Computer Team Kentucky Derby

[arkansasman]

A good friend of mine, who is a very knowledgeable horseplayer and knows some horse race computer modelers, thinks that a huge bet made on Sweetnorthernsaint in the last 2 minutes was made by a computer team. If the computer team's model had Sweetnorthernsaint as that big of an ADVANTAGE bet, then it is certain that most of the team's ADVANTAGE exacta bets and trifecta bets were keyed with Sweetnorthernsaint on top.

Does anyone have any insight or thoughts into this wager on Sweetnorthernsaint?

This reminds me of the huge bet made on Monarchos in the 2001 Florida Derby. The difference is - the North Dakota better cashed on the 2001 Florida Derby.

[sjk]

Hi Jack,

It doesn't seem like a very plausible story to me. Even with an advantaged program you are still going to lose the majority of your plays. Any play that you would worry about waiting until the last minute before the Derby and have a noticable effect on the prices would be so far out of proportion to other bets you would otherwise be able to make that it is much more gamble than investment.

It would take an awful lot of advantaged play at RD and Pha and other tracks to make back that kind of loss.

[LaughAndBeMerry]

I wish I could find the article from Saturday night in the DRF. The quote was "...a very large bet on Sweetnorthernsaint at post time brought his odds down from 7-1 to 5.50 to 1 and made him the slight favorite over Barbaro."

It was in like the third or fourth paragraph. All it was that single comment with no other information. I noticed the horse getting pounded in the last 30 seconds. If the DRF was right and it was a single bet, it almost HAD to be a computer team. No gambler is going to go to the post as they're loading to make a $500,000+ bet.

LBM

[BillW]

Late money came in on the #3 Keyed Entry also taking the odds from 35-1 down to 28-1 in the last tick. Interesting movement at posttime, no doubt.

[arkansasman]

The part you are talking about is about half way down the page on this link.

http://www.drf.com/drfNewsArticle.do?NID=74359&subs=0&arc=1

[stlseeeek]

didnt you guys think the exacta was a little light???


i thought it would pay 1k easy. the tri payed nicely, but exacta seems light

[LaughAndBeMerry]

Jack:

This is the second article. I think it appeared today. A racing buddy of mine clued me into it. The first article appeared Saturday night about 3-4 hrs after the Derby was over. That was the article I quoted.

LBM

[highnote]

A good friend of mine, who is a very knowledgeable horseplayer and knows some horse race computer modelers, thinks that a huge bet made on Sweetnorthernsaint in the last 2 minutes was made by a computer team. If the computer team's model had Sweetnorthernsaint as that big of an ADVANTAGE bet, then it is certain that most of the team's ADVANTAGE exacta bets and trifecta bets were keyed with Sweetnorthernsaint on top.

Does anyone have any insight or thoughts into this wager on Sweetnorthernsaint?

This reminds me of the huge bet made on Monarchos in the 2001 Florida Derby. The difference is - the North Dakota better cashed on the 2001 Florida Derby.

Maybe it was the same guy who bet Monarchos?

I don't think it would be any benefit to me to know who they (him/her/it?) is, but it would be interesting to know.

Do you think there is any benefit to knowing? Or are you just curious like me?

I was at CD on Saturday and made some bets. I bet the same whether there are teams or not. I love betting against computer teams. When they make mistakes the payoffs are juicier. I had a modest bet on Barbaro, so they helped me. They probably make win bets on the wrong horse 67% of the time. All I have to do is bet on factors they don't use.

If they bet Saint on Saturday it is pretty easy to tell what factor they have heavily weighted -- Speed Figures. All you have to do is give less weight to Speed Figures or else find a better way to make them.

If they bet speed figures in the Derby then they need to adjust their model for the Derby. It ain't just about speed -- it's also about stamina.

The only thing that Minister's 116 Beyer and Saint's 109 Beyer meant to me was that they'd never earn that high of a figure in the Derby.

I wonder if they bet whole Kelly or fractional Kelly? When it comes to the Derby, I think I'd bet about 1/10 Kelly or else make a Derby model. It is nothing like a typical $25,000 claiming race for older horses at 6 furlongs.

Live and learn. If they can afford to bet $500,000 then they can afford to lose $500,000. So what do you think -- they must have a bankroll of between $10 and $50 million?

[PaceAdvantage]

To all in this thread:


What exactly is your definition of a "Computer Team?"

[arkansasman]

Here is another link from a Privman article referring to a huge wager.

http://www.drf.com/drfNewsArticle.do?NID=74316&subs=0&arc=1

[arkansasman]

Maybe it was the same guy who bet Monarchos?

I don't think it would be any benefit to me to know who they (him/her/it?) is, but it would be interesting to know.

