show pools converted into odds [Archive] - Horse Racing Forum - PaceAdvantage.Com

highnote

08-11-2017, 09:39 PM

Is there any software that will convert the show money on each horse into odds,or any formula to do it longhand?
Thanks
Mikeb

Buy a copy of "Dr. Z's Beat The Racetrack" by Dr. William T. Ziemba and Dr. Donald Hausch.

They give approximations of the formulations needed to calculate the values.

The actual method is called Generalized Reduced Gradient. It's an optimization problem. It's complicated.

If you're handy with Excel you could probably solve the equations using the Solver add-on.

If your math ability is like mine then the approximation will have to suffice.

AltonKelsey

08-11-2017, 09:51 PM

Generalized Reduced Gradient ?

For what.

Is a show pool, not dark matter.

As mentioned there will be a range of payoffs, low to high. easy to figure.

Otherwise, what kind of 'odds' are you looking for?

highnote

08-11-2017, 09:58 PM

Generalized Reduced Gradient ?

GRG is one method for computing all the ijk in the money finishes.

If there are 10 horses in the race and any three of them can finish 1-2-3 (that is, ijk) then it can get pretty complicated.

What is the probability of any particular i-j-k finish?

If i and j finish first and second then any horse can finish third except for i and j.

AltonKelsey

08-12-2017, 01:07 AM

I don't think that's what the orig poster was asking for, but who knows.

highnote

08-12-2017, 01:35 AM

I don't think that's what the orig poster was asking for, but who knows.

It doesn't matter. Beat the Racetrack is a good book for people who want to have a deeper understanding of place and show betting.

hracingplyr

08-12-2017, 12:43 PM

The new software cost is.

1,297 if you have the current Black Magic.

1,797 If you are a new customer.

ARAZI91

08-12-2017, 01:16 PM

highnote

08-12-2017, 01:29 PM

Highnote

Hi , i remember first reading about Dr Z's Tote system in Nick Mordin's first Book years ago and wishing we(the UK) had a stronger Tote system.
Ziemba's book 'Beat The Racetrack' can be read online for free at openlibrary.org along with numerous other old handicapping classics.

I enjoyed the podcast interview with Ziemba that you linked recently - Thanks.
Had a look around that traders site - if you register your email they have numerous other podcasts and a nice pdf transcript of an interview they done with Edward Thorp - (Beat The Dealer - BlackJack legend)

Regarding 'The Professor' Nick Mordin - as i am from the UK used to love reading his articles in the Racing Post Weekender/Irish Field etc and know his work very well but have been wondering for a while what happened to him - Has he gave up the Game , Retired , working abroad? I even enquired to The Racing Post but even that came up blank. I know he was on good terms with Cary Fotias - The Xtras (sadly passed - Blinkers Off - great book)

Can you shed any light as he is missed in the UK.

Thanks

I'm glad you liked the link to Dr. Z's interview. It is a great podcast.

Nick and I were friends for many years and attended the races in New York regularly. He moved from NYC to Saratoga and then moved to North Carolina. Then he started consulting for a racing syndicate. I believe he moved to somewhere in the Far East. He has lived in England, S. Africa, Malasia, the U.S. and now I think he is back in Asia. I think he has wanderlust and can't stay in one place for too long.

The last time I saw him was at Cary Fotias' funeral. We used to talk by phone, but he has not contacted me in several years. I tried calling him, emailing him, and messaging him on facebook, but he seems to have disappeared.

He has family in London, but I don't know their names. They might know where he is. He is an enigma. :)

PS

I could not find Beat the Racetrack on openlibrary.org.

betovernetcapper

08-12-2017, 01:51 PM

I have a mint condition of the Dr Z book, I'm willing to part with, but I'd suggest ATR PRO, with gives you the minimum & maximum payoffs by the minute. It also allows you to see what the show payoff would be if certain horses come in. It also has a host of other features & you can try it for a month for free. It's the best I've seen out there for US racing. If your betting in the UK, there are about a thousand programs making various claims but most are related to BetFair.
An interesting feature, particularly if your playing small tracks is that you can see how much your bet will affect the odds. At some small tracks, even a $20 bet can reduce your payoff significantly.

