameTheory]

And one more benefit for those that (unlike sjk and Cratos) have not already been winning for years with their line. Most people when they try to create an oddsline will come up with something that seemingly is spot-on and they are convinced they are about to go make a ton of money betting on the overlays. (Remember that one guy here who wrote a whole book about how awesome his value line was, but he never ever once actually checked their performance betting overlays? Of course I think he is still selling picks so I guess he found out how to make money after all.)

Anyway, making a model like this with a healthy sample of lines will show you if your line is any good at all, i.e. if it is adding value to the public line. Many will find that the output is more or less an exact copy of the public line, meaning that your line is totally useless...

[TravisVOX]

Handicapping/betting at any level is a battle for information, but this is particularly true at the "computer model" level. I was reading a book the other day where it said one pro gambler follows all of the Twitter accounts for the players in his particular sport just in case a nugget of useful information comes from a tweet here or there. Obviously an extreme example, but none-the-less, it shows how all of the information (including public odds) that you have available are added value to whatever you're trying to do.

[classhandicapper]

What's not to like? The resistance to this is almost pathological, I don't get it. Presumably those of you that make full race lines (that work) worked hard to do so you're making it as good as you can.

For me, it appears to add no value over using the margin of safety concept borrowed from Warren Buffett in stock investing.

Suppose I make a horse 2-1 and he's 8/5 on the board. Obviously, blending that doesn't help. I'm not playing the horse either way.

Suppose I make a horse 2-1 and he's 5-2 on the board. Blending that doesn't help. I'm not playing the horse. 5-2 is an overlay, but there's not enough of a margin of safety for my own potential error or lack of information.

Suppose I make a horse 2-1 and he's 4-1 on the board. Blending that doesn't help. I AM playing the horse either way. 4-1 IS enough of a margin of safety to overcome my own potential errors or lack of information.

I suspect many people are intuitively thinking in terms of blending their own line with the actual odds. They are simply acknowledging their own limitations by insisting on a margin of safety before betting even if the horse appears to be a mild overlay. The trick is knowing how much of a margin of safety to use and whether your own assessments have any value at all. I think that is typically developed intuitively based on how confident they are in their thinking about the race in question and their long term results.

[GameTheory]

For me, it appears to add no value over using the margin of safety concept borrowed from Warren Buffett in stock investing.

Suppose I make a horse 2-1 and he's 8/5 on the board. Obviously, blending that doesn't help. I'm not playing the horse either way.

Suppose I make a horse 2-1 and he's 5-2 on the board. Blending that doesn't help. I'm not playing the horse. 5-2 is an overlay, but there's not enough of a margin of safety for my own potential error or lack of information.

Suppose I make a horse 2-1 and he's 4-1 on the board. Blending that doesn't help. I AM playing the horse either way. 4-1 IS enough of a margin of safety to overcome my own potential errors or lack of information.

I suspect many people are intuitively thinking in terms of blending their own line with the actual odds. They are simply acknowledging their own limitations by insisting on a margin of safety before betting even if the horse appears to be a mild overlay. The trick is knowing how much of a margin of safety to use and whether your own assessments have any value at all. I think that is typically developed intuitively based on how confident they are in their thinking about the race in question and their long term results.No problem there, but what I'm talking about is not applicable to your style of play anyway. Nevertheless, I REPEAT that what I am proposing is not so simple, and so the statements you made above "blending doesn't help here", etc are also not applicable. [to the method I am talking about] So just to avoid confusion and put it bluntly, you can't use this method (because your style of play doesn't allow it), but because of that you also can't you comment about this method from the context of your own play because that context is irrelevant.

Hmmm...still sounds confusing. (And a bit rude, but I don't mean it that way.)

[TrifectaMike]

I read these posts and empathize with the posters, because I know how difficult it is to think nonlinearly.

Mike

[Robert Goren]

I have always thought that underlays are the way to go. Money bet on factors that I and almost every other handicapper out there have no way of knowing is under bet. The best "overlays" are what appear to be "underlays".The more I look at the numbers, the more sure I become of that.

[GameTheory]

I have always thought that underlays are the way to go. Money bet on factors that I and almost every other handicapper out there have no way of knowing is under bet. The best "overlays" are what appear to be "underlays".The more I look at the numbers, the more sure I become of that.
Underlays win more, no doubt.

But just so I'm following, maybe you can expand what you mean?

[sjk]

If any of you are interested in my progress I have put together something that appears to add a couple of points of value.

I have never had a line that beat the public line on the information metric GT posted many years ago but this one appears to do so by a whisker. I need to define and test a specific implementation but that won't be hard. I think I will leave that to tomorrow and play some races today.

I have been playing the same way for 10 years so maybe it is time for a change. Thanks for the help.

[lansdale]

No problem there, but what I'm talking about is not applicable to your style of play anyway. Nevertheless, I REPEAT that what I am proposing is not so simple, and so the statements you made above "blending doesn't help here", etc are also not applicable. [to the method I am talking about] So just to avoid confusion and put it bluntly, you can't use this method (because your style of play doesn't allow it), but because of that you also can't you comment about this method from the context of your own play because that context is irrelevant.

Hmmm...still sounds confusing. (And a bit rude, but I don't mean it that way.)

Hi GT,

I admire what you're trying to do here -- the truth and value of the two-step model has been demonstrated in the real world a number of times post-Benter, but this is clearly counterintuitive for many. Also, I think, out of decorum, you imply that handicappers have some value to add to the 'public' model, when, in most cases, they don't. And, as you explained, even traditional handicappers who do have an edge, can't use this technique unless they're capable of constructing a model.

Handicappers, especially those who are successful, often have an irrational confidence in their own ideas, which prevents them from being aware of the kind of bias that Benter was taking such pains to eliminate from his model, even though even his supposedly 'biased' fundamental model was vastly more accurate than that of most handicappers.

Possibly the example of Nate Silver's success - taking the mean of several of the least biased political polls is far more predictive than any individual poll - might make this clearer to some.

Thanks again for your insights.

Cheers,

lansdale

[classhandicapper]

No problem there, but what I'm talking about is not applicable to your style of play anyway. Nevertheless, I REPEAT that what I am proposing is not so simple, and so the statements you made above "blending doesn't help here", etc are also not applicable. [to the method I am talking about] So just to avoid confusion and put it bluntly, you can't use this method (because your style of play doesn't allow it), but because of that you also can't you comment about this method from the context of your own play because that context is irrelevant.

Hmmm...still sounds confusing. (And a bit rude, but I don't mean it that way.)

Not rude at all. I may be missing the point. Just because my approach is different does not mean I'm not intellectually curios.

If you blend the odds you may create a more accurate line, but all you are really doing is changing the degree of overlay or underlay for each horse. When you blend, no horse will suddenly move from overlay to underlay or vice versa.

My Line

1. 50%
2. 25%
3. 10%
4. 15%

Public Line

1. 40%
2. 30%
3. 30%
4. 5%

Blended at the average (just for demonstration purposes).

1. 45%
2. 27.5%
3. 20%
4. 10%

On my original line there were 2 overlays. On the blended line there are still 2 overlays (#1 and #4). That will always be the case. By blending the odds all I did was change the degree of overlay or underlay to make it more accurate. I could just as easily take my original line and say the horse has to be overlaid by a certain amount before taking the bet and I'm in the same place.

I would probably also quickly learn that I need a greater margin of safety in some odds ranges than others (less linear than just averaging them).

For example, I know that an estimate of 5% could easily actually be 3% because it's tough to get that fine with just a couple of percent either way. So 25-1, 30-1, or even 35-1 might not be enough to bet even though I am calling the horse 19-1.

But with an estimate of 50% (even money), I might be able to take 8-5 with confidence because 40% is still good enough.

[raybo]

Underlays win more, no doubt.

But just so I'm following, maybe you can expand what you mean?

Maybe it's a just matter of "his" line not being the best. What "appear" to him to be overlays, when comparing to the public's odds, may not be. So, that would bias his opinions of his overlay perceptions, making "his" underlays perform better than "his" overlays.

No offense intended, of course, good lines are very difficult to achieve.

[Robert Goren]

Underlays win more, no doubt.

But just so I'm following, maybe you can expand what you mean?If you can devise a odds line that reflects the way the horses will be bet off their Past Performances(easier said than done) then if the tote odds are lower than the "PP odds" by quite a bit, then you have the likely winner. It is my contention that the amount that tote odds lower is not large enough to reflect the actual odds of winning. A case might be like this. A horse with the "PP odds" of 3/1 (a 25% chance of winning) is bet down to 3/2(40% chance of winning) would actually have a 50% of winning (even money odds) and would be a good bet. I am currently working on a math model that would predict the "PP odds" using Prime Power numbers and a math model that predict how much a horse's chances of winning increase by how much lower its "tote odds" compared to the "PP odds". I should be able to post some very preliminary findings in the couple of weeks. It is my hope to have something workable by the AQU inner meet since my data comes from race ran there.

[sjk]

I have always found that underlays run better than overlays and I have seen several others here in agreement with that finding.

The issue is whether that happens in a biased way that can be studied and understood or an unbiased way that you have to live with. The suggestion is that you can handicap your own line to identify and adjust for systematic biases.

It is probably hubris that leaves one to think that any error in analysis is due to non-random factors. I have always thought that if you handicap every factor to the best of your ability there can be little to analyze.

I am looking at it more openly here and seem to find a benefit at first look.

[raybo]

I have always found that underlays run better than overlays and I have seen several others here in agreement with that finding.

The issue is whether that happens in a biased way that can be studied and understood or an unbiased way that you have to live with. The suggestion is that you can handicap your own line to identify and adjust for systematic biases.

