The Fundamental Theorem Of Horse Betting

[DeltaLover]

Most of the poker players should be familiar of the fundamental theorem of poker (http://en.wikipedia.org/wiki/Fundamental_theorem_of_poker)

I have been thinking for long time about a similar concept for horse racing.

I would like to know if you ever thought about it and if you have a definition of it.

[Capper Al]

How could you forget the one and only absolute truth of racing. Dave even has a video class on it, The Basics of Winning.

[Overlay]

If I were to put a fundamental theorem into words, it would be as follows:

Handicapping and wagering in horse racing should be structured so as to obtain maximum pari-mutuel efficiency from both generally available and proprietary information (that is, to use that information to provide as accurate an idea as possible of each horse's true winning probability), and also to account and compensate for information that is not or cannot be known (whether as a result of the nature/availability of data sources, the deliberate efforts of others, or the random events that occur during the running of a race).

[Robert Fischer]

Every time you play a race differently from the way you would have played it if you knew the optimum probability estimate and the optimum final odds estimate, your opponents gain; and every time you play a race the same way you would have played it if you knew the optimum probability estimate and the optimum final odds estimate, they lose. Conversely, every time opponents play a race differently from the way they would have if they knew the optimum probability estimate and the optimum final odds estimate, you gain; and every time they play a race the same way they would have played if they knew the optimum probability estimate and the optimum final odds estimate, you lose.

[pondman]

I don't believe there is a perfect strategy as in blackjack. I also don't think the probabilities of racing can be defined as you can when flipping a card.

I type horses. I have a reasonable estimate of the chance of each group of winning overtime. I place the unknown at 55-60% of that chance. And so I'll only play groups that are in that 40 percentile overtime. This means I pass on the majority of races.

I've spent the last few years debating the issue of jockey value and whether it's part of the group type or part of the epsilon. I'm beginning to believe it doesn't give value to the type but decreases the epsilon. Therefore a ringer can be ridden by an unknown and still fit in the +40% group.

[Robert Goren]

The idea behind the "fundamental theorem of poker" was to get most value out of your hand. Because others players make decisions based on your play in poker, it would only apply to whales in horse racing who under stand how the other other whales in the pool make their decisions. The closest thing to it in horse race betting for anybody other the largest bettors would be the " Kelly Criterion" modified for pool size. The "fundamental theorem of poker" is a betting tool, not a discovery tool. Trying to come with something similar to pick winners is a waste of time.

[Robert Goren]

I don't believe there is a perfect strategy as in blackjack. I also don't think the probabilities of racing can be defined as you can when flipping a card.

I type horses. I have a reasonable estimate of the chance of each group of winning overtime. I place the unknown at 55-60% of that chance. And so I'll only play groups that are in that 40 percentile overtime. This means I pass on the majority of races.

I've spent the last few years debating the issue of jockey value and whether it's part of the group type or part of the epsilon. I'm beginning to believe it doesn't give value to the type but decreases the epsilon. Therefore a ringer can be ridden by an unknown and still fit in the +40% group. Most of the members of this board should memorize this post. The idea behind what pondman is say is true. You don't have to look for 40%+ winners, but you should be looking for types of horses that show a high ROI with set likelihood of winning. I recently changed my style of betting have come up with this type of bet. I have found a type that wins between 25-30% of the time at 5/1+ odds. You don't get a lot of these bets, maybe one every two weeks or so. So you only cash every other month on average. I have now moved on to looking for other types while still monitoring that group to make sure it doesn't fall into a unprofitable wager. I firmly believe that long term typing of rare overlooked horse types is the one of the ways to making decent money at this game.

[Dave Schwartz]

You don't get a lot of these bets, maybe one every two weeks or so.

Goren,

You have more patience than Job.

And certainly far more than I.


Dave

[formula_2002]

"The fundamental theorem of poker is a principle first articulated by David Sklansky that he believes expresses the essential nature of poker as a game of decision-making in the face of incomplete information."

Thanks to the tote board, determing a horse racing out come is much simpler than determining poker outcomes. Horse racing information is more complete. You actually see the "cards" everyone is holding

Before making a bet you have a very good idea of the odds you are up against. Not so in poker.

[thaskalos]

Thanks to the tote board, determing a horse racing out come is much simpler than determining poker outcomes. Horse racing information is more complete. You actually see the "cards" everyone is holding

Before making a bet you have a very good idea of the odds you are up against. Not so in poker.
I'm not sure that information is more "complete" in horse racing.

In poker, you may find yourself playing against the world's best player...but you can at least take solace in the fact that he doesn't know anything more about the actual composition of the deck than you do.

In horse racing, there are some players out there who know much more about the "composition of the deck" than the rest of us do.

[TJDave]

In horse racing, there are some players out there who know much more about the "composition of the deck" than the rest of us do.

Trainers, owners, jockeys, grooms, racing secretaries, security etc.