Do you think there is any benefit to knowing? Or are you just curious like me?

I was at CD on Saturday and made some bets. I bet the same whether there are teams or not. I love betting against computer teams. When they make mistakes the payoffs are juicier. I had a modest bet on Barbaro, so they helped me. They probably make win bets on the wrong horse 67% of the time. All I have to do is bet on factors they don't use.

If they bet Saint on Saturday it is pretty easy to tell what factor they have heavily weighted -- Speed Figures. All you have to do is give less weight to Speed Figures or else find a better way to make them.

If they bet speed figures in the Derby then they need to adjust their model for the Derby. It ain't just about speed -- it's also about stamina.

The only thing that Minister's 116 Beyer and Saint's 109 Beyer meant to me was that they'd never earn that high of a figure in the Derby.

I wonder if they bet whole Kelly or fractional Kelly? When it comes to the Derby, I think I'd bet about 1/10 Kelly or else make a Derby model. It is nothing like a typical $25,000 claiming race for older horses at 6 furlongs.

Live and learn. If they can afford to bet $500,000 then they can afford to lose $500,000. So what do you think -- they must have a bankroll of between $10 and $50 million?

John

There is no benefit for me to know either. I was just curious too.

I would say that their bankroll is at least 50 million. They use a form of Kelly, but it is what a lot of people refer to as the "whale formula." This type of formula is pool size limited.

[LaughAndBeMerry]

Jack:

The Privman article was the one I was referring to. I'm glad you found it. I was beginning to think I was crazy. I didn't see anyone else mention it any follow-up article or post I saw on any board.

Re computer teams: We're referring to the large batch betting syndicates making high volume and high dollar wagers at post-time, a la Bentner's HK team or Peter Wagner ("the North Dakota Whale") in the U.S.

[highnote]

Jack:

The Privman article was the one I was referring to. I'm glad you found it. I was beginning to think I was crazy. I didn't see anyone else mention it any follow-up article or post I saw on any board.

Re computer teams: We're referring to the large batch betting syndicates making high volume and high dollar wagers at post-time, a la Bentner's HK team or Peter Wagner ("the North Dakota Whale") in the U.S.

Where can someone make bets in batches?
Do you mean like making a thousand bets in a second or making one bet every second or so?

[LaughAndBeMerry]

Yes. Computer teams have sophisticated investment software than crawls through the pools identifying overlays, structures hundreds or thousands of bets, and then transmits them all at once with a minute or less to post by virture of a direct link to a wagering hub like Las Vegas Dissemination Corp (LVDC).

LBM

[Valuist]

I'm glad to hear it. It tells me these computer whales are betting big final figures without a full understanding of how the horse earned the figure. They also clearly must've never watched the Illinois Derby because the video clearly shows the horse was running near a very slow pace.

I didn't cash on the Derby but I was right about one thing: SNS was going to be heavily bet. I took some heat on here back in mid-April for saying he was going to be one of the favorites and I do believe my assertion was correct.

[arkansasman]

I'm glad to hear it. It tells me these computer whales are betting big final figures without a full understanding of how the horse earned the figure.
They use a lot more factors than just speed figures. Some use up to 80 factors to create their probabilities. They use factors like morning line odds coverted to probabilities, weight, post, recency weighted factors (although Benter loves these, I do not) that includes finish, weight, etc.

The models that they build to arrive at probabilities is a very complicated process and I speak from experience.

Interesting thought - if the huge bet was loaded into the tote at 1 minute to post - did that create an advantage bet on Barbaro?

[highnote]

if the huge bet was loaded into the tote at 1 minute to post - did that create an advantage bet on Barbaro?

I guess it would depend on the model. Maybe the lower odds on SNS created by the big bet actually made SNS an even better bet to some other model.

I suppose the possibilities might be limitless.

[Valuist]

They use a lot more factors than just speed figures. Some use up to 80 factors to create their probabilities. They use factors like morning line odds coverted to probabilities, weight, post, recency weighted factors (although Benter loves these, I do not) that includes finish, weight, etc.



Well something is definitely askew in their formulas. I'll take pencil and paper any day of the week over the MSFT/ORCL crowd.

[PaceAdvantage]

Well something is definitely askew in their formulas. I'll take pencil and paper any day of the week over the MSFT/ORCL crowd.

And you can come to this conclusion off of one race? The 20-horse chaotic Kentucky Derby no less?

I'm a member of the Microsoft crowd, and I had Barbaro on top. What now?