ARAZI91

08-12-2017, 02:26 PM

Is This it Highnote ? - Under 'Betting at The Racetrack'
Search under 'Racetrack'

Thanks for the info on Nick:ThmbUp:

highnote

08-12-2017, 03:07 PM

Is This it Highnote ? - Under 'Betting at The Racetrack'
Search under 'Racetrack'

Thanks for the info on Nick:ThmbUp:

Betting at the Racetrack is Dr. Z's book, too.

I searched for his first book "Beat The Racetrack", but it was not in openlibrary.org. Thanks for this!

ARAZI91

08-12-2017, 03:29 PM

Sorry Highnote , my bad - didn't know they were 2 seperate books.
Do you own 'Efficiency Of Racetrack Betting Markets' & 'Handbook Of Sports & Lottery Markets' ?- By Ziemba - Tons of interesting papers in both.

highnote

08-12-2017, 03:30 PM

davew

08-12-2017, 06:49 PM

some ADWs show percentage of pool for place and show, along with totals

AltonKelsey

08-12-2017, 07:37 PM

Yes. I have them both. Efficiency is a classic.

I've looked at a lot of these papers. Is there one thing in there that made you any money?

highnote

08-12-2017, 07:46 PM

I've looked at a lot of these papers. Is there one thing in there that made you any money?

Yes

AltonKelsey

08-12-2017, 11:23 PM

Yes

Nice to hear, I assume its not something as mundane as betting against bridge jumpers.

If not, why don't you share it.

steveb

08-13-2017, 12:13 AM

Nice to hear, I assume its not something as mundane as betting against bridge jumpers.

If not, why don't you share it.

may i ask what a bride jumper is?
thanks

AltonKelsey

08-13-2017, 12:24 AM

may i ask what a bride jumper is?
thanks

My guess, a groom that can't control himself.

The other thing, a bridge jumper, is well known here as someone that can't wait to lose all their money betting sure things to show.

highnote

08-13-2017, 12:25 AM

Nice to hear, I assume its not something as mundane as betting against bridge jumpers.

If not, why don't you share it.

It's pretty simple.

First, I am a decent computer programmer when it comes to horseracing. If you asked me to program for something other than horseracing I wouldn't be good at it because not much else that can be programmed is of interest to me.

I learned how to program the Dr. Z place and show system and automate the betting.

True story... I started with a bankroll of a couple thousand and it dwindled to $400 over a few months. I did some tweaks and in the next 6 months it ran up to over $25,000. I put that system in place 8-10 years ago. All I will say about it since then is that it does work, but not like in the book. You have to adopt it to the internet age. The math is the same, but the implementation is different.

By the way, I had been using the Dr. Z system successfully at the harness races and flats for several years before the internet. So I was confident that it would work. I used to be excited when I found 5 or 6 good bets in a day at the races. I was like a kid in a candy store when I started automatically betting the system at 25 or 30 tracks all day long on the internet!

Second, I learned everything I could about multinomial logistic regression. I have built many factor models. The hardest part is creating the predictive factors. You need a lot of them and they have to be good and they have to stay good over time, otherwise you have to replace them. So that means testing them occasionally and tweaking as necessary. It helps to have a statistical background, which I do. I also wrote code to automate the betting of the factor models, but I also had help with certain aspects of the code. Also, with the help of others I learned how to optimize the bets when there are more than one in a race -- especially in the case of exotics. This is where Generalized Reduced Gradient comes into play again.

I have done better in this multinomial logistic regression betting than in place and show, but how much better is not important. Suffice it to say, it works.

Benter and Woods have proved this.

I also learned a lot by reading the papers in "Efficiency". It's hard to attribute the knowledge I use to any one paper other than Benter's and Ziemba's, but I'd say the papers on market efficiency, favorite/longshot bias, optimal bet sizing, and others in the book gave me ideas and deeper understanding of horse race markets that helped me in many different ways -- such as getting an intuitive feel of how to create factors that would be predictive in a factor model.