It is probably hubris that leaves one to think that any error in analysis is due to non-random factors. I have always thought that if you handicap every factor to the best of your ability there can be little to analyze.

I am looking at it more openly here and seem to find a benefit at first look.

Good post! I agree that if you do a very good job of handicapping all the factors you have access to, there can be made an argument that no further analysis is needed. And, if you can do all that with a computer, then you effectively have a decent handicapping "black box". All that is left to do is the wagering decisions (and getting your line adjusted properly, for each race, or for all races (which is what was being discussed). A very good line goes a long way in solving the pass/play and wager type/wager amount questions.

[davew]

For some reason this discussion reminds me of the Monty Hall problem ->

Suppose you're on a game show, and you're given the choice of three doors: Behind one door is a car; behind the others, goats. You pick a door, say No. 1, and the host, who knows what's behind the doors, opens another door, say No. 3, which has a goat. He then says to you, "Do you want to pick door No. 2?" Is it to your advantage to switch your choice?

http://en.wikipedia.org/wiki/Monty_Hall_problem

[raybo]

For some reason this discussion reminds me of the Monty Hall problem ->

Suppose you're on a game show, and you're given the choice of three doors: Behind one door is a car; behind the others, goats. You pick a door, say No. 1, and the host, who knows what's behind the doors, opens another door, say No. 3, which has a goat. He then says to you, "Do you want to pick door No. 2?" Is it to your advantage to switch your choice?

http://en.wikipedia.org/wiki/Monty_Hall_problem

Of course, your probability doubles if you switch doors, but how does that example relate to what is under discussion here?

[GameTheory]

Not rude at all. I may be missing the point. Yes, you are. (Please continue to consider me "not rude" if I continue to sound blunt.)

If you blend the odds you may create a more accurate line, but all you are really doing is changing the degree of overlay or underlay for each horse. When you blend, no horse will suddenly move from overlay to underlay or vice versa.Go back and re-read. This is precisely the intuitive reaction I'm trying to get people to abandon because it is not true. This is the difference being linear and non-linear and whole race adjustment/renormalize and simple horse-by-horse weighted (where you get the same result either way as long as your linear weights add up to 1.0).

So, yes NEW overlays can appear, and yes NEW underlays can appear.


My Line

1. 50%
2. 25%
3. 10%
4. 15%

Public Line

1. 40%
2. 30%
3. 30%
4. 5%

Blended at the average (just for demonstration purposes).

1. 45%
2. 27.5%
3. 20%
4. 10%

On my original line there were 2 overlays. On the blended line there are still 2 overlays (#1 and #4). That will always be the case. By blending the odds all I did was change the degree of overlay or underlay to make it more accurate. I could just as easily take my original line and say the horse has to be overlaid by a certain amount before taking the bet and I'm in the same place.
That's exactly right, but is not at all what I'm suggesting. This is simple linear blending, just a weighted average, and it behaves exactly as you say. If I was good at making plotted diagrams I'd put one in here showing two overlaid curves and the new resulting line from the model, but I'm not good at that. What I'm saying is you make a model of "what actually happens" given your line and the public line. So maybe when you rate something as having a 33% chance of winning and the public also rates it at 33% chance of winning, it actually wins at at rate of 40%. NEW overlay!

Now then, it IS true of course that most of the time, the resulting probability from this new model will in fact fall in the middle of your line and the public (but that is still beneficial for bet sizing and margin of error reasons), BUT you will find, as sjk just did, that the weights between the two are not constant. (Again, think overlaid curves instead of parallel lines.)

[GameTheory]

If you can devise a odds line that reflects the way the horses will be bet off their Past Performances(easier said than done) then if the tote odds are lower than the "PP odds" by quite a bit, then you have the likely winner. It is my contention that the amount that tote odds lower is not large enough to reflect the actual odds of winning. A case might be like this. A horse with the "PP odds" of 3/1 (a 25% chance of winning) is bet down to 3/2(40% chance of winning) would actually have a 50% of winning (even money odds) and would be a good bet. I am currently working on a math model that would predict the "PP odds" using Prime Power numbers and a math model that predict how much a horse's chances of winning increase by how much lower its "tote odds" compared to the "PP odds". I should be able to post some very preliminary findings in the couple of weeks. It is my hope to have something workable by the AQU inner meet since my data comes from race ran there.Yes, I think that is right. Years ago I posted here about how betting against the morning line as-if it were the actual odds was not as laughable as it sounds for similar reasons.

[raybo]

Yes, I think that is right. Years ago I posted here about how betting against the morning line as-if it were the actual odds was not as laughable as it sounds for similar reasons.

Doesn't that assume that the M/L maker is basing his line on the handicapping of the horses' data only, and not his opinion of how the public should bet?

Or, are you saying that all of the public will bet according to the horses' data only? In which case, the morning line, if it comes from an expert handicapper, should be about the same as the actual final odds.

I'm confounded that anyone would use the M/L, unless they know that the M/L maker is an expert mimic of the way the public bets. Maybe you can convince me?

[classhandicapper]

What I'm saying is you make a model of "what actually happens" given your line and the public line. So maybe when you rate something as having a 33% chance of winning and the public also rates it at 33% chance of winning, it actually wins at at rate of 40%. NEW overlay!


OK, now I understand.

If that happens, then I think what you really need to do is fix your handicapping. You are identifying something neither of you is building in properly in a certain odds range. So figure out what it is and incorporate it.

[thaskalos]

If we blend our betting line with the crowd's in equal measure, aren't we, in fact, admitting that our opinion is no better than that of the public's?

[GameTheory]

Doesn't that assume that the M/L maker is basing his line on the handicapping of the horses' data only, and not his opinion of how the public should bet?

Or, are you saying that all of the public will bet according to the horses' data only? In which case, the morning line, if it comes from an expert handicapper, should be about the same as the actual final odds.

I'm confounded that anyone would use the M/L, unless they know that the M/L maker is an expert mimic of the way the public bets. Maybe you can convince me?The idea is that the ML influences the way the public bets. In this era of whale money, I'm not sure if that influence is still strong enough, but in years past the ML was at least partially a self-fulfilling prophecy. So you know people would see a horse was 2-1 on the ML, but at 10 MTP he's only 8-1 or something. Just the fact that his ML is low will attract money to him as non-sophisticated players start to think overlay.

But what if the ML is say 4-1 and the horse is bet down to 2-1 on the board? Even though he is a big "ML underlay" there is some portion of the public that will now try to beat him not only because his odds are low but because that 4-1 ML is bugging them. So what I'm saying is that same horse, if his ML was 2-1 instead of 4-1 would probably be bet down even more to 7/5 to 6/5. And the people betting him down usually have good reasons, and so it is now likely that that horse is a true overlay at 2-1 because he should be 7/5 and the 4-1 ML line is propping him up by influencing the betting. Follow?

And so, as a believer in using the public odds to alter my own lines, I have no problem adjusting my probability upward on a horse that the public is rating higher than me. So what happens when I bet against the ML instead of the real line? Well, maybe I rate the horse at 2-1. His ML is 4-1. The ML is the proxy for the odds, so that makes him an overlay. But now in real life he's bet down to 2-1 making it a break even or underlay. So I've just screwed myself, right? I'm making a bad bet by using the ML. Well no, because as I just explained, if he got bet down to 2-1 against a 4-1 ML, he had to "overcome" that 4-1 to do it and he really should be 7/5. Which makes him an overlay, and a good bet. (I LOVE low priced overlays.)

The reverse situation where the horse goes off even higher than his morning line, which should mean super overlay, was the more problematic situation as it often meant (if I had the horse at lower odds, meaning I would bet him against that ML) I had overestimated his chances and he was a false overlay, or would win very seldom anyway.

This is something I actually used to do so I didn't have to grind all day, just use the ML in the morning, make my bets, and that was that. AND I used the ML as if it were the public line to model "what actually happens" as described earlier, so I was actually betting against THAT line, which was already adjusted by the ML. (Further advantage of the ML is that there is no guessing what the odds are going to be, they are fixed.) So the whole method was tuned for using the ML. And it worked, and it puts you onto those low-priced overlays that a traditional real-odds approach will keep you away from. But whether it would still work now, I just don't know if the ML has that much influence over the odds anymore...

[GameTheory]

OK, now I understand.

If that happens, then I think what you really need to do is fix your handicapping. You are identifying something neither of you is building in properly in a certain odds range. So figure out what it is and incorporate it.
I would submit that although you can always improve your handicapping, there will always be information in the public line that you will never know. In short, it is impossible to "fix your handicapping" EXCEPT for doing this -- THIS fixes it so just do this.

I will state again that I absolutely guarantee that not a single human being in the world can make an unbiased line (over many races) if they do not account for the public odds in some way in their handicapping. An unbiased line is one where the 3-1 (or whatever) horses win at 25% when the public has got them at 6/5 and they still win at 25% when the public has got them at 10-1. You will never ever ever achieve that by "fixing your handicapping"...

[sjk]

In my experience if a horse is 6-1 or less on the morning line and as post time approaches he is above the morning line price and I think the horse is live the price almost always comes down.

[classhandicapper]

I would submit that although you can always improve your handicapping, there will always be information in the public line that you will never know.

In short, it is impossible to "fix your handicapping" EXCEPT for doing this -- THIS fixes it so just do this.

I will state again that I absolutely guarantee that not a single human being in the world can make an unbiased line (over many races) if they do not account for the public odds in some way in their handicapping. An unbiased line is one where the 3-1 (or whatever) horses win at 25% when the public has got them at 6/5 and they still win at 25% when the public has got them at 10-1. You will never ever ever achieve that by "fixing your handicapping"...

