The Gambler Who Cracked the Horse-Racing Code

[jay68802]

I look at this way. It is sort of like what I have done in real estate. I started out by owning 1/2 of a duplex. It took a long time to be able to turn that little investment into a business. I was a small operator because of limited $. The computer teams start with 10 duplexes. A big advantage. It took about 40 years for me to get where I am now. If I would have started with 10 duplex's I would own 15 times the amount of property that I have now. Right now I do not even have what the computer teams started with. And I was lucky, in that I avoided some downturns in the economy. Tough game to grow in, but it can be done with a good plan and some luck.

[Dave Schwartz]

Originally Posted by classhandicapper

I love it.

I'm making a huge investment of time to really tear this game apart and build personal metrics, but it's a very slow process. Even when I get something up and running well, I immediately come up with ways to make it better. In the end though, I still see the "exceptions" and know there are limits on what I can "program". That why I still feel that no matter how much the whales dominate the pools in general, they are going to keep leaving holes to exploit.

Bright Shiny Objects - The biggest stumbling block of all.

[jay68802]

Keep it simple, stupid. The approach I am developing now. no more shiny objects for me.

[Hoofless_Wonder]

Originally Posted by dilanesp

Forty Niner being sent in the Preakness

Although Pat Day gave many horses horrible rides, if memory serves he was riding to orders when Woody Stephens determined that Winning Colors wasn't going to get an easy lead. With comments in the news articles in the DRF leading up to the 1988 Preakness, it was no surprise what happened.

Jockey tendencies are a bit of a wild card, but still can be useful. Some are better at the gate, some have a better sense of pace, some know how to get a better trip on turf, etc.

Day was a nightmare for me, and I was really pleased to see him retire. His horses were so often underlays at OP or AP, and only when he went to the big circuits could you get any value. And then it was a crap shoot whether he'd give the horse a decent ride. He wasn't as good a judge of pace as some of his peers, and more of a finesse rider. He seemed to excel at getting horses to relax and getting good trips when he had the best horse. On paper, "Day on" was always good for a five length improvement. He gave Theatrical some awesome rides - not so much with Easy Goer.

And he must have had a great agent, since he got so many live mounts. He was also a fine gentleman (and probably still is), and a great ambassador for the sport. I also have to give him a tip of the hat for beating his personal demons and drug problems. A lot of jockeys couldn't do that.

[dilanesp]

Originally Posted by Hoofless_Wonder

Although Pat Day gave many horses horrible rides, if memory serves he was riding to orders when Woody Stephens determined that Winning Colors wasn't going to get an easy lead. With comments in the news articles in the DRF leading up to the 1988 Preakness, it was no surprise what happened.

Jockey tendencies are a bit of a wild card, but still can be useful. Some are better at the gate, some have a better sense of pace, some know how to get a better trip on turf, etc.

Day was a nightmare for me, and I was really pleased to see him retire. His horses were so often underlays at OP or AP, and only when he went to the big circuits could you get any value. And then it was a crap shoot whether he'd give the horse a decent ride. He wasn't as good a judge of pace as some of his peers, and more of a finesse rider. He seemed to excel at getting horses to relax and getting good trips when he had the best horse. On paper, "Day on" was always good for a five length improvement. He gave Theatrical some awesome rides - not so much with Easy Goer.

And he must have had a great agent, since he got so many live mounts. He was also a fine gentleman (and probably still is), and a great ambassador for the sport. I also have to give him a tip of the hat for beating his personal demons and drug problems. A lot of jockeys couldn't do that.

I will never forget the ride he gave me on Unbridled in the BC Classic. Post 14, start on the turn, and Day managed to save all the ground.

[cato]

A question for Dave S and/or anyone else who has knowledge of da whales -

Based primary on the giant profits reported with regard to Bentner and Woods for betting in Hong Kong it seems that they were making a decent positive ROI and not simply relying on rebates; whereas most of the stuff i've read about the US whales is that they mainly rely on their rebates to get them in the profit zone.

Is that correct and if so why are they more profitable in HK than the US?
The obvious reasons (to me) include -

a more honest game in HK, including
better regulation of (or no) drugs
published post race interviews with jockeys and trainers
bigger fields
fewer races
larger pools (harder to move the odds)
controlled and closed pool of horses that race on a regular basis
more additional information, including vet treatments, etc.

What else explains it, assuming the premise is true? Part of what started me on this path was a discussion in the article about Bentner, a relative necomer to computer programming and statisitcs (although admittedly very smart) who was able to put together a computer program that was profitable almost from the beginning (although Bentner did have some extra time to work on it while he was trying to win and raise enough money to get back to HK). Additionally it sounds like the first algrythm considered approximately 20 data points. That does not seem overly demanidng on the software programming side of things - heck

I'll bet all of the programs sold in the US have that many or more points to analyze and sythesize but none of them have aproached bentner's success (at least that i am aware of)

Best, Frank

[BCOURTNEY]

Originally Posted by cato

A question for Dave S and/or anyone else who has knowledge of da whales -

Based primary on the giant profits reported with regard to Bentner and Woods for betting in Hong Kong it seems that they were making a decent positive ROI and not simply relying on rebates; whereas most of the stuff i've read about the US whales is that they mainly rely on their rebates to get them in the profit zone.

