Confidence Level

[cmoore]

I'm going to start labeling races from 1 to 5..Five being the most confident and one being the least confident. Now I feel this will work best on races you handicap the morning or day before. I did this last Sunday and my results were this.

6 races got a 2 to 2.5 rating...................0 for 6....Didn't bet

10 races got a 3 to 3.5 rating................3 for 13...(one race I bet 3 horses and another one I bet an exacta wheel..example 1/23)

Winners added up to $35.40..A profit of $9.40.

I only bet on confidence levels of 3 or higher..So it was a winning day..I feel that this skill will improve over time and will help to eliminate races that are to competitive. I feel high odds sometimes sway me into a bet when my confidence on that particular race isn't all that high.

[CincyHorseplayer]

I used to grade races alphabetically before the races and have odds ranges on a numerical scale.For instance if I have a C-5,it's a race I have a so-so opinion about and the odds are low.No bet.

[dav4463]

I tried this and my biggest hits were on races where I had a low confidence rating for some odd reason. :confused:

[CincyHorseplayer]

I tried this and my biggest hits were on races where I had a low confidence rating for some odd reason. :confused:

That's pretty common I think.When races are wide open and most of the field can contend the payoffs usually are good.

Back to my caveman alphabetical numerical rankings,when I have a C-1 race,great payoffs,vague idea of the race I know what separates me winning and losing is distinguishing value from probability.Some of these races are throwaways,the rest really are value at it's best.Making that distinction,those fine line decisions are what separates us from mediocrity and genuine success IMO.

[fmolf]

That's pretty common I think.When races are wide open and most of the field can contend the payoffs usually are good.

Back to my caveman alphabetical numerical rankings,when I have a C-1 race,great payoffs,vague idea of the race I know what separates me winning and losing is distinguishing value from probability.Some of these races are throwaways,the rest really are value at it's best.Making that distinction,those fine line decisions are what separates us from mediocrity and genuine success IMO.
In contentious races i look to body language to eliminate horss and land me on my bet.This has led me to some nice price horses.very helpful when choosing between three 8/1 or higher horses in wide open race with vulnerable favorite.If no horse stands out or two of the three have no serious negative body signs ...no bet

[DanG]

I tried this and my biggest hits were on races where I had a low confidence rating for some odd reason. :confused:
I could not agree more. :ThmbUp:

If there was a way to rank gamblers from the ‘serious as a heart attack / to the person throwing darts; for each level of reduced ‘chaos’, you invite another level of player to see the race 20/20.

Like most general statements; tons of examples in both directions, but many players have learned to embrace ‘noise’ as their best chance to gain an edge.

[Tom]

I could not agree more. :ThmbUp:

but many players have learned to embrace ‘noise’ as their best chance to gain an edge.


That an VI. ;)

[DanG]

That an VI. ;)
I don’t know what you’re talking about Tom. ;) ;)

[markgoldie]

I could not agree more. :ThmbUp:

If there was a way to rank gamblers from the ‘serious as a heart attack / to the person throwing darts; for each level of reduced ‘chaos’, you invite another level of player to see the race 20/20.

Like most general statements; tons of examples in both directions, but many players have learned to embrace ‘noise’ as their best chance to gain an edge.
Agreed. The point being, that unless you are using some esoteric, ultra-sophisticated method of handicapping, then your level of confidence will rise in agreement with the general wagering public. This, in turn, will lower your odds on whatever opinion you hold. Additionally, if you ARE using a methodology which can become clear to you but not to the public, and it is successful, you probably are making enough money that you don't need to be worried about your degree of confidence.

But when I saw this thread, it brought something to mind that hasn't been discussed much, if at all, since I've been on the forum for about the last 3 months. That is, we have a bunch of guys talking about "programmed" wagers, which is to say selections either created by a paper system or methodology OR by a computer program, in which a consistent amount of wager is contemplated. Part of the reason for this is the fact that ROI calculations are done with a standard amount of bet (often $1 or $2). But even if ROI is calculated by percentage, it is generally based on a consistent unit wager.

But does this make sense?? It would seemingly take little thought to reveal that it doesn't. Each event should produce a different degree of opportunity even if a program just spits out a horse's name with no degree of advantage of any kind. That's because the target horse will have a specific post-time ODDS. And those odds will either be higher or lower than the average odds of a selected animal. On the other hand, if the program (or selection process) yields a degree of advantage, here we have two pieces of information (degree of advantage AND odds) with which we might successfully tweak the wagered amount.

This type of "degree of confidence" is a bit different than degree of confidence based on the handicapping layout of a race. This degree of confidence could also be called "degree of opportunity for profit" based on odds.

At any rate, personally I don't do any type of programmed wagering based on a set of handicapping criteria (either computer or paper-based). But every day here I read the posts of guys who do. So let me throw this out as a variation of the original post written by Cmoore: How many of you gentlemen who are using or working on programmed selection processes are incorporating variable wagering amounts, based on degree of advantage and/or odds?