I don't disagree with you premise that's it's close to impossible to build a perfect line. That's why I use a margin of safety. I don't have access to every bit of information or the ability to weigh it all perfectly.

But in your example, I gave the horses 33%, the public gave the horse 33%, but then it won 40%. That's the outlier where the actual win percentage isn't somewhere between my line and the public line but still an overlay.

In that case, something is missing from the public line also. I'd like to figure out what it is. It sounds very promising. ;)

[Cratos]

I would submit that although you can always improve your handicapping, there will always be information in the public line that you will never know. In short, it is impossible to "fix your handicapping" EXCEPT for doing this -- THIS fixes it so just do this.

I will state again that I absolutely guarantee that not a single human being in the world can make an unbiased line (over many races) if they do not account for the public odds in some way in their handicapping. An unbiased line is one where the 3-1 (or whatever) horses win at 25% when the public has got them at 6/5 and they still win at 25% when the public has got them at 10-1. You will never ever ever achieve that by "fixing your handicapping"...

In all due respect, this post is near "genius" with its explanation.

[raybo]

The idea is that the ML influences the way the public bets. In this era of whale money, I'm not sure if that influence is still strong enough, but in years past the ML was at least partially a self-fulfilling prophecy. So you know people would see a horse was 2-1 on the ML, but at 10 MTP he's only 8-1 or something. Just the fact that his ML is low will attract money to him as non-sophisticated players start to think overlay.

But what if the ML is say 4-1 and the horse is bet down to 2-1 on the board? Even though he is a big "ML underlay" there is some portion of the public that will now try to beat him not only because his odds are low but because that 4-1 ML is bugging them. So what I'm saying is that same horse, if his ML was 2-1 instead of 4-1 would probably be bet down even more to 7/5 to 6/5. And the people betting him down usually have good reasons, and so it is now likely that that horse is a true overlay at 2-1 because he should be 7/5 and the 4-1 ML line is propping him up by influencing the betting. Follow?

And so, as a believer in using the public odds to alter my own lines, I have no problem adjusting my probability upward on a horse that the public is rating higher than me. So what happens when I bet against the ML instead of the real line? Well, maybe I rate the horse at 2-1. His ML is 4-1. The ML is the proxy for the odds, so that makes him an overlay. But now in real life he's bet down to 2-1 making it a break even or underlay. So I've just screwed myself, right? I'm making a bad bet by using the ML. Well no, because as I just explained, if he got bet down to 2-1 against a 4-1 ML, he had to "overcome" that 4-1 to do it and he really should be 7/5. Which makes him an overlay, and a good bet. (I LOVE low priced overlays.)

The reverse situation where the horse goes off even higher than his morning line, which should mean super overlay, was the more problematic situation as it often meant (if I had the horse at lower odds, meaning I would bet him against that ML) I had overestimated his chances and he was a false overlay, or would win very seldom anyway.

This is something I actually used to do so I didn't have to grind all day, just use the ML in the morning, make my bets, and that was that. AND I used the ML as if it were the public line to model "what actually happens" as described earlier, so I was actually betting against THAT line, which was already adjusted by the ML. (Further advantage of the ML is that there is no guessing what the odds are going to be, they are fixed.) So the whole method was tuned for using the ML. And it worked, and it puts you onto those low-priced overlays that a traditional real-odds approach will keep you away from. But whether it would still work now, I just don't know if the ML has that much influence over the odds anymore...

Ok, that makes sense! But, like you, I don't think the M/L has the same affect on the way the public actually bets. At least not as much as in the past.

Thanks for the explanation! :ThmbUp:

[DLigett]

So, yes NEW overlays can appear, and yes NEW underlays can appear.



I like examples... but, this is my first try at an attachment, hope it shows up...

[raybo]

If we blend our betting line with the crowd's in equal measure, aren't we, in fact, admitting that our opinion is no better than that of the public's?

I don't think anyone in this thread thinks otherwise. But, the thing is to determine value, not improve hit rate. If you can more accurately determine true odds then you can more accurately determine value potential between the horses.

[GameTheory]

I don't disagree with you premise that's it's close to impossible to build a perfect line. That's why I use a margin of safety. I don't have access to every bit of information or the ability to weigh it all perfectly. Just as you should, nothing wrong with that! But don't waterdown my premise -- it is not "close to impossible to build a perfect line" (without reference to the odds). It is absolutely totally impossible. "Perfect" here meaning nothing more than unbiased, a perfect perfect line would have only 1s and 0s and pick 100% winners...

[DLigett]

I butchered the formula when I deleted a couple of lines. Sorry. Serves me right for trying to document a spreadsheet.

=EXP($A$9*LN(A2)+$B$9*LN(B2))

Something like...

New Prob = e ** ( Alpha * ln(My Line) + Beta * ln(Public Line) ) / denom

[thaskalos]

I would submit that although you can always improve your handicapping, there will always be information in the public line that you will never know. In short, it is impossible to "fix your handicapping" EXCEPT for doing this -- THIS fixes it so just do this.

I will state again that I absolutely guarantee that not a single human being in the world can make an unbiased line (over many races) if they do not account for the public odds in some way in their handicapping. An unbiased line is one where the 3-1 (or whatever) horses win at 25% when the public has got them at 6/5 and they still win at 25% when the public has got them at 10-1. You will never ever ever achieve that by "fixing your handicapping"...

Yes...the public line will always contain information that the individual bettor will never know, regardless of how much he "improves" his handicapping. But will this situation improve sufficiently by just blending our betting line with that of the crowd's?

And, YES...our 3/1 shots will not win 25% of the time across all the different odds ranges...but will our "blending" remedy this situation enough...or will the end result still fail to impress?

I respect what you are doing here, GT, and you are making a lot of sense...but "making sense" and "making a profit" do not always go hand-in-hand in this game.

Perhaps, a real-life demonstration...? :)

[raybo]

Yes...the public line will always contain information that the individual bettor will never know, regardless of how much he "improves" his handicapping. But will this situation improve sufficiently by just blending our betting line with that of the crowd's?

And, YES...our 3/1 shots will not win 25% of the time across all the different odds ranges...but will our "blending" remedy this situation enough...or will the end result still fail to impress?

I respect what you are doing here, GT, and you are making a lot of sense...but "making sense" and "making a profit" do not always go hand-in-hand in this game.

Perhaps, a real-life demonstration...? :)

LOL - dang, let him get his testing done first. He may come back after that and say "never mind" (for either reason, not good or really good! :lol:

[GameTheory]

Yes...the public line will always contain information that the individual bettor will never know, regardless of how much he "improves" his handicapping. But will this situation improve sufficiently by just blending our betting line with that of the crowd's?Depends if your line is any good, and different from the public. If it doesn't haven't the inherent potential to make money, then nothing can save it.

And, YES...our 3/1 shots will not win 25% of the time across all the different odds ranges...but will our "blending" remedy this situation enough...or will the end result still fail to impress?It will remedy that situation entirely. If your line sucks, it will do it by just making your line invisible, i.e. it will just end up being a copy of the public line. The public's line will be unbiased in reference to itself, of course.

I respect what you are doing here, GT, and you are making a lot of sense...but "making sense" and "making a profit" do not always go hand-in-hand in this game.See: William Benter and the millions he made.

Perhaps, a real-life demonstration...? :)We've discussed this before, and I offered to any naysayers that actually make their own full-race lines that if they provided me with a whole bunch of them (hundreds or thousands) from past races, then I'd make a model out of them and prove it to them. I am no longer a line-maker myself. The real thing you all should be complaining about is, yeah but don't we have to know the final odds? And the answer is yes, which you can't, so you must use post time odds, or make a probability distribution based on saved real-time tote data of what the final odds are going to be. That's something I have always wanted to do (probability distribution of where the odds are going to end up), but never got around to it. Plus I was being devilishly clever by using the ML and actually having it work...

[thaskalos]

LOL - dang, let him get his testing done first. He may come back after that and say "never mind" (for either reason, not good or really good! :lol:

Let me explain the problem as I see it, Raybo...using an example that we are all familiar with:

We have an obvious contender to whom we've assigned the 2/1 favoritism...but this horse is 5/1 at post time. The classic "dead-on-the-board" horse. Now...we could blend our line with that of the crowd's, and call this horse a 7/2 shot...but what usually happens is that the horse will run like a 10/1 shot...suggesting that the 5/1 odds he went off at were held at 5/1 only because the uninformed masses were unaware of some important information about this horse which was known only to the "few". In this case, "blending" may INDEED improve our line...but not NEARLY enough.

[raybo]

Let me explain the problem as I see it, Raybo...using an example that we are all familiar with:

We have an obvious contender to whom we've assigned the 2/1 favoritism...but this horse is 5/1 at post time. The classic "dead-on-the-board" horse. Now...we could blend our line with that of the crowd's, and call this horse a 7/2 shot...but what usually happens is that the horse will run like a 10/1 shot...suggesting that the 5/1 odds he went off at were held at 5/1 only because the uninformed masses were unaware of some important information about this horse which was known only to the "few". In this case, "blending" may INDEED improve our line...but not NEARLY enough.

I understand what you're saying, but you are thinking linearly. That example would probably be an example of "on the edge" and be treated as such in a non-linear process.

