Place Vs. Fave/Long Exacta Running Tab

[SniperSB23]

Originally Posted by blackthroatedwind

The truth of this contest is that an aberational result which caused place bets to appear the better bet would have sent a TERRIBLE message to people. This idea that place betting is a sound strategy is flat out wrong and yet another reason people have trouble making money betting horses. The idea is to find ways to maximize one's profits or returns. Place and show betting, simply put, minimizes returns. Place and show bettors are suckers and losers at the windows.

You are of course evaluating this on the basis of someone looking to make money through wagering as opposed to people wagering as a hobby. For people wagering as a hobby I don't think they are going to be happy to know that statistically they made the right play when they bet an 80-1 shot under the favorite in the exacta and it winds up second behind another horse and they get nothing. It is actually more Game Theory than anything. Some people are risk seekers and want the best longterm payout and are willing to come up with nothing on several occassions when they are right on a long shot. Others are risk averse and want the utility (or happiness) of getting a payoff every time. That happiness of winning every time you are right is worth more to them than the difference in both methods for the longterm.

This is why it drove me nuts majoring in statistics and economics. They contradicted each other in so many ways. I could give a 100% statistically correct answer to a question on an economics exam and get zero points for it.

[Dunbar]

Originally Posted by blackthroatedwind

This is an absurd mentality. What does " if " mean? If some of those ridiculous big priced horses at Gulfstream hadn't won, or had run second to the favorite, the exacta would have been substantially higher. " IF " is absurd to use.

That's correct. However, when the difference after 250 bets is less than many of the individual exacta payoffs, you can be reasonably sure that the result is statistically meaningless. If one side or the other had finished ahead by 3 or 4 exacta bets, we could at least say that the result supported one side or the other.

--Dunbar

[blackthroatedwind]

Originally Posted by Dunbar

That's correct. However, when the difference after 250 bets is less than many of the individual exacta payoffs, you can be reasonably sure that the result is statistically meaningless. If one side or the other had finished ahead by 3 or 4 exacta bets, we could at least say that the result supported one side or the other.

--Dunbar

I understand, and I believe the place bets had a strong run of luck in this short sample.

I understand what you said in the previous post but I am not concerned with people who play the game for a hobby ( and only wish there were more of them ) as I think people who are following this honestly want to better their play and results. Place betting is a major mistake if one is trying to do that.

[Dunbar]

Originally Posted by blackthroatedwind

There is no need to wait for future results. You can just as easily use past results.

Are you volunteering? ;>) I'd be interested in past results, as long as the begin and end points are clearly stated before whoever does the looking starts checking.

Originally Posted by blackthroatedwind

The reason the exacta is better is for a variety of reasons. One, the breakage hurts the place payoffs enough to raise the takeout to at least a couple points higher than the 15 or 16%.

No, it doesn't. The average place payoff was about $15. The average breakage on a $2 bet is 10 cents except in NY where it is 5c. 10 cents out of $15 is 0.67%. 5 cents out of $15 is 0.33% So the takeout climbs to 16.67% in Kentucky and 15.33% at AQ. Those increases hardly made a dent in the gap with the exacta takeout. (Besides, with an average exacta payoff of $51, the exacta takeout climbed 0.20% in Kentucky and 0.10% in NY due to breakage). In short, breakage has negligible effect on the contest.

Originally Posted by blackthroatedwind

Another is that favorites are overbet in place pools for the most part and as they finish first or second around 50% of the time they will artificially deflate place payoffs on the other horse.

That's an interesting idea, but I'd like to see some data. It should be very easy to substantiate that. If it's true, then place bets placed on favorites should lose at more than the track take. (I can't think of any other definition of "overbet" that would apply here.)

Originally Posted by blackthroatedwind

But, the biggest reason is that with exactas you are effectively making a parlay of two different results, one horse to win and another to place, with a takout of roughly 20%. It's not wholly dissimilar to making a football parlay on the point spread AND the over/under while only paying the vig for one bet. So, even though the exacta takout is, say, 20% vs. 15% for place, you are effectively ( though not exactly ) lowering it to 10% on each outcome. Now, when you add on the actual increase that breakage brings to the place takeout you have a substantially better mathematical proposition.