As I say, I haven't heard this discussed much.

[sjk]

I scale bet size based on egde and probability within certain upper and lower bounds. Also raise or lower a bit based on recent success at the track.

[JBmadera]

when it comes to maidens my ears really perk up when Chris really likes a race - I've seen him hit numerous 50-1 shots...how the heck he does it I have no clue....but then again I don't know how electricity works but that doesn't stop me from turning on the lights each morning.....:ThmbUp:

jb

[lansdale]

Agreed. The point being, that unless you are using some esoteric, ultra-sophisticated method of handicapping, then your level of confidence will rise in agreement with the general wagering public. This, in turn, will lower your odds on whatever opinion you hold. Additionally, if you ARE using a methodology which can become clear to you but not to the public, and it is successful, you probably are making enough money that you don't need to be worried about your degree of confidence.

But when I saw this thread, it brought something to mind that hasn't been discussed much, if at all, since I've been on the forum for about the last 3 months. That is, we have a bunch of guys talking about "programmed" wagers, which is to say selections either created by a paper system or methodology OR by a computer program, in which a consistent amount of wager is contemplated. Part of the reason for this is the fact that ROI calculations are done with a standard amount of bet (often $1 or $2). But even if ROI is calculated by percentage, it is generally based on a consistent unit wager.

But does this make sense?? It would seemingly take little thought to reveal that it doesn't. Each event should produce a different degree of opportunity even if a program just spits out a horse's name with no degree of advantage of any kind. That's because the target horse will have a specific post-time ODDS. And those odds will either be higher or lower than the average odds of a selected animal. On the other hand, if the program (or selection process) yields a degree of advantage, here we have two pieces of information (degree of advantage AND odds) with which we might successfully tweak the wagered amount.

This type of "degree of confidence" is a bit different than degree of confidence based on the handicapping layout of a race. This degree of confidence could also be called "degree of opportunity for profit" based on odds.

At any rate, personally I don't do any type of programmed wagering based on a set of handicapping criteria (either computer or paper-based). But every day here I read the posts of guys who do. So let me throw this out as a variation of the original post written by Cmoore: How many of you gentlemen who are using or working on programmed selection processes are incorporating variable wagering amounts, based on degree of advantage and/or odds?

As I say, I haven't heard this discussed much.

Mark,

You don't call it by name, but if you're talking about proportional betting, I assume you're talking about the Kelly Criterion, a method of betting proportionally according to one's advantage. Benter is probably the best known of those who have used it, but it was and is, AFAIK, used by all the Hong Kong betting teams. If you're interested in the 'biography' of Kelly betting, you might like 'Fortune's Formula' by William Poundstone.

Cheers.

lansdale

[markgoldie]

Mark,

You don't call it by name, but if you're talking about proportional betting, I assume you're talking about the Kelly Criterion, a method of betting proportionally according to one's advantage. Benter is probably the best known of those who have used it, but it was and is, AFAIK, used by all the Hong Kong betting teams. If you're interested in the 'biography' of Kelly betting, you might like 'Fortune's Formula' by William Poundstone.

Cheers.

lansdale
Wasn't aware of Kelly Criterion, but it sure makes sense that a guy a smart as Bentner would use this type of advantage exploitation. It just seemed to me that we hear a lot about the results of given systems here on the forum but I just can't remember anyone pointing out that the ROI results were proportionally bet. Instead, we get a lot about strike rates and average payoffs but these criteria are essentially meaningless when you are proportionally betting.

[sjk]

Wasn't aware of Kelly Criterion, but it sure makes sense that a guy a smart as Bentner would use this type of advantage exploitation. It just seemed to me that we hear a lot about the results of given systems here on the forum but I just can't remember anyone pointing out that the ROI results were proportionally bet. Instead, we get a lot about strike rates and average payoffs but these criteria are essentially meaningless when you are proportionally betting.

I don't agree that strike rates and average payoffs are meaningless. You increase your ROI somewhat by betting Kelly or some variant thereof but if your process is not profitable to begin with no betting scheme is going to make it so.

It was recently pointed out on another thread that when your perceived edge doubles your real edge should go up but it has probably not doubled because the longer priced overlays do not get there as often as the shorter priced overlays. So betting more on them should add to your profits but not by as much as you might hope.

Checked my records and varying the bet size has only increased my ROI by 1.5% over 12 years (over 50,000 races).

[markgoldie]

I don't agree that strike rates and average payoffs are meaningless. You increase your ROI somewhat by betting Kelly or some variant thereof but if your process is not profitable to begin with no betting scheme is going to make it so.

It was recently pointed out on another thread that when your perceived edge doubles your real edge should go up but it has probably not doubled because the longer priced overlays do not get there as often as the shorter priced overlays. So betting more on them should add to your profits but not by as much as you might hope.