[pondman]

Thanks to the tote board, determing a horse racing out come is much simpler than determining poker outcomes. Horse racing information is more complete. You actually see the "cards" everyone is holding

Before making a bet you have a very good idea of the odds you are up against. Not so in poker.

???
A 2 is a 2.
A 3 is a 3.

All hell can break loose in a race. I've bet horses that have died in races-- snapped their legs or just fell over-- and that was in high end BC races. I wish I could read that in the tote. What is the value of a horse that finishes 5th in a MSW at Del Mar? Do you really believe the crowd can tell you? I don't. If you study it enough you can get a reasonable estimate. But the tote...

[Tape Reader]

Thanks to the tote board, determing a horse racing out come is much simpler than determining poker outcomes. Horse racing information is more complete. You actually see the "cards" everyone is holding.

And that is why I bet off the tote board. I don't know what I don't know in reading the PPs.

I know what I know when I read the tote board.

[Robert Fischer]

"Everytime you bet the race the same way that you would have bet it if you could somehow see the race's actual outcome ahead of time...you gain. And everytime you bet the race differently than you would have bet it had you been able to see the race's actual outcome ahead of time...you lose.

Similarly, everytime the competition bets the race the same way that they would have bet it if they could somehow see the race's outcome ahead of time...THEY gain. And everytime they bet the race differently than they would have bet it had they been able to see the race's actual outcome ahead of time...they lose."

Thaskalos (who likes to leave as little as possible to chance) :)

Your theorem may be closer to Sklansky's poker theorem than mine, and as a bonus it is easier to read as well. :ThmbUp:

My theorem is constructed with the practical limits of the player in mind(we don't own a time machine), but Sklansky's poker theorem where he talks about "seeing all your opponent's cards" is analogous to your theorem.

So I went another step and referenced the 'estimates'.

Not sure why I originally felt the need to do that. :confused:

Anyway both are probably interesting to horseplaying theorists. :D

[DeltaLover]

"Everytime you bet the race the same way that you would have bet it if you could somehow see the race's actual outcome ahead of time...you gain. And everytime you bet the race differently than you would have bet it had you been able to see the race's actual outcome ahead of time...you lose.

Similarly, everytime the competition bets the race the same way that they would have bet it if they could somehow see the race's outcome ahead of time...THEY gain. And everytime they bet the race differently than they would have bet it had they been able to see the race's actual outcome ahead of time...they lose.

:ThmbUp:

This is exactly my defintion of the fundamental theorem of poker and what I had in mind when opening the thread!

Only this is I cannot see Thask's posting :confused:

[Dave Schwartz]

Originally Posted by formula_2002
Thanks to the tote board, determing a horse racing out come is much simpler than determining poker outcomes. Horse racing information is more complete. You actually see the "cards" everyone is holding

Before making a bet you have a very good idea of the odds you are up against. Not so in poker.

What you don't have is a solid set of probabilities.

In poker, you may well have that: a mathematical computation that is relatively precise and correctable.

In horse racing, you do not have that. In other words, if you give a horse a 40% chance of winning, if he wins, were you right? If he loses were you wrong?

[Robert Fischer]

:ThmbUp:

This is exactly my defintion of the fundamental theorem of poker and what I had in mind when opening the thread!

Only this is I cannot see Thask's posting :confused:
http://www.paceadvantage.com/forum/showthread.php?t=100585

parallel universe :cool:

[Robert Goren]

If you knew the outcome of a race before hand, you bet every penny you had on the winner somehow at 0 MTP. If you tried that in poker, your opponent would fold and you win only the antes.

[DeltaLover]

:eek:

So, I opened a thread with exactly the same topic as one that was discussed six months ago...

I am sorry about that, I don't know how I missed it.. :(

[formula_2002]

In horse racing, you do not have that. In other words, if you give a horse a 40% chance of winning, if he wins, were you right? If he loses were you wrong?

obviously not.

[formula_2002]

If you knew the outcome of a race before hand, you bet every penny you had on the winner somehow at 0 MTP. If you tried that in poker, your opponent would fold and you win only the antes.
how would he know what you know?

[Robert Goren]

how would he know what you know? He doesn't have to. He knows you are willing risk your entire fortune, that enough. Not only does he fold, He will probably get up and leave the table. I know I would.

[Robert Goren]

A lot of poker is betting just the right amount to convince your opponent that you have a good hand but not good enough to beat his. Enough poker stuff now and back to horse racing.

[Robert Fischer]

:eek:

So, I opened a thread with exactly the same topic as one that was discussed six months ago...

I am sorry about that, I don't know how I missed it.. :(

Great minds think alike ?

no problemo, makes for an interesting discussion, and I was hoping some of the serious poker players would contribute.

