Change Your Game Forever

[thaskalos]

Originally Posted by Hoofless_Wonder

I have not noticed the late money to move the odds on that large of percentage of races - maybe one in five? With the exception of the Aussie racing at night with the ultra small pools, it seems the odds are usually within a tick or two of what they were with a couple minutes to post.

But I completely agree that an accurate betting line is outside the ability of most players - one of the dangers of using a computer is that if you want a "betting line" calculated, you can get one. How accurate it might be is a whole different ball game.

I would go as far as to say that a profitable betting line is outside the ability of all but the very few.

As far as the late changes in the odds are concerned...it isn't a matter of how often these drastic late changes come about; it's about whether or not we feel comfortable betting on a supposed overlay...when we know that it could become an underlay by the time the race is over.

I recently bet on a horse at Woodbine which went off at odds of 3/2 in a full field. The race actually started with the horse showing 3/2 on the board...which I thought presented great value. The horse eventually paid $3.80 to win.

It may not happen very often...but it erodes one's confidence in "value wagering" just the same.

And it happens much more often at the minor tracks.

[Hoofless_Wonder]

Originally Posted by thaskalos

I would go as far as to say that a profitable betting line is outside the ability of all but the very few.

As far as the late changes in the odds are concerned...it isn't a matter of how often these drastic late changes come about; it's about whether or not we feel comfortable betting on a supposed overlay...when we know that it could become an underlay by the time the race is over.

I recently bet on a horse at Woodbine which went off at odds of 3/2 in a full field. The race actually started with the horse showing 3/2 on the board...which I thought presented great value. The horse eventually paid $3.80 to win.

It may not happen very often...but it erodes one's confidence in "value wagering" just the same.

And it happens much more often at the minor tracks.

Excellent example. I forgot about those sudden "steam" avalanches at Woodbine. I've been stung on a few horses there at say 8 or 10-1 that have their odds cut in half as they run the race.

I guess you have to take something like Dave's approach and come up with an adjusted line that calculates what the final odds will be, or just live with it. Sometimes the odds drops are somewhat predictable, but certainly not always.

This is one of the reasons I miss the "fixed odds" wagering that IASBET offers, but of course not to U.S. citizens any more. There was no guessing if you thought late money was going to come down on your selection.

[AndyC]

Originally Posted by Dave Schwartz

Such a play may be hugely profitable but not an appropriate play with a substantial bet (relative to bankroll).

I don't know of a money management system that suggests an inappropriate sized wager for a low probability winner. So people making substantial wagers on low probability horses that have "pure value" really need a better grasp on money management.

[AndyC]

Originally Posted by PICSIX

It's really very simple, but, I couldn't see it through the "value" haze I was in...."value" doesn't equal winner. Who in their right mind bets a horse to win that isn't the most likely winner (in his or her estimation)?

Let me reverse the question. Who in their right mind would bet a horse simply because they were the most likely winner but the odds were far less than the probability of winning? So let's roll some dice. I will let you bet on a 7, 4, or 10. The odds for a 7 will be 3-1 and the odds on the 4 and 10 will both be 20-1. If this were horse racing you would tell me I am not in my right mind if I didn't bet the 7. Most likely winner doesn't equal profit.

Your problem has nothing to do with "value betting" and everything to do with your ability to assess value.

[PICSIX]

Originally Posted by AndyC

Your problem has nothing to do with "value betting" and everything to do with your ability to assess value.

Me and 99% of the people that post here!! That's my point....it's nearly impossible to make an accurate odds line and even more impossible to incorporate the late odds changes of today.

Most of us are better off handicapping the most likely winner and betting when the odds are 2-1 instead of passing and watching the win because our acceptable odds were 5-2.

[AndyC]

Originally Posted by PICSIX

Me and 99% of the people that post here!! That's my point....it's nearly impossible to make an accurate odds line and even more impossible to incorporate the late odds changes of today.

Most of us are better off handicapping the most likely winner and betting when the odds are 2-1 instead of passing and watching the win because our acceptable odds were 5-2.

It all depends on your objective. If you just crave the action then fire away. If you are trying to end up in the black then you really need to know that what you are betting is more likely than not an overlay.

I believe that the best way to find value is to specialize. Be it tracks, trainers, types of races, types of bets, physical handicapping, trip handicapping, etc. Be one of the best at a specialty and finding value will become much easier. Too many people "play" the races expecting to win. They have enough knowledge to make enough winning plays to keep them coming back but lack the expertise necessary to become a regular winner.

[garyscpa]

Originally Posted by PICSIX

In all honesty, the "value wagering" discussed here took me from a profitable weekend warrior to a losing one. I spent a long long time trying to figure out why.

It's really very simple, but, I couldn't see it through the "value" haze I was in...."value" doesn't equal winner. Who in their right mind bets a horse to win that isn't the most likely winner (in his or her estimation)?

