R-square test of various U.S. Tracks?

[maddog42]

I have always had the feeling that the smaller US tracks were easier to beat.The bigger tracks seemed to be a little harder. Has anyone out there done an R-square test of the public odds for various tracks? I don't want to steal any ones research or handicapping niche, but would like an answer (yes or no) to this question.

Another aspect of this is Wong's assertion that the Hong Kong tracks became much more efficient between 1988 and 2008, presumably because of Benter and his ilk. How about the U.S. market efficiency and its improvement?

[classhandicapper]

Another aspect of this is Wong's assertion that the Hong Kong tracks became much more efficient between 1988 and 2008, presumably because of Benter and his ilk. How about the U.S. market efficiency and its improvement?

I don't think there's any question that the US market is more efficient than it was a few decades ago when I started.

Part of it is that a lot of information that only a few players had access to privately (speed and pace figures, trainer patterns, replays etc...) are now available to everyone and part of it is that the casual player is dropping out. There also used to be a ton of dead money in the pools from the casual OTB player that enjoys scratch offs etc.. now instead. Now there isn't (except perhaps on very big race days like Derby, Preakness, Belmont, Breeder's Cup and at Saratoga).

[HalvOnHorseracing]

I have always had the feeling that the smaller US tracks were easier to beat.The bigger tracks seemed to be a little harder. Has anyone out there done an R-square test of the public odds for various tracks? I don't want to steal any ones research or handicapping niche, but would like an answer (yes or no) to this question.

Another aspect of this is Wong's assertion that the Hong Kong tracks became much more efficient between 1988 and 2008, presumably because of Benter and his ilk. How about the U.S. market efficiency and its improvement?
I have no specific data on an r-squared test, but I was a small track player (out of necessity) for a number of years. Annecdotally, I would say that you got more than your share of overpays and underpays, whenever a big (or medium) fish moved. Today, I get the sense that the favorites are often overbet in the WPS pools, and that can give some opportuities if you like another horse. I believe your best chance at the small tracks is to get into the more difficult verticals, but you have to balance your bets because the pools can be small.

One day at Mountaineer there were so many long prices that if you bet every horse in every race and boxed the field in the trifecta you'd have made good money.

[Secondbest]

I have always had the feeling that the smaller US tracks were easier to beat.The bigger tracks seemed to be a little harder. Has anyone out there done an R-square test of the public odds for various tracks? I don't want to steal any ones research or handicapping niche, but would like an answer (yes or no) to this question.

Another aspect of this is Wong's assertion that the Hong Kong tracks became much more efficient between 1988 and 2008, presumably because of Benter and his ilk. How about the U.S. market efficiency and its improvement?
At the big tracks it used to the sharks vs the minnows.Now it the sharks vs the orcas

[green80]

In the days before the OTB's and ADW's there was quite a difference in tracks as far as the skill of the crowd. Now any advantage you have due to the inept crowd at the track is quickly wiped out by money from the ADW's. Your best shot now is tracks the whales and CRW's tend to ignore due to low or no rebates.

[Capper Al]

Trust the force. Play a few very large tracks on fast days and play small tracks on fast days, and compare after a few cards on each.

[betovernetcapper]

I think HK racing is very efficient when it's efficient and when it's not, the outcome is due to the blanket finishes that occur with many turf races rather than trainer manipulation.

Re the smaller tracks, I've found that often the horses connections can play a much greater part then in larger tracks. At SA or SAR any trainer that gets stall space has a degree of competence. At both tracks most of the jockeys are skilled. At a small track you may find more then half of a given field being trained by people that haven't won a race in a year or two. For many years Dave Gall was the leading rider ( in terms of wins ) in the US. This wasn't because he was the best jockey in the US, he was just overwhelmingly superior to the jockey colony at FP. Some of those guys road more like cowboys then jockeys.

[ultracapper]

I think HK racing is very efficient when it's efficient and when it's not, the outcome is due to the blanket finishes that occur with many turf races rather than trainer manipulation.

Re the smaller tracks, I've found that often the horses connections can play a much greater part then in larger tracks. At SA or SAR any trainer that gets stall space has a degree of competence. At both tracks most of the jockeys are skilled. At a small track you may find more then half of a given field being trained by people that haven't won a race in a year or two. For many years Dave Gall was the leading rider ( in terms of wins ) in the US. This wasn't because he was the best jockey in the US, he was just overwhelmingly superior to the jockey colony at FP. Some of those guys road more like cowboys then jockeys.

Replace Gall with Baze and FP with BM and GG and you haven't changed the meaning of the sentence one bit. Seldom is the best jockey the leading win rider for the year. The small and mid tracks are always dominated by someone. When racing for limited purses, when your horse is ready to go, you have to get serious and put the best guy on it you can.