This may be correct, and it's what I thought you were getting at when you said there was a "mathematical" reason earlier. Maybe it should be obvious to me, but it isn't yet. I plan to keep gathering data for another 250 bets, but at the same time I'm working on the math of a simple example that I hope will demonstrate which is better.

Originally Posted by blackthroatedwind

The truth of this contest is that an aberational result which caused place bets to appear the better bet would have sent a TERRIBLE message to people. This idea that place betting is a sound strategy is flat out wrong and yet another reason people have trouble making money betting horses. The idea is to find ways to maximize one's profits or returns. Place and show betting, simply put, minimizes returns. Place and show bettors are suckers and losers at the windows.

I think that's excellent general advice. However, I don't think it excludes the conclusion that place bets on > 10-1 horses are a better bet than betting the same horses in exactas using the race fav in the 1st position of the exacta.

--Dunbar

[blackthroatedwind]

NY does not break to ten cents above $10 I don't believe.

I'm too weary of this to argue anything else. I may spend a few hours up at DRF going through past data....if not this week then soon.

[SniperSB23]

Honestly I don't think you can say that one is right or wrong all the time. Perhaps on average the best play is the favorite over your horse in the exacta but really the best play is to examine the percentage of your horse and the favorite in the place pool and project a payout and then examine what the exacta probable pays. I think there are plenty of cases where these pools would tell you that for that particular race betting place is a better play.

[Dunbar]

Originally Posted by blackthroatedwind

NY does not break to ten cents above $10 I don't believe.

I'm too weary of this to argue anything else. I may spend a few hours up at DRF going through past data....if not this week then soon.

Interesting about NY. Didn't realize that.

If you do get into the past data, I'll be very interested in how it comes out.

Also...

Originally Posted by Dunbar

The average place payoff was about $15. The average breakage on a $2 bet is 10 cents except in NY where it is 5c. 10 cents out of $15 is 0.67%. 5 cents out of $15 is 0.33% So the takeout climbs to 16.67% in Kentucky and 15.33% at AQ.

I should have written "the average loss to breakage...", not "the average breakage...". But the percentages are correct, except that the NY cost should be the same as the KY/CA tracks for place payoffs above $10.

--Dunbar

[Dunbar]

It's all over, folks! Well, sort of over....

Day 34 (No dark days included) April 18, 2007
Carryover differential: PLACE BETTING +$80.50

AQ, KE, SA

AQ: Race1-- PL 15.40 EX 64.00
AQ: Race2-- PL 14.20
AQ: Race4-- PL 32.40


Today's races: EX $64.00, PL $62.00
Today's differential: EX +$2.00

Final Totals: EX $3853.40, PL $3770.90
OVERALL DIFFERENTIAL: EX +$82.50

Total Number of 10-1's or higher in spots 1 or 2: (250) Will Stop=250
Total Exactas so far: 75 (30%)

Total races run at all tracks during the contest--755 (we stopped as soon as we hit the 250th place bet. I'm including 3 races at Keenland in the total of "races run".)



Congrats go to Grits for winning the bet with Randall.

The exacta bets won by $82.50. That's 2.2% more than the Place bets. However, the $82.50 margin is less than many of the exacta bets. It looks like we will need a LOT more races to see a difference that is big enough to consider meaningful.

My intention is to follow this out for another 250 place bets. However, if BTWind comes back with some historical data that is compelling, I'll probably hang it up.

--Dunbar

[golfer]

During this whole excercise, I've never understood why it has to be THE favorite on top of your longshot play? Why, if you like a longshot, would you just robotically play the favorite on top of your horse anyway? What about using a few horses, or even the all button (although this might dilute your return on investment)? Andy, my question has always been, would I be better off playing my choice to place (obviously to win as well), or underneath (and on top) in exactas. I would reasonably put my horse underneath all that I believe have a chance to win, no?