Is that correct and if so why are they more profitable in HK than the US?
The obvious reasons (to me) include -

a more honest game in HK, including
better regulation of (or no) drugs
published post race interviews with jockeys and trainers
bigger fields
fewer races
larger pools (harder to move the odds)
controlled and closed pool of horses that race on a regular basis
more additional information, including vet treatments, etc.

What else explains it, assuming the premise is true? Part of what started me on this path was a discussion in the article about Bentner, a relative necomer to computer programming and statisitcs (although admittedly very smart) who was able to put together a computer program that was profitable almost from the beginning (although Bentner did have some extra time to work on it while he was trying to win and raise enough money to get back to HK). Additionally it sounds like the first algrythm considered approximately 20 data points. That does not seem overly demanidng on the software programming side of things - heck

I'll bet all of the programs sold in the US have that many or more points to analyze and sythesize but none of them have aproached bentner's success (at least that i am aware of)

Best, Frank

Benter and Woods did as well as they did because they went through all the maturity stages of professional handicapping and they did not stop after Stage 1.

Stage 1: Horse selection strategy and approach - Possibilities
Stage 2: Bet optimization strategy and approach - Probabilities
Stage 3: Wagering strategy and approach (using 1 and 2) - Profitability

Most individual players cannot have a reasonable conversation past Stage 1. Benter built a multinomial logistic regression model and learned how to exploit it to correctly size wagers from their selection process, then further added the wagering strategy (a modified kelly approach) based on positive expectations from the runners to construct the tickets. This generated a profit, it wasn't the selection process alone. Stages 2 and 3 require sophisticated mathematical skills and software. This is the difference between the minnows and the sharks.

You can just simply focus on making your selection process unique from the selection process of the sharks, such that you don't land on the syndicate horses. This increases an individual players chance at returning a profit versus the syndicated teams. Players should at a minimum dabble in Stage 2 and Stage 3 if possible. You cannot refine and re-refine Stage 1 Ad nauseam into a profitable approach, although this is the curse of 97% of all horse players, the hamster wheel of eternity.

[PaceAdvantage]

It also helped IMMENSELY that Benter & Woods were one of the FIRST to enter the market...they didn't have much if any competition doing what they were doing.

Now that a lot of others have caught on, I'm guessing it's not as profitable as it was initially...even in Hong Kong.

[Dave Schwartz]

Is that correct and if so why ARE they more profitable in HK than the US?
The obvious reasons (to me) include -

a more honest game in HK, including
better regulation of (or no) drugs
published post race interviews with jockeys and trainers
bigger fields
fewer races
larger pools (harder to move the odds)
controlled and closed pool of horses that race on a regular basis
more additional information, including vet treatments, etc.

First, the word should be WERE, not ARE. Those guys have been gone from HK for a long time.

The story I have, from two guys who both worked for Benter in the 1980s, I hear that he was making around 6%. As BCourtney said, they had very little competition.

BTW, when you meet people in Asia or Australia involved in racing "the Benter way," it seems that they all worked for Benter at one time or another. This is what originally caused the competition to form: smart guys who knew it could be done either going out on their own or finding an angel investor.

Side story...

Every time I'd get a call from some new group who was starting up, looking for experienced horse racing developers, the first question they'd ask was, "Do you program in FoxPro?"

FoxPro is a long-nearly-dead d-Base clone that is still hanging on by a thread. Perhaps it has even made a little progress in recent years.

This is a trickle-down result because Benter used FoxPro.

To hear the Asian programmers talk about it, they all stole a copy of Benter's program when they left. (I actually had a copy myself at one time, but it was so simple that I just did not see the value.)

Benter's software was very high on things like time-decayed variables, especially things like finish position and class levels.

Remember that HK is an enclosed horse community (i.e. no shippers except for invitational races). This makes for a very distinctly class-sensitive system, with weight imposts being a huge part of the classes.

That is, win a race, go up in class or carry significantly more weight at this level. Much like Greyhound racing in the US.

Back then, Benter also "gathered his own data."

Back to meaningful stuff.

The story that I heard was that the new competitors were cutting into Benter's advantage. Remember that it does not take new winners using the same/similar approach to take your ROI down. It just demands an increase in people who wind up on the same horses for whatever reason.

The ROI dropped from +6, to +4, then +2. After that, if I recall correctly, things began to unravel much faster. Benter being a really bright guy, decided to hang it up when it got to +1%.

As for these questions...

a more honest game in HK, including
better regulation of (or no) drugs
published post race interviews with jockeys and trainers
bigger fields
fewer races
larger pools (harder to move the odds)
controlled and closed pool of horses that race on a regular basis
more additional information, including vet treatments, etc.

All the above probably contributed to the Perfect Storm created by Benter.

No real idea about the "honesty."

Bigger fields certainly helped, but with the favorites winning in the 40s (even with large fields) there certainly was a high level of "best horse wins."

When I was working in Japan, my client, who was the money handler for a large player - "whale" by our standards and terminology. (His job was to make deposits or withdrawals from each track agent's betting money; everything was cash.)