[Robert Fischer]

Every time you make a different parimutuel investment than the way you would have invested if you could see and understand your opponents' trading algorithm, they gain. Every time you make the same parimutuel investment you would have made if you could see and understand your opponents' trading algorithm, they lose. Conversely, every time your opponents make a different parimutuel investment from the way they would have if they could see and understand your trading algorithm, you gain. Every time they make the same parimutuel investment they would have made if they could see and understand your trading algorithm, you lose.

another interesting perspective :ThmbUp:

[MPRanger]

Most of the poker players should be familiar of the fundamental theorem of poker (http://en.wikipedia.org/wiki/Fundamental_theorem_of_poker)

I have been thinking for long time about a similar concept for horse racing.

I would like to know if you ever thought about it and if you have a definition of it.


My view on it:

In Limit Hold 'em and poker generally there is a mathematically best way to play every hand. Not only would all players play perfectly if everyone knew everyone else's hole cards. Even with the unknown hole cards there is still a mathematically best way to play a basic strategy. (No limit which I was born to play is a little different.)

52 exact cards are in the mix every round. Those cards are shuffled and spread around and over time everybody gets the same cards as everyone else. There is an optimal way to play each hand.

Obviously racing contenders are not identical and each race has it's own finger print even if it fits into some framework or guideline.

The only way to win over the long run in horse racing is to bet on overlays. Period.

There it is. That's the fundamental theorem of racing. THE ONLY WAY TO WIN OVER THE LONG RUN IS TO BET ON OVERLAYS.

There is not only a single kind of winner in racing. There are subsets of winners who may be overlays. Subsets can have vastly different characteristics. There are any number of strategies to find overlays in racing. Not a single optimum systematic method.

[raybo]

In Poker, there are times when you actually know, without a doubt, that you will win the hand, 100%. That is never true in racing.

[thaskalos]

In Poker, there are times when you actually know, without a doubt, that you will win the hand, 100%. That is never true in racing.

The two games are fundamentally different...and they don't have much in common.

Even the notion that in horse racing the battle is among the players and not against the house can be debated, and does not apply as much as it does in poker...because, when the house takes 26% of an exotics pool right off the top...then the house becomes just as big an adversary as the other players -- if not more so.

[Hoofless_Wonder]

Rather than a fundamental theorem with generally precise mathematical constraints, I see horse racing as much more analogous to fluid dynamics. I had the opportunity to study this while a meteorologist in the Air Force back in the '80s.

You have many variables, some known (trainer, jockey, post, earnings), some estimated (pace, speed figures from PPs, lengths beaten), and some partially or totally unknown (connections trying, number of combos used in the super pool, heart failure imminent), and some that are outright incorrect.

The horseplayer attempts to combine all known information, filling in the blanks for some of it, ignoring some of it, weighting all of it, forecasting a result and then determining if and where the inefficiencies in the parimutuel pools lie.

Sometimes it's as easy as forecasting hot weather in Florida in July. And sometimes it's more like trying to forecast fog, thunderstorms, freezing rain or snowfall amounts. In the thermodynamics and kinematics course I attended at Texas A&M, I recall the professor describing air parcel movements with something like 27 equations and solving for 27 unknowns. The problem is that many of the values approach zero, and are dropped out to achieve a solution, but of course even small variables influence the models of air movement over the longer run.

It's the same way with horse racing. The horse that gets bumped at the start may lose all chance in a sprint, but overcome it to easily win a route. There are poor estimates and more accurate estimates, and as MPRanger points out, the only way to stay ahead of the crowd is with overlays - where the risk/reward ratio is positive, and where the player has simplified their process from reality so that it produces long-term earnings.

Or, at least for me, be happy with nailing a race now and then, and being entertained. Just like golf, it's not the 112 score and the three dozen bad shots I made that keeps me coming back - it's the one 18 foot putt, the one 250 yard drive that stayed in the fairway and the six iron to within three feet of the hole during the round that I remember and that keeps me motivated...

[raybo]

The two games are fundamentally different...and they don't have much in common.

Even the notion that in horse racing the battle is among the players and not against the house can be debated, and does not apply as much as it does in poker...because, when the house takes 26% of an exotics pool right off the top...then the house becomes just as big an adversary as the other players -- if not more so.

I think the fundamental difference between horse race betting and poker/blackjack betting is that in horse racing there is no fundamentally "correct" way to handicap and bet. Our handicapping must be very accurate in its separation of true contenders from true non-contenders, and our wagering must accurately reflect the risk/reward ratio, while being based on a vast amount of unknowns. And, even when we are accurate enough in both of those things, we still can't rely on the projected outcome actually taking place.

In other words we can be "perfectly" correct in our total approach, and still lose.

[Horseplayersbet.com]

In Poker, there are times when you actually know, without a doubt, that you will win the hand, 100%. That is never true in racing.
Except when you have an all in a horizontal wager, though you can't wager once you know you will win...well, you can actually, if you make cover bets when you have all or most horses in a last leg and you want to guarantee certain profits.