[ultracapper]

As for efficiency, from my observations at the Socal tracks, the favs are overbet just about every race. There is a 2-1 or 8-5 that should be double that in just about every race. At the same time, the fav that used to pay $6.40 now pays $3.40. There's a spot in there where there just aren't as many winners as there used to be. In Socal, the $16 to $20 winner is definitely rarer than it used to be. Now, if they have a legitimate shot, the get bet down to 9/2, and if they don't have a legitimate shot, they get lost on the board. I see sometimes 3 or 4 days in a row at SA without a $15 to $20 winning horse. Watch sometime. It won't be long before they go on a 25 to 30 race streak with nothing paying in that window. 70%-75% of the races in the streak will pay less than $15, and the remainder will pay over $20. Watch, it'll make you go...WHAT??? It's a dead zone sometimes. And unfortunately for us that were raised on the old "don't ever bet on anything below 6-1" philosophy, when you see a horse at 7-1 or 8-1, you know there is total confusion in the handicapping population concerning that horse. Crowd can't get a read on it.

[Cratos]

I have always had the feeling that the smaller US tracks were easier to beat.The bigger tracks seemed to be a little harder. Has anyone out there done an R-square test of the public odds for various tracks? I don't want to steal any ones research or handicapping niche, but would like an answer (yes or no) to this question.

Another aspect of this is Wong's assertion that the Hong Kong tracks became much more efficient between 1988 and 2008, presumably because of Benter and his ilk. How about the U.S. market efficiency and its improvement?
I don't know about the HK racing market, but looking at the US race wagering market, efficiency doesn't fit the EMH definition in the race wagering market as applied to stock market prices.

Also how can the coefficient of determination (r-squared) be applied to the nonlinear curve in the US wagering market?

I hope an explanation can be given beyond anecdotal rhetoric.

[garyscpa]

I don't know about the HK racing market, but looking at the US race wagering market, efficiency doesn't fit the EMH definition in the race wagering market as applied to stock market prices.

Also how can the coefficient of determination (r-squared) be applied to the nonlinear curve in the US wagering market?

I hope an explanation can be given beyond anecdotal rhetoric.

It's the resistance, of the bettors.

[Pensacola Pete]

The whales beat the little tracks to death because the rebates are generally higher.

And that's a good thing. The easiest way to beat the game now is to use the whales' predictability to advantage.

[Dave Schwartz]

And that's a good thing. The easiest way to beat the game now is to use the whales' predictability to advantage.

Considering that the money shows up in the pool after the gate opens, how do you do that?

[thaskalos]

The whales beat the little tracks to death because the rebates are generally higher.

And that's a good thing. The easiest way to beat the game now is to use the whales' predictability to advantage.

Are you suggesting that you can tell what these whales will do, BEFORE they do it?

Because they won't give you the opportunity to wager AFTER them.

[Cratos]

Considering that the money shows up in the pool after the gate opens, how do you do that?
A good reminder Dave and that is a reason that the EMH does not apply to the US racing wagering market; this market is inherently inefficient.

[Capper Al]

A good reminder Dave and that is a reason that the EMH does not apply to the US racing wagering market; this market is inherently inefficient.

If I'm reading this correctly, an efficient market would have a tight correlation between post odds and real value(something Quirin proved). Wouldn't an inefficient market be more desirable for wagering especially considering the take outs? In an inefficient market one could cheer pick their spots when the public misjudges.

[Dave Schwartz]

A good reminder Dave and that is a reason that the EMH does not apply to the US racing wagering market; this market is inherently inefficient.

Dick Mitchel once said that in order to be a professional horse player one must be able to tell the difference between a horse that was worth 7/2 and a horse that was worth 3/1. That's the difference between 22 and 25%.

Aside from the fact that there aren't like 4 people on the planet who can do that, today you must not only know the difference but must predict what the odds will actually be.


HOWEVER...

This thread got me to thinking about a whole new bunch of potential metrics.

Depending upon how your sort the pacelines for each horse, the metrics will "explain" a different view of the horse.

I am going to work on these and report back.


Dave

[Cholly]

Considering that the money shows up in the pool after the gate opens, how do you do that?

I’m curious: Is it a certainty that the so-called whales are laying off 100% of their bets at the last tick? I would have thought they are feeding the pools all along, to protect the price they want to get on a horse. There must be quite a bit of cat & mouse going on between the whales regarding the prices on horses, no?

Thank you in advance for your answer, and if it makes any difference, I only play the NYRA tracks plus CD & GP.