He was also known to make large wagers of his own on some of the same races he bet for his boss. (This was an automatic firing offense.)

I asked him, "What is the largest bet you ever made on a horse race with your own money?"

Very sheepishly, he said, "Not that much." When I pressed him, he continued with his embarrassed attitude and finally, took out his calculator to figure the exchange rate.

$20,000 USD is what he said. I was pretty impressed because I can tell you that I've never made a single bet anywhere near that. My largest bet ever was $2,400 spread over 3 horses (and my hands were shaking).

I asked if the horse won. He said it did. "What did it pay?"

"$117 was his response."

WHAT!

Well, that was for a $5 unit, but still, that is the equivalent to $46.80!

So, yes, pool size matters.


Dave

PS: BTW, when we speak of "Asian Racing," remember that the people involved are ALL westerners. Only once did I ever meet a team employee that was not round-eyed and he was a translator.

[Parkview_Pirate]

Originally Posted by dilanesp

I will never forget the ride he gave me on Unbridled in the BC Classic. Post 14, start on the turn, and Day managed to save all the ground.

Yes, that was Day at his finest. It helped to have the best horse too, which Unbridled was on Derby Day as he rolled by Summer Squall (and Day) with ease.

In the BC Classic I ended up making a small bet on Ibn Bay, but did have a saver Q with Unbridled, but thought his outside post (and Day) would result in a tough trip. My mistake. Great call by Durkin ("Ibn Bay is STILL there!") and a great race.

*******

Dave, excellent response on HK racing and the side story. It's definitely a closed system, and highly regulated. Considering their purse structure vs. handle, they must funnel a lot of money towards the plant, supervision and the like (although HK is an expensive place to live).

For some reason, it sticks in the back of my mind that takeout may have gone up there too over the years, which of course would affect everyone's returns. But I'm not sure about that.

[BCOURTNEY]

Originally Posted by PaceAdvantage

It also helped IMMENSELY that Benter & Woods were one of the FIRST to enter the market...they didn't have much if any competition doing what they were doing.

Now that a lot of others have caught on, I'm guessing it's not as profitable as it was initially...even in Hong Kong.

Yes, when you fly around in a little Cessna plane in the US from track to track hauling in a computer the size of a briefcase to place bets until the pools can't sustain your wagers any longer and you have to head to the gambling capital of the world being first is a good thing.

[BCOURTNEY]

Originally Posted by Dave Schwartz

Side story...
[INDENT]Every time I'd get a call from some new group who was starting up, looking for experienced horse racing developers, the first question they'd ask was, "Do you program in FoxPro?"

FoxPro is a long-nearly-dead d-Base clone that is still hanging on by a thread. Perhaps it has even made a little progress in recent years.

This is a trickle-down result because Benter used FoxPro.

Hi Dave, on the historical piece, I believe originally they had obscure pick like operating system then moved to Revelation (a database which is a type of post-relational database known as a multi-value database) Then later outgrowths (from people that used to work there from the HK teams) were rewritten in Foxpro after converting all the code and databases, converting from Revelation to Foxpro or DBase was quite common in that period of time.

Does that sound right or conflict with your understanding? Maybe we should query Benter himself, fun history here.

[dilanesp]

Originally Posted by PaceAdvantage

It also helped IMMENSELY that Benter & Woods were one of the FIRST to enter the market...they didn't have much if any competition doing what they were doing.

Now that a lot of others have caught on, I'm guessing it's not as profitable as it was initially...even in Hong Kong.

+1

Indeed, this is true about just about every successful investment or gambling or business strategy.

Success attracts others who do the same thing, which then cuts the profitability over time.

[dilanesp]

Originally Posted by Parkview_Pirate

Yes, that was Day at his finest. It helped to have the best horse too, which Unbridled was on Derby Day as he rolled by Summer Squall (and Day) with ease.

In the BC Classic I ended up making a small bet on Ibn Bay, but did have a saver Q with Unbridled, but thought his outside post (and Day) would result in a tough trip. My mistake. Great call by Durkin ("Ibn Bay is STILL there!") and a great race.

*******

Dave, excellent response on HK racing and the side story. It's definitely a closed system, and highly regulated. Considering their purse structure vs. handle, they must funnel a lot of money towards the plant, supervision and the like (although HK is an expensive place to live).

For some reason, it sticks in the back of my mind that takeout may have gone up there too over the years, which of course would affect everyone's returns. But I'm not sure about that.

Ibn Bay was a great pick and you really deserved to win that (and would have if Day hadn't given his horse that amazing trip).

[Dave Schwartz]

Originally Posted by BCOURTNEY

Hi Dave, on the historical piece, I believe originally they had obscure pick like operating system then moved to Revelation (a database which is a type of post-relational database known as a multi-value database) Then later outgrowths (from people that used to work there from the HK teams) were rewritten in Foxpro after converting all the code and databases, converting from Revelation to Foxpro or DBase was quite common in that period of time.

Does that sound right or conflict with your understanding? Maybe we should query Benter himself, fun history here.

Sounds right to me. When I got to it, everything was pretty much FoxPro. (That was 2002.)