[GameTheory]

Let me explain the problem as I see it, Raybo...using an example that we are all familiar with:

We have an obvious contender to whom we've assigned the 2/1 favoritism...but this horse is 5/1 at post time. The classic "dead-on-the-board" horse. Now...we could blend our line with that of the crowd's, and call this horse a 7/2 shot...but what usually happens is that the horse will run like a 10/1 shot...suggesting that the 5/1 odds he went off at were held at 5/1 only because the uninformed masses were unaware of some important information about this horse which was known only to the "few". In this case, "blending" may INDEED improve our line...but not NEARLY enough.Seriously? Stop with the linear people, just stop. If that is in fact what "usually happens", then the blending I propose would absolutely remedy that -- THAT'S THE POINT. It models what "usually happens".

[raybo]

Seriously? Stop with the linear people, just stop. If that is in fact what "usually happens", then the blending I propose would absolutely remedy that -- THAT'S THE POINT. It models what "usually happens".

LOL - it is hard to stop thinking linearly, because most of us have always done that. As you say, that is the purpose of modeling your new merged line with what actually happens. The model finds the problem and the new line must solve it, if I am understanding this stuff anyway.

[thaskalos]

Seriously? Stop with the linear people, just stop. If that is in fact what "usually happens", then the blending I propose would absolutely remedy that -- THAT'S THE POINT. It models what "usually happens".
I am sorry...I am always very reluctant to participate in threads of this type, because I can seldom grasp the topic at hand -- nor can I offer anything of value to the conversation. I still keep reading, though...hoping that something new will eventually seep through this thick skull of mine. If only it were possible to teach an old dog new tricks... :)

[raybo]

I am sorry...I am always very reluctant to participate in threads of this type, because I can seldom grasp the topic at hand -- nor can I offer anything of value to the conversation. I still keep reading, though...hoping that something new will eventually seep through this thick skull of mine. If only it were possible to teach an old dog new tricks... :)

I think, if you really paid attention, that you may have learned something after all. When you model, you run data and results into a database, then you find out what "normally" happens in all circumstances (at least all circumstances that you are inputting into it). Then, based on what normally happens, you find solutions that closely model those normal results. In this case, the dead money horse "normally" performs poorly, so the solution is that his odds should be higher than what they are, because the model says they should be higher.

That's the difference between linear and non-linear, the adjustments are not straight/linear, they are curved/non-linear, and the largest deviations/adjustments are at both ends of the curve.

GT, if I'm all wet please correct me, this stuff is quite a bit over my head too.

[thaskalos]

I think, if you really paid attention, that you may have learned something after all. When you model, you run data and results into a database, then you find out what "normally" happens in all circumstances (at least all circumstances that you are inputting into it). Then, based on what normally happens, you find solutions that closely model those normal results. In this case, the dead money horse "normally" performs poorly, so the solution is that his odds should be higher than what they are, because the model says they should be higher.

That's the difference between linear and non-linear, the adjustments are not straight/linear, they are curved/non-linear, and the largest deviations/adjustments are at both ends of the curve.

GT, if I'm all wet please correct me, this stuff is quite a bit over my head too.
I grasp this much, Raybo. Where I part ways with this level of thinking is in the assumption that what WAS, will still carry out into the future. Based on what I've experenced, the game is very chaoc...and past betting patterns -- or past "discoveries" -- do not necessarily make future repeat appearances. A repeat occurrence that we've attributed to one factor...could easily be taking place because of something else.

As I said...this "modeling" theory is not making sense to me...and the fault could easily be mine.

[GameTheory]

I grasp this much, Raybo. Where I part ways with this level of thinking is in the assumption that what WAS, will still carry out into the future. Based on what I've experenced, the game is very chaoc...and past betting patterns -- or past "discoveries" -- do not necessarily make future repeat appearances. A repeat occurrence that we've attributed to one factor...could easily be taking place because of something else.
I never understand this line of thinking coming from a handicapper. "History doesn't repeat itself!" Then how in the world do you handicap any race? It is all based on lessons learned from previous races, is it not?

[raybo]

I grasp this much, Raybo. Where I part ways with this level of thinking is in the assumption that what WAS, will still carry out into the future. Based on what I've experenced, the game is very chaoc...and past betting patterns -- or past "discoveries" -- do not necessarily make future repeat appearances. A repeat occurrence that we've attributed to one factor...could easily be taking place because of something else.

As I said...this "modeling" theory is not making sense to me...and the fault could easily be mine.

Sounds like you need to have a long talk with DL, because I'm pretty sure he knows all about this stuff. As a matter of fact, I think the project he is working on is quite similar to what is being discussed here. Huge database of race data and results, huge input/circumstances set, and as a result, a pretty good number of large sample models. In live play, these models would be compared against each race's "circumstances"/factors set, and the one that matches those exact circumstances are used for that race.

The same thing would happen with this odds line adjustment method. Each horse's "circumstances" (including the odds) are matched against the models, and the line's odds are adjusted according to what the model for those circumstances and normal finish positions suggests. This could mean that your "dead on the board" 5/1 horse's odds line gets adjusted above 5/1, not just somewhere between your 7/2 line and the public's 5/1 line, because if these circumstances in the past have resulted in horses finishing, not where a 5/1 horse should finish, but worse than that, then the adjustments could well put his new odds above 5/1. Unless I'm all wet, of course. :lol: (I'm trying to learn something too)

[thaskalos]

I never understand this line of thinking coming from a handicapper. "History doesn't repeat itself!" Then how in the world do you handicap any race? It is all based on lessons learned from previous races, is it not?
I don't handicap like others do, GT. I don't look for history to repeat itself. To me, the future is always new and chaotic...and it's that "chaos" that I have learned to take advantage of. I bet the most on races that other players avoid like the plague...and I treat every race as the once-in-a-lifetime event that it is -- to me at least.

I may not be a "computerized" player...but there is NOTHING conventional about what I do.

[raybo]

I don't handicap like others do, GT. I don't look for history to repeat itself. To me, the future is always new and chaotic...and it's that "chaos" that I have learned to take advantage of. I bet the most on races that other players avoid like the plague...and I treat every race as the once-in-a-lifetime event that it is -- to me at least.

I may not be a "computerized" player...but there is NOTHING conventional about what I do.

LOL - the response to this should be interesting. I know what mine would be, but I have a feeling GT's will be much more exciting.

[thaskalos]

LOL - the response to this should be interesting. I know what mine would be, but I have a feeling GT's will be much more exciting.
What would YOUR response be?

[GameTheory]

I don't handicap like others do, GT. I don't look for history to repeat itself. To me, the future is always new and chaotic...and it's that "chaos" that I have learned to take advantage of. I bet the most on races that other players avoid like the plague...and I treat every race as the once-in-a-lifetime event that it is -- to me at least.

I may not be a "computerized" player...but there is NOTHING conventional about what I do.
If you are not betting randomly (in which case you are definitely losing, and losing a lot), then you are using something from the past to guide you. Like the information that most people avoid the races you bet on -- how did you find that out? Because that's been the case up until now. How do you know that you don't handicap like others? Because that's been true in the past. Do you assume that today just may be the day when everyone else starts to bet just like you? I doubt it, and why? Because it hasn't happened yet. Everything you know, everything you have learned has been shaped what has happened or not happened in the past -- what other way could it be? (Other than random.) I am sorry but there is no way around that unless you have a time machine and are looking at the future.

[thaskalos]

If you are not betting randomly (in which case you are definitely losing, and losing a lot), then you are using something from the past to guide you. Like the information that most people avoid the races you bet on -- how did you find that out? Because that's been the case up until now. How do you know that you don't handicap like others? Because that's been true in the past. Do you assume that today just may be the day when everyone else starts to bet just like you? I doubt it, and why? Because it hasn't happened yet. Everything you know, everything you have learned has been shaped what has happened or not happened in the past -- what other way could it be? (Other than random.) I am sorry but there is no way around that unless you have a time machine and are looking at the future.

What does any of this have to do with history "repeating itself"? What part of racing "history" should I expect to "repeat"? Are the handicapping "factors" to whom I've attributed the results of the last 500 races supposed to demonstrate the same effectiveness over the NEXT 500 races? Are the tote board betting patterns of today supposed to make a repeat appearance at some future point in time? Is the "system" that worked miracles at Arlington this year supposed to be trusted NEXT year?

Looking at the past for subtle hints is one thing; expecting the past to "repeat" itself, is something else altogether. This would have been a much easier game if history really "repeated" itself.

[traveler]

Yes, I think that is right. Years ago I posted here about how betting against the morning line as-if it were the actual odds was not as laughable as it sounds for similar reasons.

You would be 100% correct.

[GameTheory]

What does any of this have to do with history "repeating itself"? What part of racing "history" should I expect to "repeat"? Are the handicapping "factors" to whom I've attributed the results of the last 500 races supposed to demonstrate the same effectiveness over the NEXT 500 races? Are the tote board betting patterns of today supposed to make a repeat appearance at some future point in time? Is the "system" that worked miracles at Arlington this year supposed to be trusted NEXT year?

Looking at the past for subtle hints is one thing; expecting the past to "repeat" itself, is something else altogether. This would have been a much easier game if history really "repeated" itself.You are setting an arbitrary level of abstractness (or specificness, same thing) and claiming "nothing repeats!". As an example, do you not expect that next year that the public favorite will win at a higher rate than the public second favorite over all races just as it has in every other year known? That's history repeating itself, nothing complicated about it. But if we make it very specific, like do we expect the favorite rate to be the same as this year down to 5 decimal points, then no, that probably won't happen.

But anyway, if you make your line with some consistent method (e.g. like from a factor model), then yes you can expect the relationship your lines have had with the public lines in the past to continue in the future, more or less. Of course you update your models as you go along.

[raybo]

What would YOUR response be?