I have been doing this for close to 30 years. I wouldn't recommend it to anyone. The stock market is easier and the market is a whole lot bigger. But racing is fun. So there's that aspect of it. Having fun is important.

highnote

08-13-2017, 12:47 AM

may i ask what a bride jumper is?
thanks

Let's say there is an extreme favorite in a race and there are only 4 or 5 horses in the race. Since the minimum payout to show is $2.10 per $2.00 wagered, sometimes the betting on the extreme favorite will command 95% or more of the show pool. This is also called a minus pool because racetracks actually pay out more than they take in when this happens.

The bridge jumper is the guy who bets perhaps $80,000 to show on the horse when there will be only a total of $100,000 in the pool. He is hoping to get a quick 5% return on his investment.

However, sometimes he loses all $80,000.

AltonKelsey

08-13-2017, 02:24 AM

That's great highnote (http://www.paceadvantage.com/forum/member.php?u=298), gives all us handicappers struggling with speed figs , stats , track bias, and physical handicapping hope for a better way.

I'm sure I speak for many of you.

highnote

08-13-2017, 09:03 AM

may i ask what a bride jumper is?
thanks

Steve,
I tried writing code to make speed figures using your method. But I could never figure out how to iterate to make a daily track variant. The figures work so-so. I refer to them when I handicap a race but other people make better ones.

Longshot6977

08-13-2017, 01:16 PM

may i ask what a bride jumper is?
thanks

Check these out:

https://www.etsy.com/market/bride_jumper

OR:

If you meant bridge jumper,
https://blog.winningponies.com/2011/09/07/bridge-jumping-not-for-the-faint-of-heart/

"Simply stated, a bridge jumper is a bettor who places an unusually large wager on a single horse, usually to show. While these bets vary in size it’s not unusual to see them approach $100,000 or more. Because show pools have a mandatory payout of $2.10 or $2.20 (depending on racing jurisdiction) no matter how low the odds actually are, if a bridge jumper is successful, he’ll get a small return on his huge bet no matter what the true odds are. This creates something that tracks refer to as a “minus pool” because more is paid out to the bridge jumper than the track receives via the takeout. Of course if the bridge jumper’s horse finishes out of the money, this will result in huge show payoffs for the other horses in the field, and the bridge jumper may just feel like hopping off the nearest bridge to the racetrack."

highnote

08-13-2017, 02:07 PM

The thing about bridge jumpers is that they might not actually be bridge jumpers. Read Ziemba's paper called "Locks at the Racetrack".

When there is a large amount of money bet on one horse to show it is possible to construct a bet where you cannot lose if you bet on all the horses in the proper proportion.

So even though the big favorite finishes out of the money the supposed bridge jumpers could still turn a profit.

steveb

08-13-2017, 07:26 PM

Steve,
I tried writing code to make speed figures using your method. But I could never figure out how to iterate to make a daily track variant. The figures work so-so. I refer to them when I handicap a race but other people make better ones.

thanks firstly for the brid(g)e! jumper explanation.

what can i say about speed figures?

it's quite possible that others could make them better i guess.

it's also quite possible that you have not coded it properly, or, understood properly what i have said, or even if i have said how to make them.

in fact, i am pretty sure i haven't said how to make them, other than putting up a screenshot of how the woods syndicate were trying to emulate my way of doing it.
if i had put it up here, then there would have been no need to privately show others how to do it, would there?

even if you did it with the exact same method as i would, it would also depend on how YOU classified race types.

i think we have some wires crossed somewhere.

steveb

08-13-2017, 07:33 PM

Check these out:

https://www.etsy.com/market/bride_jumper

OR:

If you meant bridge jumper,
https://blog.winningponies.com/2011/09/07/bridge-jumping-not-for-the-faint-of-heart/

"Simply stated, a bridge jumper is a bettor who places an unusually large wager on a single horse, usually to show. While these bets vary in size it’s not unusual to see them approach $100,000 or more. Because show pools have a mandatory payout of $2.10 or $2.20 (depending on racing jurisdiction) no matter how low the odds actually are, if a bridge jumper is successful, he’ll get a small return on his huge bet no matter what the true odds are. This creates something that tracks refer to as a “minus pool” because more is paid out to the bridge jumper than the track receives via the takeout. Of course if the bridge jumper’s horse finishes out of the money, this will result in huge show payoffs for the other horses in the field, and the bridge jumper may just feel like hopping off the nearest bridge to the racetrack."

thanks, and yes i meant bridge!!

just one question, does the operator of the show pool take the hit, or does the dividend on the other horses that show pay less to compensate.

in australia's case they're robbing peter to pay paul.
as in australia, it is the latter, or was, as i am out of it these days, so may be left behind as far as changes go.