[blackthroatedwind]

Originally Posted by golfer

During this whole excercise, I've never understood why it has to be THE favorite on top of your longshot play? Why, if you like a longshot, would you just robotically play the favorite on top of your horse anyway? What about using a few horses, or even the all button (although this might dilute your return on investment)? Andy, my question has always been, would I be better off playing my choice to place (obviously to win as well), or underneath (and on top) in exactas. I would reasonably put my horse underneath all that I believe have a chance to win, no?

You should never bet to place. That's where you start. I'll try to answer better when I'm more awake tomorrow.

[docicu3]

I have wandered in here on page 18 so I probably have missed far more than I will be able to use successfully but if I take this one step further......

Andy if you have an angle on a LS say at better than 10-1 and you really like a horse to win which you play with your typical "X" or unit play. Does not the concept of "saver bet" at times on long odds with horses you have already played to win make sense. Although I want the 33-1 payoff 27/$2 and 14/$2 are not returns that will likely make you a loser for the day.
I thought the most important thing at the end of the day,other than making some kind of profit was to be able to look back at all of your wagers and objectively evaluate risk and return favorably. You can certainly on occasion win by playing stupidly, albeit not very often, but making sound wagers should eventually give you a decent chance long term to be a winner over all.

An example is a 33-1 shot that paid 68-27-14 allow you to still make quite a profit if the horse finished 2nd or 3rd.

I certainly agree that playing short odds will likely make betters go broke but there has to be a time when investing in returns that can be worth the risk at certain potential ROI is a sound play.

DrD

[blackthroatedwind]

It doesn't matter how you do on any given day except psychologically. And, when your " longshot " runs second and you get nothing you need to remember the times you won five times as much as you would have with a place bet when you hit your exacta savers.

[SentToStud]

Originally Posted by docicu3

I have wandered in here on page 18 so I probably have missed far more than I will be able to use successfully but if I take this one step further......

Andy if you have an angle on a LS say at better than 10-1 and you really like a horse to win which you play with your typical "X" or unit play. Does not the concept of "saver bet" at times on long odds with horses you have already played to win make sense. Although I want the 33-1 payoff 27/$2 and 14/$2 are not returns that will likely make you a loser for the day.
I thought the most important thing at the end of the day,other than making some kind of profit was to be able to look back at all of your wagers and objectively evaluate risk and return favorably. You can certainly on occasion win by playing stupidly, albeit not very often, but making sound wagers should eventually give you a decent chance long term to be a winner over all.

An example is a 33-1 shot that paid 68-27-14 allow you to still make quite a profit if the horse finished 2nd or 3rd.

I certainly agree that playing short odds will likely make betters go broke but there has to be a time when investing in returns that can be worth the risk at certain potential ROI is a sound play.

DrD

DrD,
I disagree fairly strongly. I think the ONLY time a place or show wager can make any sense is when you are betting against a huge minus pool chalk you think is vulnerable to running out of the pool. But that's not what we're talking here.

I'm certain if you took a look at every 30-1 shot that ran first or second over a large enough sample, you will find a higher ROI betting those horses just to win as opposed to splitting your wager WP or WPS.

If you like a horse at long odds, it's fair to assume you find percieved value because you feel the favorite is underlaid. In other words, a horse you feel is fair at 10-1 that is going off at 20-1 is doing so most likely because a favorite is going off at 4/5 instead of 5/2.

In my mind, the only way to play this horse is to win and in vertical exotics with the non-overlaid chalk contenders. Favored horses (deserving chalk and otherwise) are overbet in Place/Show pools and are increasingly overbet in vertical exotics.

I think if a horse is 4/5 and is deserving, it's fruitless to try to beat the race betting any horse to win or playing vertically. "Deserving" is, of course, subjective but if 40-50% of the pool is correct, you are just not going to find value elsewhere. In other words, if I see a horse at 15-1 that I thought was fair at 10-1 in a race with a short favorite worthy of its odds level, I'm going to pass because it's very likely that I'm wrong in my opinion of the longshot's fair odds and there's no way to beat the place or exacta pool.