I would have said basically the same thing GT just said, everything you do to handicap and wager today, is based on experiences from your past, and your past is your history. Nobody is saying that history will repeat itself, exactly, because that would make perfection possible. But, it will repeat itself enough that you can stay ahead of the game, as long as you don't stop remembering what has happened along the way, which will gradually adjust your past and continually update your present methods. A constantly evolving database (history) will definitely continually evolve your methods, unless you stop playing that is.

As GA stated, anyone that is not selecting horses and betting totally randomly, is using history to do both.

[thaskalos]

I would have said basically the same thing GT just said, everything you do to handicap and wager today, is based on experiences from your past, and your past is your history. Nobody is saying that history will repeat itself, exactly, because that would make perfection possible. But, it will repeat itself enough that you can stay ahead of the game, as long as you don't stop remembering what has happened along the way, which will gradually adjust your past and continually update your present methods. A constantly evolving database (history) will definitely continually evolve your methods, unless you stop playing that is.

As GA stated, anyone that is not selecting horses and betting totally randomly, is using history to do both.

I politely disagree with the concept of "history repeating itself"...and I mean no disrespect whatsoever by doing so. I respect differing points of view...and I never claim to possess any sort of knowledge which "separates" me from other competent players of this game. Different opinions are the very essence of this game.

To me...there is an artificial component of dishonesty (for lack of a better word) inherent in this game...and that casts a certain shadow of doubt on what tomorrow may bring. I have done well in this game for more than a few years now...but I don't dare presume that this will necessarily continue into the future. Things change. And if I can't be sure of the continuation of my OWN performance in this game...then what level of certainty can I attach to OTHER aspects of this game, as far as the future is concerned?

[classhandicapper]

I politely disagree with the concept of "history repeating itself"...and I mean no disrespect whatsoever by doing so. I respect differing points of view...and I never claim to possess any sort of knowledge which "separates" me from other competent players of this game. Different opinions are the very essence of this game.

To me...there is an artificial component of dishonesty (for lack of a better word) inherent in this game...and that casts a certain shadow of doubt on what tomorrow may bring. I have done well in this game for more than a few years now...but I don't dare presume that this will necessarily continue into the future. Things change. And if I can't be sure of the continuation of my OWN performance in this game...then what level of certainty can I attach to OTHER aspects of this game, as far as the future is concerned?

If you are saying that the game is constantly going through subtle changes that impact both handicapping and how the crowd is betting, I agree with that.

[thaskalos]

If you are saying that the game is constantly going through subtle changes that impact both handicapping and how the crowd is betting, I agree with that.

Yes, I am saying that...and I am also saying one other thing: This game is getting tougher for ALL of us...whether we are "conventional" handicappers or not. There is a stigma attached to the "conventional handicapper"...IMO. The game has supposedly passed him by, and his methods are no longer working, if they EVER did. His edge has been stolen by the more resourceful player...who has embraced the computer as a main handicapping staple. I hear "computerized" handicappers saying that the game has not gotten tougher for them...but I don't believe them. When I say that I am seriously thinking about giving up this game, it isn't because some other handicapper has stolen my edge; my methods are as effective today as they've EVER been. The change in the game is of a more FUNDAMENTAL nature...and it's affecting ALL the players -- whether they want to admit it or not.

I know that the vast majority of the players don't play this game daily, like I do...so they may be failing to notice the dilapidated state that this game has devolved to during the weekdays. A small field means a small chance that the "crowd" will make the sort of mistake that's needed in order to overcome the onerous takeouts that all of us in this game are burdened with. There are less betting opportunities during the weekdays than there have been in years past...and there is no relief in sight -- either in the form of increased field sizes...OR in the form of some much-needed reduction in the takeouts. Maintaining the same takeouts as the fields gradually decrease in size is a recipe for disaster, as far as the horseplayer is concerned...and no handicapping "improvement" of ANY sort is capable of rectifying this situation.

WAR has been waged against the horseplayer by the racing industry itself...and the serious player is fighting for his life -- and losing. For those who see this game as an exciting "entertainment" venue...there might be no real cause for concern. But the profit-motivated player better take notice...because things are bad out there -- and getting worse. And my hunch is that the computerized player is in the same boat with everyone else in this fight.

[TravisVOX]

I politely disagree with the concept of "history repeating itself"...

Not to go too far down this path but I think you're drastically under-valuing the knowledge gained from past events. While I do agree with the idea that racing is constantly changing from the perspective of the horseplayer, to imply that past experiences matter not seems like a hasty generalization.

That being said, you are very right in saying this...

Yes, I am saying that...and I am also saying one other thing: This game is getting tougher for ALL of us...

[GameTheory]

On the one hand, it sort of just seems like a sign of the times. In other words, of course horse racing is dying as it is a sport from another age, and it just isn't going to make it in the modern world. But then you look at some of the other countries where horse racing is absolutely huge. I don't have much faith in the future of North American racing, I'd probably just move to one of those other places if I just had to bet on horses...

[lansdale]

Yes, I am saying that...and I am also saying one other thing: This game is getting tougher for ALL of us...whether we are "conventional" handicappers or not. There is a stigma attached to the "conventional handicapper"...IMO. The game has supposedly passed him by, and his methods are no longer working, if they EVER did. His edge has been stolen by the more resourceful player...who has embraced the computer as a main handicapping staple. I hear "computerized" handicappers saying that the game has not gotten tougher for them...but I don't believe them. When I say that I am seriously thinking about giving up this game, it isn't because some other handicapper has stolen my edge; my methods are as effective today as they've EVER been. The change in the game is of a more FUNDAMENTAL nature...and it's affecting ALL the players -- whether they want to admit it or not.

I know that the vast majority of the players don't play this game daily, like I do...so they may be failing to notice the dilapidated state that this game has devolved to during the weekdays. A small field means a small chance that the "crowd" will make the sort of mistake that's needed in order to overcome the onerous takeouts that all of us in this game are burdened with. There are less betting opportunities during the weekdays than there have been in years past...and there is no relief in sight -- either in the form of increased field sizes...OR in the form of some much-needed reduction in the takeouts. Maintaining the same takeouts as the fields gradually decrease in size is a recipe for disaster, as far as the horseplayer is concerned...and no handicapping "improvement" of ANY sort is capable of rectifying this situation.

WAR has been waged against the horseplayer by the racing industry itself...and the serious player is fighting for his life -- and losing. For those who see this game as an exciting "entertainment" venue...there might be no real cause for concern. But the profit-motivated player better take notice...because things are bad out there -- and getting worse. And my hunch is that the computerized player is in the same boat with everyone else in this fight.

Hi Thask,

You seem to have latched on the the 'stolen edge' phrase from my earlier post. As I said, this was a poor choice of words because it implies some kind of conspiracy against the average horseplayer, which I don't think exists. But the reality, as has often been stated on this site, is that the game is dying, and part of that involves the dwindling level of opportunity that you mention. And, as everyone seems to agree, the parimutuel market has become more efficient - and has possibly reached EMH in the lower-odds range, according to some. But I don't think the traditional handicapper is thus stigmatized - after all 99+% of horseplayers (including those using information from a computer) are conventional handicappers - i.e., their picks aren't based on mathematical models. And players who have successfully developed the latter, which involves the requisite math background, and from what I understand, also significant time and effort, are far and few between.

And although the game is getting tougher, this small group with successful models (which, let's remember, is not everyone using a computer to handicap) does have one advantage - scale. On any given day, they can choose from a vastly larger group of races to bet, so although they're facing the same diminished level of opportunity, it's on another scale. I think sjk mentioned that he formerly made something like 80 bets on an average weekend, now it's down to 30 or so - and that's with zero handicapping.

For those traditional horseplayers whose livelihood, or at least a significant part of it, is based on the game, GT's idea isn't bad. Thoroughbred racing seems to be thriving in other parts of the world - maybe one way to go. Regardless, gambling professionally has never been an easy, unstressful way to make a living, and that's one reality that will never change.

Cheers,

lansdale

[Robert Goren]

Hi Thask,

You seem to have latched on the the 'stolen edge' phrase from my earlier post. As I said, this was a poor choice of words because it implies some kind of conspiracy against the average horseplayer, which I don't think exists. But the reality, as has often been stated on this site, is that the game is dying, and part of that involves the dwindling level of opportunity that you mention. And, as everyone seems to agree, the parimutuel market has become more efficient - and has possibly reached EMH in the lower-odds range, according to some. But I don't think the traditional handicapper is thus stigmatized - after all 99+% of horseplayers (including those using information from a computer) are conventional handicappers - i.e., their picks aren't based on mathematical models. And players who have successfully developed the latter, which involves the requisite math background, and from what I understand, also significant time and effort, are far and few between.

And although the game is getting tougher, this small group with successful models (which, let's remember, is not everyone using a computer to handicap) does have one advantage - scale. On any given day, they can choose from a vastly larger group of races to bet, so although they're facing the same diminished level of opportunity, it's on another scale. I think sjk mentioned that he formerly made something like 80 bets on an average weekend, now it's down to 30 or so - and that's with zero handicapping.

For those traditional horseplayers whose livelihood, or at least a significant part of it, is based on the game, GT's idea isn't bad. Thoroughbred racing seems to be thriving in other parts of the world - maybe one way to go. Regardless, gambling professionally has never been an easy, unstressful way to make a living, and that's one reality that will never change.

Cheers,

lansdaleWhile 99% of the bettors maybe conventional handicappers, the money bet by them is far short of 99% of the pools.