AltonKelsey

08-13-2017, 09:49 PM

thanks firstly for the brid(g)e! jumper explanation.

i think we have some wires crossed somewhere.

Wires crossed is a feature here

highnote

08-13-2017, 10:12 PM

thanks, and yes i meant bridge!!

just one question, does the operator of the show pool take the hit, or does the dividend on the other horses that show pay less to compensate.

in australia's case they're robbing peter to pay paul.
as in australia, it is the latter, or was, as i am out of it these days, so may be left behind as far as changes go.

This is my understanding, although, I may not be correct about all of it, but, in general, I think I am:

Bettors who receive a rebate on their bets and who make their bets over the internet pay their share of the minus pool they helped create.

Bettors who bet at the track and who may or may not be primarily responsible for the minus pool do NOT have to pay for the minus pool. The tracks take the hit. That is, on-track bettors receive the full $2.10 or $2.20 payoff.

The minimum payoff is $2.10 or $2.20 to show on a $2 wager depending on the jurisdiction. If the actual payoff is less than $2.10, the track, by law, still has to pay out $2.10 or $2.20 on every wager.

Now, advance deposit wagering companies that might pay a 5% or even 10% rebate on the bet sometimes get charged by the host racetrack for its share of the minus pool. The ADWs in turn pass on a share of the charge to its customers who made the bet.

I gathered a lot of data on this as a study for a group that asked me to see what happened if a bet was made on races with a minus pool. Not every bet that would have been made into a minus pool was on the horse that caused the minus pool. Sometimes the bets won a lot of money when they were NOT on the horse that caused the minus pool and when the horse that was taking all the money finished out of the money. However, this was offset by the occasional losses of the bets on those horses that produced the big payoffs and also losses on the favorite that ran out of the money, or the losses on bets on the favorite when it was in-the-money because of the share of the minus pool that had to be paid back. I looked at all the possible scenarios, tallied the results, and found that show betting into minus pools broke even for the system that was being used to identify the bets. On the one hand the bets broke even for the bettor. I can't remember if the ADW still made a profit from the minus pool bets, even though share of the minus pool loss was paid back to them.

Keep in mind that it might be the case that rules vary from one ADW to another. Host racetracks might also have varied rules.

As the esteemed Senator Bird of the Great State of West Virginia once said, "The secret to success in the senate is knowing the rules." The same is likely true in racetrack betting.

highnote

08-13-2017, 10:20 PM

i think we have some wires crossed somewhere.

I am certain I do not make my figures the way you suggested. We had some pretty lengthy posts back and forth and you corrected some of my errors for which I thank you.

You pointed out that Beyer figures are just percentages whether he knows it or not. As I dug deeper into the math behind the making of the figures you were proved correct.

Surely, the mathematician Bill Quirin understood Beyer figures, but even so, Quirin's method was different from Beyers' and I am not convinced his figures are less predictive than Beyers'. Quirin's might be more profitable if they are reasonably good and used by fewer people.

This could be studied, but it's hard to find time for everything. At some point, you realized there is more to life than chasing money and now, I seem to recall, you chase birds -- the winged kind.