[Cratos]

If I'm reading this correctly, an efficient market would have a tight correlation between post odds and real value(something Quirin proved). Wouldn't an inefficient market be more desirable for wagering especially considering the take outs? In an inefficient market one could cheer pick their spots when the public misjudges.
An efficient market is a derivative of perfect competition which says that the correlation is between perfect information and market prices.

Therefore if the EMH is extrapolated to the horserace wagering market this market becomes efficient when all information is made equally available to all bettors which allows them to EFFICIENTLY wager along their risk curves, but given the current dramatic changes in the odds at some tracks after the horses are in the gate and the betting window is closed the wagering market in horseracing is inefficient.

[Dave Schwartz]

Is it a certainty that the so-called whales are laying off 100% of their bets at the last tick?

My sources have led me to believe that years ago they used to start at around 5 minutes to post but that the practice has changed as the odds predictions (on their end) have improved.

Do yourself a favor... Take a single day and note the exact pool amounts at the first flash after the tote board shows zero minutes and compare that to (what I call) the Winner Circle tote. You will be shocked at the results, especially at a high-rebate track.

At tracks with little or no rebate it will be quite different.

[Cholly]

My sources have led me to believe that years ago they used to start at around 5 minutes to post but that the practice has changed as the odds predictions (on their end) have improved.

Do yourself a favor... Take a single day and note the exact pool amounts at the first flash after the tote board shows zero minutes and compare that to (what I call) the Winner Circle tote. You will be shocked at the results, especially at a high-rebate track.

At tracks with little or no rebate it will be quite different.

Changed in that they're betting more money earlier?

Isn't a large part of that money coming in > 0 mtp small players like me betting $100 from their computer at home and punching it in at the last second? Surely the large players have algorithms that based on the obvious handicapping factors, multi-race will pays, opening odds (stable $), odds at 15, 10, & 5 mtp tell them what the final odds will be--they don't have to wait until the last second, they just have to keep from tilting the board too early, no?

[JohnGalt1]

What I notice at some smaller/medium tracks is 2 minutes to post there is very little money in the win pool.

Then zero to 1 minute to post the money comes pouring in, which I believe is from ADW's, and other track's simulcast money, and of course some from on track betters whose bets are accounted on the tote as soon as they are made.

I don't find it suspicious at all.

Ernie Dahlman said he bets huge early and then everyone, including the TV handicapper, notices his 1-9 selection offers no value and most play other horses.

The odds on his horse rises to what he expected. If he bet late, his huge bet kills the odds, with no time to recover.

[HalvOnHorseracing]

Dick Mitchel once said that in order to be a professional horse player one must be able to tell the difference between a horse that was worth 7/2 and a horse that was worth 3/1. That's the difference between 22 and 25%.

Aside from the fact that there aren't like 4 people on the planet who can do that, today you must not only know the difference but must predict what the odds will actually be.



There are a lot of things that make horseracing a difficult game and zoning in on value might be the toughest one. And you are right - predicting the final crowd odds makes the takes that much tougher.

Mitchell made a critical point about being able to accurately assess win probability, but I believe you can escape the necessity to be as precise as Mitchell suggests as long as you are willing to pass potentially good bets. Even the best aren't getting it precise 100% of the time, so it makes sense to build a margin or error, or a premium, into the bet. So if you think the horse is 25%, perhaps don't bet at less than 9/2, or whatever premium factor you want to apply. Even if you are exceptional at setting odds, you can't bet a 25% horse at 3-1 and do better than break even, so you have to demand some level of premium. Recordkeeping tells you whether you are a good oddsmaker.

[Dave Schwartz]

Isn't a large part of that money coming in > 0 mtp small players like me betting $100 from their computer at home and punching it in at the last second? Surely the large players have algorithms that based on the obvious handicapping factors, multi-race will pays, opening odds (stable $), odds at 15, 10, & 5 mtp tell them what the final odds will be--they don't have to wait until the last second, they just have to keep from tilting the board too early, no?

They are betting all there money as the horses enter the gate.

Somewhere around 55-60% of all the money shows up after the gate opens.


Mitchell made a critical point about being able to accurately assess win probability, but I believe you can escape the necessity to be as precise as Mitchell suggests as long as you are willing to pass potentially good bets. Even the best aren't getting it precise 100% of the time, so it makes sense to build a margin or error, or a premium, into the bet.

I agree. For anyone but the big operations it is just not feasible to analyze data at the level or with that accuracy.

Personally, I have not used the tote board to analyze value in 2 years.

[no breathalyzer]

Indiana Grand is the worst.. which is why i refuse to play there. i'm not gonna play guess which horse gets hammered while half the field is loaded! Last time i paid attention there like 85/90% of the win pool came in with 1/0 mins to post time. Which seemed extremely high considering the pool size