[thaskalos]

Hi Thask,

You seem to have latched on the the 'stolen edge' phrase from my earlier post. As I said, this was a poor choice of words because it implies some kind of conspiracy against the average horseplayer, which I don't think exists. But the reality, as has often been stated on this site, is that the game is dying, and part of that involves the dwindling level of opportunity that you mention. And, as everyone seems to agree, the parimutuel market has become more efficient - and has possibly reached EMH in the lower-odds range, according to some. But I don't think the traditional handicapper is thus stigmatized - after all 99+% of horseplayers (including those using information from a computer) are conventional handicappers - i.e., their picks aren't based on mathematical models. And players who have successfully developed the latter, which involves the requisite math background, and from what I understand, also significant time and effort, are far and few between.

And although the game is getting tougher, this small group with successful models (which, let's remember, is not everyone using a computer to handicap) does have one advantage - scale. On any given day, they can choose from a vastly larger group of races to bet, so although they're facing the same diminished level of opportunity, it's on another scale. I think sjk mentioned that he formerly made something like 80 bets on an average weekend, now it's down to 30 or so - and that's with zero handicapping.

For those traditional horseplayers whose livelihood, or at least a significant part of it, is based on the game, GT's idea isn't bad. Thoroughbred racing seems to be thriving in other parts of the world - maybe one way to go. Regardless, gambling professionally has never been an easy, unstressful way to make a living, and that's one reality that will never change.

Cheers,

lansdale

Hi lansdale,

I hope you don't mind that I like to use that word from your previous post. :) I do it not as a dig to you...but because "stole" neatly describes some arguments of the same type that i've seen in the past. And yes...I have already acknowledged that "hi-tech" handicapping can be a major advantage when employed by an already competent player. I only wish that I could have made a decision years earlier to educate myself in that aspect of the game.

And, no...even though GT's suggestion to move to another racing-friendly country might appeal to some...it just isn't for me. I think I'll be able to find something to occupy myself with after the "death" of my favorite game. :)

[raybo]

Yes, the game is getting tougher to beat. We have noticed, with my program, that although hit rates are on par with previous years, the average prices we are getting this year at almost all tracks, and at quite a few tracks last year, are lower. There are probably multiple reasons for that, as have been stated.

Can profit still be had? Yes. Can a living still be made? Probably by a select few, but I'd hate to have to try to live off of only the profits from race wagering, in today's world. Is the game dying? It certainly looks like it. Is there hope for the future? Not that I can see, as of now. Will the industry "snap to" and make significant changes that will positively affect the players, and thus the game? Are you kidding? Why would they? They never have in the past, except for an "insignificant" number of venues, and even those cannot make up for much, as more and more serious bettors recognize the advantage those few tracks offer and bet good horses down there, more and more as time goes by. Fewer tracks, means smaller crops, which means little or no increase in field sizes, etc., and that all relates to even fewer profitable opportunities. Many are proposing closing all but a few major tracks, in order to give the player a better chance, with supposedly larger fields and larger pools. Well, maybe initially that might happen to some degree, but I don't believe that strategy will work for more than a short period of time. By doing so, more and more players will be forced to leave the game (because fewer will have access to those few tracks, especially as governments continue to take the internet out of the picture), faster than ever before. Fewer players, and fewer wagered dollars means even those few large tracks remaining will not be able to keep their heads above water for long. A snowball effect will take place, hastening the demise of our game.

That's the way I see it, anyway.

It's a sorry state of affairs, for sure.

[traynor]

Yes, I am saying that...and I am also saying one other thing: This game is getting tougher for ALL of us...whether we are "conventional" handicappers or not. There is a stigma attached to the "conventional handicapper"...IMO. The game has supposedly passed him by, and his methods are no longer working, if they EVER did. His edge has been stolen by the more resourceful player...who has embraced the computer as a main handicapping staple. I hear "computerized" handicappers saying that the game has not gotten tougher for them...but I don't believe them. When I say that I am seriously thinking about giving up this game, it isn't because some other handicapper has stolen my edge; my methods are as effective today as they've EVER been. The change in the game is of a more FUNDAMENTAL nature...and it's affecting ALL the players -- whether they want to admit it or not.

I know that the vast majority of the players don't play this game daily, like I do...so they may be failing to notice the dilapidated state that this game has devolved to during the weekdays. A small field means a small chance that the "crowd" will make the sort of mistake that's needed in order to overcome the onerous takeouts that all of us in this game are burdened with. There are less betting opportunities during the weekdays than there have been in years past...and there is no relief in sight -- either in the form of increased field sizes...OR in the form of some much-needed reduction in the takeouts. Maintaining the same takeouts as the fields gradually decrease in size is a recipe for disaster, as far as the horseplayer is concerned...and no handicapping "improvement" of ANY sort is capable of rectifying this situation.

WAR has been waged against the horseplayer by the racing industry itself...and the serious player is fighting for his life -- and losing. For those who see this game as an exciting "entertainment" venue...there might be no real cause for concern. But the profit-motivated player better take notice...because things are bad out there -- and getting worse. And my hunch is that the computerized player is in the same boat with everyone else in this fight.

I think the main reason why computer bettors do better (and have little fear of deterioration in the game) is that they DON'T "play this game daily." The only time they bet is when they have an edge. No edge, no profit, no bet. Pretty simple stuff.

I can think of very few things more boring than "playing this game daily" in some endless procession of sameness, week after week, month after month, year after year.

[cashmachine]

I know that the vast majority of the players don't play this game daily, like I do...so they may be failing to notice the dilapidated state that this game has devolved to during the weekdays. A small field means a small chance that the "crowd" will make the sort of mistake that's needed in order to overcome the onerous takeouts that all of us in this game are burdened with. There are less betting opportunities during the weekdays than there have been in years past...and there is no relief in sight -- either in the form of increased field sizes...OR in the form of some much-needed reduction in the takeouts. Maintaining the same takeouts as the fields gradually decrease in size is a recipe for disaster, as far as the horseplayer is concerned...and no handicapping "improvement" of ANY sort is capable of rectifying this situation.


Why don't you try your hand in Hong-Kong?

[thaskalos]

I can think of very few things more boring than "playing this game daily" in some endless procession of sameness, week after week, month after month, year after year.
Though I try...I can't take you seriously when you talk like this. Could you please tell me what "exciting" activity these computer "geniuses" engage in to avoid the boredom associated with daily betting?

I submit that the person who thinks playing the horses is "boring" hasn't held a real job in his life. This game might be a lot of things...but it's far from the boring endeavor that you make it out to be. So interesting a game is this, in fact...that most people pay substantial money for the privilege of playing it. :)

[classhandicapper]

I agree there are fewer opportunities now than years ago.

I think that's party because the competition is thicker (from both traditional and computer users) and partly because a lot of the dead money has moved on to slots, sports, lotteries etc...

But there are still lots of opportunities. I screwed up an obvious one yesterday.

There's no way Bayern should have been 7-2 in the race at PARX. It was a speed favoring golden rail track and he was the inside main speed (with an excuse for his last).

I screwed it up because I came into the day loving Tapiture and not liking Bayern as the 2nd choice on the assumption of an honest track or bad rail because that's how PARX tends to play.

By post time though, everyone on paceadvantage knew the race had changed dramatically because of the way the track was playing. I'm sure many people adjusted, bet him, and got great value. I didn't because it's never easy for me to reverse my position going into a day by 180 degrees and do the complete opposite of what I intended.

If anything, my example suggests we should have a thread on the mental aspects of the game.

[classhandicapper]

I think the main reason why computer bettors do better (and have little fear of deterioration in the game) is that they DON'T "play this game daily." The only time they bet is when they have an edge. No edge, no profit, no bet. Pretty simple stuff.



Where is the evidence that computer bettors do better?

I've known quite a few winners over 40 years and I still haven't met a single player that beats the game consistently using a computer model.

Even if I grant that some of the people posting here are winners even though I don't know them personally, there's no evidence whatsoever that it's easier for them or that there are more of them etc... If anything, the thing I am taking away from this discussion is that there are multiple ways to achieve profits but everyone has some holes in their game and approach.

[thaskalos]

I agree there are fewer opportunities now than years ago.

I think that's party because the competition is thicker (from both traditional and computer users) and partly because a lot of the dead money has moved on to slots, sports, lotteries etc...

But there are still lots of opportunities. I screwed up an obvious one yesterday.

There's no way Bayern should have been 7-2 in the race at PARX. It was a speed favoring golden rail track and he was the inside main speed (with an excuse for his last).

I screwed it up because I came into the day loving Tapiture and not liking Bayern as the 2nd choice on the assumption of an honest track or bad rail because that's how PARX tends to play.

By post time though, everyone on paceadvantage knew the race had changed dramatically because of the way the track was playing. I'm sure many people adjusted, bet him, and got great value. I didn't because it's never easy for me to reverse my position going into a day by 180 degrees and do the complete opposite of what I intended.

If anything, my example suggests we should have a thread on the mental aspects of the game.

You said a mouthful there, CH...BOOKS could be written on the mental aspect of this game. I don't care HOW good a handicapper you are, or how ADVANCED the tools at your disposal happen to be...it all boils down to those last few seconds when you are ready to put your money on the line. The REAL game takes place between our ears...

[traynor]

Though I try...I can't take you seriously when you talk like this. Could you please tell me what "exciting" activity these computer "geniuses" engage in to avoid the boredom associated with daily betting?

I submit that the person who thinks playing the horses is "boring" hasn't held a real job in his life. This game might be a lot of things...but it's far from the boring endeavor that you make it out to be. So interesting a game is this, in fact...that most people pay substantial money for the privilege of playing it. :)

Some believe the same thing about routine jobs--that they are so challenging, exciting, and rewarding that they are justified (as in what is called "the Microsoft Model") in working 60-70-80 hours a week at a job that most would consider just that--a job--and no more. "Salaried" employees (as opposed to "hourly" employees limited to a finite number of work hours) are often expected to show their dedication and love for their job by putting in lots of (unpaid) extra time and effort. And many are more than happy to do so.