I think WP or WPS wagering is a bit of a psychological phenomenon people use to cash tickets and, to an extent, preserve bankroll on a short term basis.

So I think the only time to bet a horse at what I percieve is an overlay price is when there is a false favorite simply because there is no other way over the long run that makes my horse an attractive overlay. And given that the bad chalk is increasingly overbet in the vertical pools, I think it's mandatory to use my horse in verticals with the other non-chalk contenders. These are the kind of races and betting opps we live for and you have to get your money in the the right way when you can.

Sorry to ramble. Thats my opinion.

[blackthroatedwind]

While I agree with a lot of what you said I think you are making a mistake in the latter part of your post, or perhaps I am misreading you, as suppose you like that 15-1 shot ( and I understand what you mean when you say maybe you were " wrong " ) but also think the favorite is deserving. Isn't that a situation where at the very least you should have an exacta with your horse under the favorite ( and perhaps a small reverse )?

As an add-on, not relevant to what you posted necessarily, I think people need to pay more attention to how much exactas pay when a strong favorite runs second. To me these are far and away the best value payoffs available. I can't tell you how many 3-1 shots I have seen beat odds on favorites with the exacta paying more than double the win price. If a horse is considered over 50% to win the race just imagine what it's chances are of finishing second....especially when the second or third choice wins the race. It's worth noting.

[Dunbar]

Originally Posted by blackthroatedwind

Another is that favorites are overbet in place pools for the most part and as they finish first or second around 50% of the time they will artificially deflate place payoffs on the other horse.

Originally Posted by Dunbar

That's an interesting idea, but I'd like to see some data. It should be very easy to substantiate that. If it's true, then place bets placed on favorites should lose at more than the track take. (I can't think of any other definition of "overbet" that would apply here.)

I'm re-thinking the highlighted part of what I wrote. It turns out to NOT be easy to substantiate it.

For a 15% track takeout, the average return for a $2 place bet should be 1.70. When btwind says that favorites are overbet in the place pool, I think that means that $2 bet on all favorites must return an average of less than 1.70.

Let's say we look at 500 favorites and take the average payout of a $2 place bet on each one, including using $0 payout for the losers. If our average comes out to be $1.64, which would suggest that place bets on favs lose at 18%, is that difference from $1.70 statistically meaningful? I don't think so. I estimate the standard error of the mean* on those 500 bets would be 0.14. That's more than twice as large the 0.06 difference between $1.70 and $1.64. That, in turn, means there there is a very good chance that the "real value" for place bets on favs is $1.70, even though our 500-bet average is $1.64. In other words, the difference of 0.06 would be statistically meaningless.

To get the standard error down to 0.03, you'd need to look at the place payoff of 10,000 favorites!

I'm putting the details and math in a footnote, so that if I'm making a blunder, someone can point it out.

--Dunbar

* If we looked at 500 $2 place bets on favorites, about half of them will return $0, and the rest will return $2.10 up to $7 or more. But about 2/3 of the payoffs would lie in the range of $0 to $5. So we can estimate the standard deviation (of place payouts on favorites) at about $3.

When you take an average of something, you can estimate the uncertainty in that value with "the standard error of the mean". You get the standard error by dividing the standard deviation by the square root of the number of values in your average. For 500 place bets, the square root of 500 is 22. So, the standard error on the mean would be $3.00/22 = 0.14.

If you are trying to demonstrate that 2 numbers are different, you'd like them to differ by AT LEAST one standard error, preferably two standard errors. In the case of distinguishing a 15% takeout from a 20% takeout, the difference in average values (0.06) is much less than even one standard error after 500 bets.

If we looked at 10,000 favs, then square root of 10,000 is 100. The standard error becomes $3/100 = 0.03. Now, an observed difference of 0.06 would be meaningful. (because it would be 2 standard errors.)

[Dunbar]

What I just wrote in my previous post also applies to the "Contest" between place bets and exacta bets, except even MORE so because of the larger range of payoffs involved. (a larger range of payoffs results in a larger standard deviation.)