Checked my records and varying the bet size has only increased my ROI by 1.5% over 12 years (over 50,000 races).
What I meant there was that the ROI cannot be derived or predicted by the strike rate and average payoff. Naturally these are not meaningless statistics, just non-predictive of a particular ROI in a proportional-betting plan of attack.

It should also be pointed out that a proportional betting plan coupled with a small positive ROI could easily indicate that you are losing money on your less-advantageous plays. Therefore, you need to carefully monitor your secondary plays to make sure that this is not the case, because if it is, you need to be more selective and jettison these plays altogether. It's okay for the secondaries to drag the overall ROI down as long as they are producing black-type numbers in their own right.

[Overlay]

I think that the distribution of odds in a personal betting line can be useful in gauging the confidence level associated with a race (assuming that "confidence" is referring to isolating the likely winner), especially if the odds have an objective basis to them. Personal records can also help in establishing a confidence level for cases where your assessment of a particular horse's chances differs significantly from the public's.

[bisket]

mark i don't use computer programs when i'm handicapping a race. one thing i never do is look at odds before picking which horse or horses i'm going to play. i always adjust my wager amounts according to odds on raceday. the higher the odds the more $ bet. i usually wait until 2 to 3 minutes before post to raise my bet amounts. now sometimes in a hard to handicap race i'll throw an extra horse or so in a tri if i know the tri will really pay. or i'll increase my wager on the tri. i usually have a win bet on every race that i play, and this is the bet that i'll increase just before post. its only according to odds, and i'm very regimented on this. two big things to always do if anyone decides to try this method of betting. NEVER LOOK AT ODDS BEFORE PICKING HORSES. greed is a powerful thing to try to overcome while making decisions like this. and always raise your bets IN THE SAME INCREMENTS ALL THE TIME. i think one of the biggest mistakes in handicapping is wagering different amounts on the exact same plays. being consistent with your bets always puts the percentages in your favor if your good at what you do!!!!!

[markgoldie]

I don't look at odds either when I handicap a race because they can taint your handicapping process. However, after carefully handicapping the race to the best of my ability, I consult the odds as a final check of my work, so to speak. Sometimes I will see that the public is betting a particular horse on which I have a low number and I will ask myself, why? Once in a while, I will see something that I hadn't before and it might change my mind about the horse's prospects. In that sense, I use the public as a handicapping partner in a manner of speaking, getting a possible different perspective on the race.

But (and this is a big "but"), if I can't find anything attractive in the pps of the horse they are betting, OR if I can see why they're betting but I disagree, I will still eliminate such an animal. I don't mind going against the public at all since this is the only way you will ever find some sort of positive wagering value. I should say that I also use the Brisnet Prime Power number as another "handicapping partner." Unlike the public, this number is relentlessly logical because it is produced by an algorithm and because it is, you get to know what factors it "likes" and "dislikes" in a horse's pps. That's why I am always fascinated when it produces a number that appears out of character. In this case, I turn the pps upside down, trying to get an insight into what the Prime Power Number is "thinking". And, believe it or not, sometimes I just come away scratching my head. I have no idea how or why the number should be so high (or low). Anecdotally (because I don't keep stats on these things), it seems when the number is mystifying, it is more likely wrong than right based on the performance of the horse in the target race.

I bet differing amounts depending on how much I like the situation. Even in races which fit my criteria for play I will often pass altogether. This, I guess, you might call the ultimate scaling back of a wager. On the other hand, if I really like a scenario, I have and will go way above my normal unit play. It took me many years to learn to "pull the trigger" in races where I am going against the wagering tide. This is an easy thing to advocate on a theoretical basis, but hard to do with actual cash. But on any list of "keys to success" it has to be near the top.

[bisket]

i pull the trigger maybe once a month or so, but boy its really fun when i do. especially when i get a tax withholding satament when i collect!

[lansdale]

I don't agree that strike rates and average payoffs are meaningless. You increase your ROI somewhat by betting Kelly or some variant thereof but if your process is not profitable to begin with no betting scheme is going to make it so.

It was recently pointed out on another thread that when your perceived edge doubles your real edge should go up but it has probably not doubled because the longer priced overlays do not get there as often as the shorter priced overlays. So betting more on them should add to your profits but not by as much as you might hope.

Checked my records and varying the bet size has only increased my ROI by 1.5% over 12 years (over 50,000 races).

SJK,

I know you're a very sharp handicapper, so I hope you can take this in the spirit in which it was intended, but Kelly can't increase your ROI, i.e. your edge, it does exponentially increase your EV - expected value. I remember you stating that you had a long-term 10% or 12% ROI, which is excellent, but if you had been betting Kelly correctly, you should be a millionaire by now. My guess is either your edge is not what you think, or more likely, your Kelly betting isn't sized correctly. Kelly has been proven time and again to be the most effective technique for increasing any given bank, assuming an edge. Ed Thorp extended its use from blackjack, where it's used by all pros, to the hedge fund world, where it has helped him make billions.