To rationalize such behavior, those engaging in such behavior must convince themselves their work is challenging, exciting, and rewarding. Much like the enthusiasm expressed by some about horse racing. In both cases, it is subjective opinion manipulated internally to avoid the cognitive dissonance of engaging in boring, repetitive endeavors that would otherwise result in serious introspection and questioning of one's real motives.

[raybo]

Though I try...I can't take you seriously when you talk like this. Could you please tell me what "exciting" activity these computer "geniuses" engage in to avoid the boredom associated with daily betting?

I submit that the person who thinks playing the horses is "boring" hasn't held a real job in his life. This game might be a lot of things...but it's far from the boring endeavor that you make it out to be. So interesting a game is this, in fact...that most people pay substantial money for the privilege of playing it. :)

To the sophisticated computer player, every race is just like any other race, the computer spits our some selections and you either bet or you don't. The extreme consistency of that kind of play does become boring. The losing streaks and winning streaks are just part of the method and a means to an end, no need to get excited or depressed, just keep the thing maintained and updated, and keep looking at potential bets. And, a few of those guys don't even have to decide to bet or not, or place the wagers, the program does it all. Talk about boring!

[Robert Goren]

Some believe the same thing about routine jobs--that they are so challenging, exciting, and rewarding that they are justified (as in what is called "the Microsoft Model") in working 60-70-80 hours a week at a job that most would consider just that--a job--and no more. "Salaried" employees (as opposed to "hourly" employees limited to a finite number of work hours) are often expected to show their dedication and love for their job by putting in lots of (unpaid) extra time and effort. And many are more than happy to do so.

To rationalize such behavior, those engaging in such behavior must convince themselves their work is challenging, exciting, and rewarding. Much like the enthusiasm expressed by some about horse racing. In both cases, it is subjective opinion manipulated internally to avoid the cognitive dissonance of engaging in boring, repetitive endeavors that would otherwise result in serious introspection and questioning of one's real motives.The hope for advancement is the driving force for most salaried employment(and more money and more prestige). Horse racing is closer to an addiction. The bet is the fix although not many handicappers will admit it. Most non horse players see the game for what it is.

[Robert Goren]

To the sophisticated computer player, every race is just like any other race, the computer spits our some selections and you either bet or you don't. The extreme consistency of that kind of play does become boring. The losing streaks and winning streaks are just part of the method and a means to an end, no need to get excited or depressed, just keep the thing maintained and updated, and keep looking at potential bets. And, a few of those guys don't even have to decide to bet or not, or place the wagers, the program does it all. Talk about boring!The thrill is in developing the model.

[raybo]

The thrill is in developing the model.

Exactly! I get more satisfaction, and have more fun, working in Excel than I ever got from handicapping and wagering. I really hate gambling with a passion!

For me, racing is just a part time job, at least it was before Texas made it illegal to wager online. Now, it's just something to do, and I find myself getting more into old school analysis again, since I have more time and I'm not wagering real money anymore. Before, it was all automated, get the tracks tested in "All Record" mode, select the methods and minimum odds for each track, get the picks from the program, watch the odds at a minute to post, and bet or pass. That was not fun, just going through the motions. Boring. A lot like working on an assembly line, installing the same part hour after hour, day after day.

I don't know, maybe one of these days I'll take the laptop and find a quiet place outside on the patio and wager a couple of cards at the simulcast building. Won't be any time soon though, too much trouble, and expense, driving the 89 miles to Lone Star Park, paying for parking and admission, dealing with the people and the noise while standing in line at the windows, etc..

[traynor]

The thrill is in developing the model.

Indeed. And stayng one-and-a-half jumps in front of the other modelers, milking the cash cows to a trickle, and moving on before they realize what is happening. Nothing is ever good enough, nothing ever works forever, and a moment of indulgence or inattention can cause serious losses. What a great life!

[Tom]

It's getting pretty deep in here, Gus. :rolleyes:
Originally Posted by Robert Goren
The thrill is in developing the model.

I think it was suggested early on that some people like to talk about handicapping, while others like to do it. this kind of agrees with that.

[raybo]

It's getting pretty deep in here, Gus. :rolleyes:


I think it was suggested early on that some people like to talk about handicapping, while others like to do it. this kind of agrees with that.

Some are only interested, or get excited, about making money, others get excited about racing in general, the handicapping and gambling, others fall into both categories. I happen to be one that hates the gambling side and gets very mentally tired while old school capping, and very bored with automated capping. If it weren't for the profit potential, I would probably have never have gotten as involved as I did, or continued for all these years. I usually don't even watch the live race videos unless it's a big race.

[Robert Goren]

Making money at gambling is always boring. If it isn't, then you have a problem. The most boring thing in the world is counting cards at blackjack. Playing Holdem is a close second.

[thaskalos]

Making money at gambling is always boring. If it isn't, then you have a problem. The most boring thing in the world is counting cards at blackjack. Playing Holdem is a close second.
Where does working at a real job rate?

[Robert Goren]

Where does working at a real job rate?Pretty boring too.

[thaskalos]

Some believe the same thing about routine jobs--that they are so challenging, exciting, and rewarding that they are justified (as in what is called "the Microsoft Model") in working 60-70-80 hours a week at a job that most would consider just that--a job--and no more. "Salaried" employees (as opposed to "hourly" employees limited to a finite number of work hours) are often expected to show their dedication and love for their job by putting in lots of (unpaid) extra time and effort. And many are more than happy to do so.

To rationalize such behavior, those engaging in such behavior must convince themselves their work is challenging, exciting, and rewarding. Much like the enthusiasm expressed by some about horse racing. In both cases, it is subjective opinion manipulated internally to avoid the cognitive dissonance of engaging in boring, repetitive endeavors that would otherwise result in serious introspection and questioning of one's real motives.

You have probably spent some time in politics...because you've mastered the art of sounding impressive without really saying anything. :)

This "boredom", which you associate with gambling, is not only a GAMBLING problem; it's a LIFE problem. LIFE ITSELF is highly repetitive and boring. And the gamblers are not the only ones who engage in the behavior of convincing themselves that their work is "challenging, exciting and rewarding". Some "intellectuals" do the same thing with their very LIVES.

Yes...gambling, when it's done right, is BORING. But it isn't more boring than WORKING for a living. Regular employment is EXCRUCIATINGLY boring...and yet...people have adapted quite nicely to it. It's funny to hear people with REAL jobs complaining about the boring nature of "gambling"...

[GameTheory]

The "execution phase" of actually betting for the computer player is a mind-numbing grind. There is no intellectual challenge to it, but it requires attention, and adherence to strict rules. This is why I did things like bet against the ML (just make all my bets in the morning or after scratches were announced and be done with it) or total automation of the betting process (i.e. "live" betting, but the computer does it for me).

Putting together a database and maintaining it is also no great excitement, but creating the algorithms, etc that actually do the handicapping or bet calculations is challenging and interesting anyway.

A "real" handicapper, especially one that goes to the track and looks at live animals has a much more interesting job, well...at least until he gets bored with it.

[Tom]

Making money at gambling is always boring. If it isn't, then you have a problem. The most boring thing in the world is counting cards at blackjack. Playing Holdem is a close second.

Says who?
I know lots of people who will totally disagree with you.

[raybo]

The "execution phase" of actually betting for the computer player is a mind-numbing grind. There is no intellectual challenge to it, but it requires attention, and adherence to strict rules. This is why I did things like bet against the ML (just make all my bets in the morning or after scratches were announced and be done with it) or total automation of the betting process (i.e. "live" betting, but the computer does it for me).

Putting together a database and maintaining it is also no great excitement, but creating the algorithms, etc that actually do the handicapping or bet calculations is challenging and interesting anyway.

A "real" handicapper, especially one that goes to the track and looks at live animals has a much more interesting job, well...at least until he gets bored with it.

Exactly! Once you step over the line to full automation, you are much more likely to be able to take all emotion out of the whole process, which is a good thing for your bottom line, but very boring at the same time.

[raybo]

Says who?
I know lots of people who will totally disagree with you.

LOL - yeah i get all jumpidy and excited, I click a button to get the selections, then sit for several minutes doing nothing, then at less than 1 MTP I compare the tote odds with my minimum requirement, if my selections' odds are equal or above my minimum I bet, if not I pass. Very stimulating and great fun!! :sleeping:

[Tom]

You choose to remove all emotion for the "game."
Others do not.

[raybo]

You choose to remove all emotion for the "game."
Others do not.

As long as your emotions don't affect your decision making that's great. Not many can do that.

[GameTheory]

Exactly! Once you step over the line to full automation, you are much more likely to be able to take all emotion out of the whole process, which is a good thing for your bottom line, but very boring at the same time.
Once it is fully automated, it is interesting because instead of watching it I'm out riding my bike or doing photography or at least doing a computer job for somebody. I just check the results at night, and prepare for the next day...

[classhandicapper]

As long as your emotions don't affect your decision making that's great. Not many can do that.

I can't to the degree that would optimize my results.

If I bet a horse off some trip and the horse loses, I am more likely to at least use him in some exotics in the future just to avoid the pain of watching him win the next time at an even bigger price.

If I have a strong opinion on a horse at the start of the day, but the track conditions change dramatically, it's very easy to me to adjust enough to pass the race, but not so easy for me to do a 180 turn and throw that horse out and bet someone else instead.