If I were in a fight with Mike Tyson, would you bet on me because the odds of Iron Mike dropping dead, just before he hit me, offered "value"

Yes.

[TrifectaMike]

Originally Posted by thaskalos

IMO...the negatives of "value wagering" are not discussed enough on this board.

If I weren't as experienced or as jaded as I am, and I just went by what I read on this board... then I would get the impression that there are many winning players out there who are in possession of accurate betting lines...which are capable of accurately assessing the supposed value presented on the tote board. And that all these winning players, who possess these accurate betting lines, have nothing at all to fear from this avalanche of late money that shows up during the last seconds of wagering.

I've never seen even one of the "value player" here explain how he handles this "late money" problem. How can this late money not wreck the long-term plans of all these non-whale "value players"? Do they think that the affect of this late money evens out in the long run?

It doesn't even out. The effect is NOT a nice normal distribution which has symmetry. It is a very complicated process, which requires several different oddslines.

Mike

[Dave Schwartz]

It doesn't even out. The effect is NOT a nice normal distribution which has symmetry. It is a very complicated process, which requires several different oddslines.

Oh, we have a new contestant, Johnny!

Please expand on THIS idea! I am all ears.

(Not something completely foreign to me, though.)

[cj]

Originally Posted by thaskalos

I've never seen even one of the "value player" here explain how he handles this "late money" problem. How can this late money not wreck the long-term plans of all these non-whale "value players"? Do they think that the affect of this late money evens out in the long run?

Easy, avoid short prices. If you are betting horses 5 to 1 and up, it not only evens out, the odds swings tilt in your favor. You just have to be good at picking horses 5 to 1 and up!

[TrifectaMike]

Originally Posted by Dave Schwartz

Oh, we have a new contestant, Johnny!

Please expand on THIS idea! I am all ears.

(Not something completely foreign to me, though.)

Now, Dave calm yourself down. You don't believe I would give details. Just enough info to let Thask know that it is being done and what maybe required.

Mike

[ElKabong]

Originally Posted by cj

Easy, avoid short prices. If you are betting horses 5 to 1 and up, it not only evens out, the odds swings tilt in your favor. You just have to be good at picking horses 5 to 1 and up!

something we totally agree on here.

however just last weekend I bet 2 horses at 5-1 (1 mtp) or better, one went down to 5-2 from 5-1.....the other was 6-1, bet down to 3-1. Lost both the bets, to be honest seeing the odds drop made the losses easier, as strange as that sounds

But the 5-1 rule is one to live by (at least at the simo), imo

[Robert Goren]

Originally Posted by ElKabong

something we totally agree on here.

however just last weekend I bet 2 horses at 5-1 (1 mtp) or better, one went down to 5-2 from 5-1.....the other was 6-1, bet down to 3-1. Lost both the bets, to be honest seeing the odds drop made the losses easier, as strange as that sounds

But the 5-1 rule is one to live by (at least at the simo), imo

What track were you betting?

[TrifectaMike]

There are many misconceptions and misinformation on how a useable oddsline is created.

Here are the steps in general:

Step 1: Data analysis

Each predictor has to be carefully studied. The majority of predictors are hierachical and should be modeled as such. Models for the predictors need to be defined.

Step 2: Binary Regression ( winners and losers )

Determine the structure of the model...test, test, and more testing.
Confounding variables, random effects, overfitting problems, etc. Compare with other modelling procedures (for ex: compare Binary to Random Forest for confirmation of significance).... combine Binary with Random Forest, etc.

Step 3: Model the public

Determine apriori what you expect the public to do. test, test and more testing.

Step 4: Follow the odds or Maximize ROI

Optimize the model by odds smoothing or NOT!!!!!!!!!!!!!!!!!!!!
Penalize model for coincidence with tote-odds and optimize model with respect to the inverse of the public's confidence.

Step 5: Know the distribution about each predictors coefficient

The odds for each horse are not deterministic. They also have a distribution. ( Ex: Horse A 4/5 to 1-1 with a probability of .90)

Finally tweak, test and test some more and monitor the process carefully.

This is it in a general sense. There are many details along the way that need to be addressed.

Mike

[erikeepper]

Originally Posted by Clocker

Which you are not going to share?

English is not my native language so it's hard for me to explain everything in detail. I have also received PMs agreeing with me.

I honestly thought (oh btw, and I am not going to ask my money back, I will support Daves work) there wasnt really nothing new FOR ME. Maybe I am just lucky. I have been a very close friend with a guy who plays almost 10 million euros per year. He has taught me a lot of things and also subsconsiously I have learned a lot from wathing him. I was like "yes, that is true, tell me what I dont already know" during the whole course and then it ended....

I will give it 2nd go, maybe I misunderstood something. Like I said earlier, that was my first impression of the course.

I agree with Thask that the "point" and "ranking" system the most thought-provoking, although I do that already also.