Whether our wires are crossed or not, I learned a lot from your posts. And that's why I spend time on PaceAdvantage. The little bits of wisdom I pick up here and there and the stimulating, lively, and passionate discussions keep me coming back. :)

And maybe most important, is the chance for me to give back some of what I've learned and help others. Hopefully, I don't lead them too far astray! LOL

The only downside is that sometimes you have to tolerate the occasional post from a sourpuss or a troll. But it's easy enough to ignore those types, too.

steveb

08-14-2017, 02:05 AM

You pointed out that Beyer figures are just percentages whether he knows it or not. As I dug deeper into the math behind the making of the figures you were proved correct.

they're only percentages where the winners are concerned.
not the beaten runners.

the attached will show how to get the variant and the basic speeds for the races of a meeting in its most basic way.
it is obvious that the 2000 metre race is a long way astray(and the 1000M one to a lesser extent), but the parts that will account for that are not included.
but it will likely be slow pace accounting for it, or something else just as easily discoverable.
but it has no place in this very basic example.
unless you know how the standards are calculated and the class(expected speed) columns are figured, then you can't accurately emulate it.
because how they were found is part of the entire method.
anyway, because that race is so far astray, the basic way will make all the other races wrong too, because that bad one (wrongly) has the same weighting as any other of the races.

in the attached the variant is simply the average of the difference between expected and actual.
and it is very obviously wrong in this example, but i was just displayng the basics.
but when you know how to weight the different races, rather than just using the plain average, it works very well.
and your variant will be accurate too, but i still cant see any use for the variant other than giving you an indication of how fast or slow the track was running for that meeting.

and the smaller your deviation between actual and expected, then the more accurate it will normally be, although a number far too high for the expected, might just be a horse that is on the way to better things.

highnote

08-14-2017, 03:58 AM

they're only percentages where the winners are concerned.
not the beaten runners.

the attached will show how to get the variant and the basic speeds for the races of a meeting in its most basic way.
it is obvious that the 2000 metre race is a long way astray(and the 1000M one to a lesser extent), but the parts that will account for that are not included.
but it will likely be slow pace accounting for it, or something else just as easily discoverable.
but it has no place in this very basic example.
unless you know how the standards are calculated and the class(expected speed) columns are figured, then you can't accurately emulate it.
because how they were found is part of the entire method.
anyway, because that race is so far astray, the basic way will make all the other races wrong too, because that bad one (wrongly) has the same weighting as any other of the races.

in the attached the variant is simply the average of the difference between expected and actual.
and it is very obviously wrong in this example, but i was just displayng the basics.
but when you know how to weight the different races, rather than just using the plain average, it works very well.
and your variant will be accurate too, but i still cant see any use for the variant other than giving you an indication of how fast or slow the track was running for that meeting.

and the smaller your deviation between actual and expected, then the more accurate it will normally be, although a number far too high for the expected, might just be a horse that is on the way to better things.

That's great. I like the way you laid it out in the spread sheet. It makes a lot of sense.

It's been so long since I have made speed figures using what I thought was your method that I forgot to mention that making good standard times is also difficult.

I won't go into my whole method because it would take to long to explain each step. Plus, I don't want to waste your time. You've been very helpful.

But in a nutshell, I converted times to speed figures and then kept adjusting the standard times until they equaled the class average speed figure, or something like that.

When I tested them in a model they were significant. That is, they were more predictive than random. I doubt they are good enough by themselves to turn a profit from betting them, but at least it is better to have them in my model than not.

steveb

08-14-2017, 04:37 AM

That's great. I like the way you laid it out in the spread sheet. It makes a lot of sense.

It's been so long since I have made speed figures using what I thought was your method that I forgot to mention that making good standard times is also difficult.

I won't go into my whole method because it would take to long to explain each step. Plus, I don't want to waste your time. You've been very helpful.

But in a nutshell, I converted times to speed figures and then kept adjusting the standard times until they equaled the class average speed figure, or something like that.