If I am ahead for the day, I am more likely to pull back in a very marginal situation to lock in the good day.

If I am behind for the day, I am more likely to bet in a very marginal situation to try to the turn the day around.

On the flip side of all this is that there's nothing like the thrill of winning, and more importantly for me, being right. There's a huge emotional high.

I bet a stupid $2 on a 45-1 mule at Stockton yesterday that ran a solid second. The $90 meant nothing, but you should have heard me screaming during race when I saw I was live and had come up with something especially good.

The money is almost meaningless to me. I think it's all "pain avoidance", the pleasure of being right, and ego.

[thaskalos]

I can't to the degree that would optimize my results.

If I bet a horse off some trip and the horse loses, I am more likely to at least use him in some exotics in the future just to avoid the pain of watching him win the next time at an even bigger price.

If I have a strong opinion on a horse at the start of the day, but the track conditions change dramatically, it's very easy to me to adjust enough to pass the race, but not so easy for me to do a 180 turn and throw that horse out and bet someone else instead.

If I am ahead for the day, I am more likely to pull back in a very marginal situation to lock in the good day.

If I am behind for the day, I am more likely to bet in a very marginal situation to try to the turn the day around.

On the flip side of all this is that there's nothing like the thrill of winning, and more importantly for me, being right. There's a huge emotional high.

I bet a stupid $2 on a 45-1 mule at Stockton yesterday that ran a solid second. The $90 meant nothing, but you should have heard me screaming during race when I saw I was live and had come up with something especially good.

The money is almost meaningless to me. I think it's all "pain avoidance", the pleasure of being right, and ego.
It must be nice to be in this kind of financial position... :)

[raybo]

Once it is fully automated, it is interesting because instead of watching it I'm out riding my bike or doing photography or at least doing a computer job for somebody. I just check the results at night, and prepare for the next day...

I hear ya. When I was able to bet online, I found myself doing other things, outside of racing, when I didn't have a possible bet coming up (my methods do not include automated betting, so I still had to place the wagers manually).

Now that I can't bet online anymore, I only run the bigger races, $250k and up. I've recently bought a used dirt bike, motocross (after 40 years of not riding one), built a trailer to haul it with, and plan to ride as often as possible. Horse racing is getting less and less important to me. No big loss really, although the extra money did come in handy.

[classhandicapper]

It must be nice to be in this kind of financial position... :)

I never played horses full time. I always had a full time job until about 7-8 years ago. Since then I've worked either part time or full time almost steadily with the one added rule that I have to be enjoying what I am doing or I'll quit. ;)

I was never in the position of "having" to win.

[Jeff P]

I submit to you the idea that if betting on horses is 'boring' you might not be betting enough.


-jp

.

[traynor]

I submit to you the idea that if betting on horses is 'boring' you might not be betting enough.


-jp

.

I think anyone who seeks (or finds) "excitement" betting on horse races needs to get a real life.

[thaskalos]

I think anyone who seeks (or finds) "excitement" betting on horse races needs to get a real life.

Jeff is right...some of you guys just ain't betting enough. :)

[AndyC]

I submit to you the idea that if betting on horses is 'boring' you might not be betting enough.


-jp

.

The act itself of betting is boring. The intellectual process of analyzing and handicapping in preparation for a bet is what draws and keeps people playing along with the satisfaction of solving the puzzle.

[thaskalos]

The act itself of betting is boring. The intellectual process of analyzing and handicapping in preparation for a bet is what draws and keeps people playing along with the satisfaction of solving the puzzle.

Andy, could you please do me a favor? Could you ask Beyer if he finds betting on horses to be "boring"...and then come back and tell us what he said? :)

[Tom]

I'll take playing the races every day over working on some computer programs.
To each his own.

We have guys who find racing boring telling us how play the game.

Yeah, right.

[thaskalos]

I'll take playing the races every day over working on some computer programs.
To each his own.

We have guys who find racing boring telling us how play the game.

Yeah, right.
:ThmbUp:

You got these guys saying that betting on the horses is boring...and I get pissed of because I can't find more horses to bet. :)

[whodoyoulike]

Making money at gambling is always boring. If it isn't, then you have a problem. The most boring thing in the world is counting cards at blackjack. Playing Holdem is a close second.

Maybe you're just not doing it correctly? And, I think I do have a problem. But, I'm having fun otherwise I wouldn't do it.

[raybo]

I submit to you the idea that if betting on horses is 'boring' you might not be betting enough.


-jp

.

No doubt, but some of us don't expect to get rich, just cruise along at a more comfortable level.

[raybo]

The act itself of betting is boring. The intellectual process of analyzing and handicapping in preparation for a bet is what draws and keeps people playing along with the satisfaction of solving the puzzle.

Maybe in the big races, but no way I want to sit here and handicap races and solve puzzles all day, every day. let me click some buttons and watch the tote, that's all I need.

[raybo]

I'll take playing the races every day over working on some computer programs.
To each his own.

We have guys who find racing boring telling us how play the game.

Yeah, right.

I would never dare try to tell you how to play the races. You wouldn't listen anyway, I'd just be wasting my time.

[raybo]

:ThmbUp:

You got these guys saying that betting on the horses is boring...and I get pissed of because I can't find more horses to bet. :)

There's lots and lots of races being run, if you can't find races to bet, that's your own fault.

[Overlay]

To me, what would be worse than boring would be time spent in playing or attempting to beat any wagering game with a negative expectation on each and every bet that could never be overcome. At least time invested in handicapping or other "Type W" games still has the theoretical possibility (however small or difficult to realize) of achieving consistent profit through skill as the light at the end of the tunnel.

[Cratos]

The "execution phase" of actually betting for the computer player is a mind-numbing grind. There is no intellectual challenge to it, but it requires attention, and adherence to strict rules. This is why I did things like bet against the ML (just make all my bets in the morning or after scratches were announced and be done with it) or total automation of the betting process (i.e. "live" betting, but the computer does it for me).

Putting together a database and maintaining it is also no great excitement, but creating the algorithms, etc that actually do the handicapping or bet calculations is challenging and interesting anyway.

A "real" handicapper, especially one that goes to the track and looks at live animals has a much more interesting job, well...at least until he gets bored with it.

This is so true; I find enjoyment and excitement in working on and finding solutions to some of the many handicapping methods that is no more than fiction.

This has led me to better understand the quote: “When the man with money meets the man with education, the man with money gets the education and the man with education gets the money.”

Horseracing handicapping is far behind most major sports in the use of modern technology even though its foundation is built on the science of physics. A horserace simply stated is no more than a set of objects moving through space over a surface and the laws of physics directly apply because of distance, mass, velocity, and acceleration.

Yes, you need a fundamental understanding of horserace handicapping, but add that understanding to the handicapper’s competent knowledge of math and statistics and it will move him/her ahead of many horseplayers.

And couple their math and science acumen with modern computer technology in terms of software development and model building they are on the road to a “non-boring” profitable lifestyle.

[raybo]

And couple their math and science acumen with modern computer technology in terms of software development and model building they are on the road to a “non-boring” profitable lifestyle.

I'd go for that! Just turn on the computer, open the software and never look at it again the rest of the day. :jump:

[Tom]

Sounds like life in the fast lane. :eek:
To each his own.
But that was my original point.

[raybo]

Sounds like life in the fast lane. :eek:
To each his own.
But that was my original point.

If you're referring to my post, wrong, I live life in the slow lane, I just prefer complete automation when it comes to horse racing. I spent many, many years analyzing the game before betting, old school (after a few years it became a grind and boring), and when I began playing for real, that was old school also, except for using Excel for data file import and PPs display (still old school capping). Finally I completely automated the capping portion, which is great as a time saver and mistake canceller, but is also boring. The wagering portion was always boring, as I'm not a gambler at heart like most other players, profit potential was/is my only motivation.

Your original point, "We have guys who find racing boring telling us how play the game.", is false, nobody here is trying to "tell" you anything. We know better.

[traynor]

Jeff is right...some of you guys just ain't betting enough. :)

That sounds a bit like "If you don't find crossword puzzles exciting, it is because your crossword puzzle isn't big enough." I don't find money--regardless of amount--"exciting." Useful, yes. Not much more. I don't see any difference between wagering amount x and wagering amounts y or z. While the betting is going on, it is little more than interim scorekeeping.

Pretending something boring is exciting is a subjective interpretation--not a statement of external reality. It may be the result of spending too much time around poor folk obsessed with money, and not enough time around folk who are less impressed with it because they do not consider the acquisition of money as being all that difficult.

[thaskalos]

That sounds a bit like "If you don't find crossword puzzles exciting, it is because your crossword puzzle isn't big enough." I don't find money--regardless of amount--"exciting." Useful, yes. Not much more. I don't see any difference between wagering amount x and wagering amounts y or z. While the betting is going on, it is little more than interim scorekeeping.

Pretending something boring is exciting is a subjective interpretation--not a statement of external reality. It may be the result of spending too much time around poor folk obsessed with money, and not enough time around folk who are less impressed with it because they do not consider the acquisition of money as being all that difficult.

No...I've also spent considerable time with those fortunate folks. And I've regretted the fact that their condition didn't prove more contagious...

[Maximillion]

On the one hand, it sort of just seems like a sign of the times. In other words, of course horse racing is dying as it is a sport from another age, and it just isn't going to make it in the modern world. But then you look at some of the other countries where horse racing is absolutely huge. I don't have much faith in the future of North American racing, I'd probably just move to one of those other places if I just had to bet on horses...

Not to go off topic,but i wonder if any serious attempt has ever been made to promote our NA product to other countries,since they seem to still be in love with the game itself.