Unless one side is WAY better than the other, it is going to be impossible to detect the difference with any sort of statistical confidence without looking at thousands of place bets.

With that in mind, I see no point in tracking an additional 250 place bets, as I had said I would earlier. To resolve the Contest issue, we need someone with a huge queriable (is that a word?) database of results, like I think Cramer used to do for Barry Meadow's monthly publication.

--Dunbar

btw, I think the 3 posts above by btwind and S2S are particularly good.

[golfer]

OK, I understand no place betting. But, assuming a $100 total investment, how much do you play with your longshot on top/win bet compared to underneath in exacta... $70/30? $80/20? $90/10?

[SentToStud]

Originally Posted by blackthroatedwind

While I agree with a lot of what you said I think you are making a mistake in the latter part of your post, or perhaps I am misreading you, as suppose you like that 15-1 shot ( and I understand what you mean when you say maybe you were " wrong " ) but also think the favorite is deserving. Isn't that a situation where at the very least you should have an exacta with your horse under the favorite ( and perhaps a small reverse )?

As an add-on, not relevant to what you posted necessarily, I think people need to pay more attention to how much exactas pay when a strong favorite runs second. To me these are far and away the best value payoffs available. I can't tell you how many 3-1 shots I have seen beat odds on favorites with the exacta paying more than double the win price. If a horse is considered over 50% to win the race just imagine what it's chances are of finishing second....especially when the second or third choice wins the race. It's worth noting.

I guess it's nuanced, as most handicapping theories are. The shorter the deserved favorite, the less likely you are going to get any value elsewhere, win-wise or in the verticals. I suppose the Exacta play you mentioned is viable, but only if you're getting the right prices there. Usually it's not there, at least to the extent that you're beating the t/o. And some of it depends on personal style and preference. For me, unless the Ex pool is real favorable, I'd rather pass on races like this and look for a better (or maybe just different) single race opp.

[Dunbar]

Originally Posted by SentToStud

I suppose the Exacta play you mentioned is viable, but only if you're getting the right prices there. Usually it's not there, at least to the extent that you're beating the t/o. And some of it depends on personal style and preference. For me, unless the Ex pool is real favorable, I'd rather pass on races like this and look for a better (or maybe just different) single race opp.

As a general rule, there is more likely to be value in the play btwind mentioned than any other play.

I spent a fair amount of time looking for overlays in the exacta pool, given 2 assumptions: (1) that the win odds are approximately correct, and (2) that the odds of a horse coming in 2nd to another (particular) horse coming in 1st can be approximated by artificially removing the 1st place horse from the odds list and recalculating the odds on the remaining horses. (the Harville formula assumption). There are many races where either or both those assumptions are poor. Still, IMO they are a good starting point for evaluating bets in the other pools.

In the exacta pool, the situation that most often showed value was precisely the one that btwind singled out--using a heavy fav as the 2nd part of an exacta.

I found that result surprising, because the tendency of many bettors to play an exacta box works against it.

--Dunbar

[SentToStud]

Originally Posted by Dunbar

As a general rule, there is more likely to be value in the play btwind mentioned than any other play.

I spent a fair amount of time looking for overlays in the exacta pool, given 2 assumptions: (1) that the win odds are approximately correct, and (2) that the odds of a horse coming in 2nd to another (particular) horse coming in 1st can be approximated by artificially removing the 1st place horse from the odds list and recalculating the odds on the remaining horses. (the Harville formula assumption). There are many races where either or both those assumptions are poor. Still, IMO they are a good starting point for evaluating bets in the other pools.

In the exacta pool, the situation that most often showed value was precisely the one that btwind singled out--using a heavy fav as the 2nd part of an exacta.

I found that result surprising, because the tendency of many bettors to play an exacta box works against it.

--Dunbar

Thanks and I appreciate you sharing as well as BTW's thoughts. I do think the Exacta pools are very important. When playing from home, I use two computers. One is tuned into the Exacta pool grid about 90% of the time. To me, it's the biggest advantage of wagering over the internet.