Cheers,

lansdale

[sjk]

Lansdale,

My computer keeps up with my results so I could tell you which races I have bet over the past 10 years (outside of the occasional visit to the track), how much I bet and how much I got paid if anything. I have no concern that I am mistaken as to the ROI.

I don't bet Kelly although as I pointed out above I vary my bet in relation to edge, odds and track in play. The largest bets are about 8 times the size of the smallest bets.

I don't use the bankroll concept and have never really believed in it. I have a certain size bet that I am confortable making and do so automatically and dont mind if I lose 50 in a row. I have increased the size of the basic bet somewhat over the years but probably by no more than a factor of 2 or 3.

If someone used Kelly and increased the bet size in relation to a growing bankroll the size of his bets would get very large in relation to what he is comfortable with and accustomed to playing and in some cases in relation to pool size. I imagine the larger players must come to grips with spreading among a number of pools so as not to dilute their own prices too much.

I have never made betting a full time thing or really gone after maximizing the profits to be made from it as I like my regular work and have good income from mainstream sources.

Thanks for you interest.

SJK

[bisket]

i look at odds for a second check also. although i preffer to discuss choices with others. i like to pick other handicappers brain to see why they like a certain horse. anything that helps me to be thorough is a worthwhile endeavor. you want to be confident in the horse your playing, and considering all the options only helps you to gain confidence.

[lansdale]

Lansdale,

My computer keeps up with my results so I could tell you which races I have bet over the past 10 years (outside of the occasional visit to the track), how much I bet and how much I got paid if anything. I have no concern that I am mistaken as to the ROI.

I don't bet Kelly although as I pointed out above I vary my bet in relation to edge, odds and track in play. The largest bets are about 8 times the size of the smallest bets.

I don't use the bankroll concept and have never really believed in it. I have a certain size bet that I am confortable making and do so automatically and dont mind if I lose 50 in a row. I have increased the size of the basic bet somewhat over the years but probably by no more than a factor of 2 or 3.

If someone used Kelly and increased the bet size in relation to a growing bankroll the size of his bets would get very large in relation to what he is comfortable with and accustomed to playing and in some cases in relation to pool size. I imagine the larger players must come to grips with spreading among a number of pools so as not to dilute their own prices too much.

I have never made betting a full time thing or really gone after maximizing the profits to be made from it as I like my regular work and have good income from mainstream sources.

Thanks for you interest.

SJK

SJK,

Sorry to have misinterpreted your original post to mean that you yourself used Kelly bet-sizing, when, in fact, you were describing a customized system of bet variation. I think it's worth mentioning, though, for those unfamiliar with Kelly, that there are many kinds of bet variation, some worthless, some effective, that have nothing to do with the Kelly Criterion.

As for your comments on comfort level, there's no question you should play in whatever way is best for you. But what if we framed the issue of comfort level differently. What if you thought of your total betting capital as an investment fund with completely automated betting? If there was nothing do so but simply watch your investment grow? Would you be uncomfortable making too much money? And since the edge you have is much better than anything the stock market offers, the variance would be considerably less. Of course, I'm saying this less to you personally, than trying to illustrate a point.

Best of liuck, although you certainly don't need it.

Cheers,

lansdale

[sjk]

What I do is pretty close to fully automated and I have found that automation has rid me of the occasional instance of pruning some of the larger bets.

The actual bet size is formula driven I can raise a constant in the formula to make my bets bigger. I think I will scale things up a bit more. Thanks for the push.

[lansdale]

What I do is pretty close to fully automated and I have found that automation has rid me of the occasional instance of pruning some of the larger bets.

The actual bet size is formula driven I can raise a constant in the formula to make my bets bigger. I think I will scale things up a bit more. Thanks for the push.

SJK,

That sounds like a move in the right direction. FWIW, and even though you are not using Kelly I can add that most pro blackjack teams use 1/3 or 1/2 Kelly. William Benter used 1/2 Kelly. This is because, as many have observed, including Dave Schwartz on this site, 'full Kelly' can be a rough ride. Using fractional Kelly smooths out the variance. I'm not a mathematician, but if you were using 1/3 Kelly I would say your bank should be doubling every six months. Not bad.

If you do become interested in Kelly, the original 1956 paper is online, and I did mention Poundstone's book 'Fortune's Formula'. Although it may seem unrelated Don Schlesinger's 'Blackjack Attack', a bible of the blackjack world, contains much useful information on Kelly betting. Don, who is a retired options trader is a brilliant guy. Also a friend of the previously mentioned Ed Thorp.

Cheers,

lansdale