When I tested them in a model they were significant. That is, they were more predictive than random. I doubt they are good enough by themselves to turn a profit from betting them, but at least it is better to have them in my model than not.

that is basically right, but you have to start with EVERY type of race getting the same expected figure , which for me is a 100.
then you iterate, and check...then you alter the times AND the race types, not just the times.
if you can't classify the races sensibly then it's going to be hard.
in those instances it's solved(classifying) using the official ratings by regression analyiss.

i know how good they WERE, because the syndicates had already told me that what i was doing for them, was more accurate than what they themselves could do.
thus what i put here a long time back was their attempt at emulating me.
I only put it up here because they pissed me off.
when alan died i was owed a lot of money, and ongoing profit share, but because our agreement was vocal only, his daughter would not do the right and proper thing.

and it is quick in that you can find classes and times in about 30 seconds depending on your computer i guess.
they could not get as accurate, depsite alan's son and others, spending months on the new country.
which is why they sought my hlep in the first place.

anyway, if anybody can win(enough to live) off time alone, then i would be very surprised.
certainly not me, and certainly none of the teams that i knew stuff about.

highnote

08-14-2017, 05:24 AM

that is basically right, but you have to start with EVERY type of race getting the same expected figure , which for me is a 100.
then you iterate, and check...then you alter the times AND the race types, not just the times.

I did that, but I am sure I did not do it as well as you. I can send you some code examples by private msg when I get a chance.

if you can't classify the races sensibly then it's going to be hard.
in those instances it's solved(classifying) using the official ratings by regression analyiss.

If I get motivated maybe I will try that. I just made figures for the most popular classes. I'll just avoid betting on the unpopular classes. I just guessed at the standard times when my sample sizes were small.

i know how good they WERE, because the syndicates had already told me that what i was doing for them, was more accurate than what they themselves could do.
thus what i put here a long time back was their attempt at emulating me.
I only put it up here because they pissed me off.
when alan died i was owed a lot of money, and ongoing profit share, but because our agreement was vocal only, his daughter would not do the right and proper thing.

It seems like they must have gone underground. I haven't heard anything about them in several years.

anyway, if anybody can win(enough to live) off time alone, then i would be very surprised.
certainly not me, and certainly none of the teams that i knew stuff about.

Those days were probably over not long after Beyer wrote his book.

Thanks again!

steveb

08-14-2017, 06:09 AM

It seems like they must have gone underground. I haven't heard anything about them in several years.

when alan died, his 2ic (paul l) tried to wrestle control of the syndicate.
i don't know exactly what happened, just that there was a split, where alan's daughter victoria got control, and the paul bloke and others left and started their own syndicate.

i know that woods and associates as it became known, was still going when i decided racing was no longer for me, just a few years back.
that was where your friend would have been, as i have seen correspondence between nick m and a woman called ciara who was there at the time(still?).
don't ask me how i get to see these things.
if they are still going now, i don't know, but i do know that the pari-mutuel handle in australia is in freefall, and they had bet there(among other places) since 2006 so maybe they have left that market.

i don't know about the Paul L one.

another i was involved with, is no more, and that was full of other ex woods people.

i guess it had to happen sooner or later, but i am pretty sure something is happening in malta at the moment.
as many names i know from the old days, appear to be domiciled there.
so i am guessing they are still involved in gambling, but at what i don't know.

highnote

08-14-2017, 11:29 AM

when alan died, his 2ic (paul l) tried to wrestle control of the syndicate.
i don't know exactly what happened, just that there was a split, where alan's daughter victoria got control, and the paul bloke and others left and started their own syndicate.

i know that woods and associates as it became known, was still going when i decided racing was no longer for me, just a few years back.
that was where your friend would have been, as i have seen correspondence between nick m and a woman called ciara who was there at the time(still?).
don't ask me how i get to see these things.
if they are still going now, i don't know, but i do know that the pari-mutuel handle in australia is in freefall, and they had bet there(among other places) since 2006 so maybe they have left that market.

i don't know about the Paul L one.

another i was involved with, is no more, and that was full of other ex woods people.

i guess it had to happen sooner or later, but i am pretty sure something is happening in malta at the moment.
as many names i know from the old days, appear to be domiciled there.
so i am guessing they are still involved in gambling, but at what i don't know.

That's interesting. Someone just posted recently asking me what happened to Nick M. I have tried contacting, but he seems to have disappeared. He hasn't posted to facebook in several years. A company bought his domain name when it wasn't renewed by him.

He has lived all over the world, so he could be anywhere. That is his nature.

I wish I could talk to him. We had a lot of fun going to the races together. He taught me a lot about paddock handicapping.