[Valuist]

Absolutely. With so much coverage of the Derby preps, so much time to anticipate strategy, and so much public money on the table, the Kentucky Derby is the single best betting race of the year....any year. IMO, somebody just looked at a printout and fired away. If they watched tapes of his races, they would've known he's never taken dirt before and it was 99% likely to happen. They also should've known he'd been the beneficiary of cozy pace situations. Sure he finished big at Hawthorne, but under the circumstances, he pretty much had to. There's no excuse for a serious bettor to make such a public play. But I guess it could be worse; somebody bet $15,000 offshore on Brother Derek.....at 3-1 odds no less. That might've been the dumbest bet of 2006.

[PaceAdvantage]

Yes, but you certainly can't take these isolated incidences (even if it DOES happen to be the Kentucky Derby) and extrapolate them to conclude these guys don't know what they're doing. That would be silly.

You're almost saying that at that level of play, they should be winning every wager they make.

In reality, all they need is an EDGE. Just like a casino. They don't have to win every bet....hell, they don't even need to win 20% of their bets, depending on their EDGE.....

[highnote]

Yes, but you certainly can't take these isolated incidences (even if it DOES happen to be the Kentucky Derby) and extrapolate them to conclude these guys don't know what they're doing. That would be silly.

You're almost saying that at that level of play, they should be winning every wager they make.

In reality, all they need is an EDGE. Just like a casino. They don't have to win every bet....hell, they don't even need to win 20% of their bets, depending on their EDGE.....

Plus, you don't know if this was their only bet. And you don't know who made the bet, for that matter. Maybe it was a bookmaker laying off some of their liabilities. Maybe someone layed Sweetnorthernsaint at low odds and was backing him at higher odds to lock in a risk free profit.

Nowadays you can't just assume it was a flat out win bet -- especially in a race like the Kentucky Derby that is a global wagering event.

[the_fat_man]

I'm glad to hear it. It tells me these computer whales are betting big final figures without a full understanding of how the horse earned the figure. They also clearly must've never watched the Illinois Derby because the video clearly shows the horse was running near a very slow pace.

I didn't cash on the Derby but I was right about one thing: SNS was going to be heavily bet. I took some heat on here back in mid-April for saying he was going to be one of the favorites and I do believe my assertion was correct.

That's incredibly funny stuff. Do you ever own up to your comments?

You 'didn't cash' cause you were all over SNS after the ILLINOIS Derby. Going on and on about his big win. Need we dig up the evidence?

COme on. OWN UP. You got your butt kicked. You're clueless and more importantly, you lack credibility. Now, that's representing the facts.

:D

P.S. Nice touch mentioning that you actually watched the video.

:lol::lol:

[chickenhead]

I think you got the wrong guy, fat man.

Don't remember Val touting SNS, quite the opposite. First post I can find by him on SNS, 4-09:

Sweetnorthernsaint was really more the beneficiary of the slow early/fast late kind of trip. I gave him a 106 speed figure (I use a similar scale to Beyer). But I still wonder if the pace is fast, will he get sucked into those fast fractions?

and his opinion pretty much built from there...quite negative all the way through.

[Valuist]

Chickenhead-

Thank you very much. I remember The Fat Man didn't like the horse but I certainly didn't expect him to be changing my words around and ripping me for telling people in advance a horse would be an underlay.

The Fat Man-

Actually, I did cash a couple matchup plays on the race (including one that was posted). I missed on tris and exactas, even though I had Barbaro on top in a few tickets and had Steppenwolfer in the run-up slots. But I'm not upset; it wasn't like Steppenwolfer was close to beating BGC. And I was the first person that I know of who said SNS would be one of the favorites and I took criticism for it on here.

Getting back to the original point about the Computer Group, Swetyjohn brings up a good point. Maybe it wasn't their only bet; at least for their sake I hope it wasn't. Given the history of the Derby and how longshots often occupy the second or third slots, a win bet on any lower priced horse really isn't the proper play (yes even the winning Barbaro). If the win bet was in 6 figures, one has to think it could've been put to better use in various exotic pools. I would consider this a fundamental wagering mistake on their part. The Derby is not the race to load up on a 5-1 shot to win. And no, I wouldn't expect them to hit every bet or every other bet, I would expect a little more esoteric wagering strategy than betting 100 thousand (or whatever the number was) to win.

[Hank]

The exacta was actually an OVERLAY I had fair pay on it a 409.20

This is an interesting thread, my comrades went down in FLAMES with the saint also. I tried to convince them that their speed pace numbers would betray them in this spot.My OWN numbers were trying to give me SM and SNS,but I was not buying.:cool:

[sjk]

Perhaps someone can help me address the question I tried to raise at the top of the thread.

Even if you think you have a 20% edge, is it rational to be $1million on the Derby with a 20% chance of cashing and an 80% chance of blowing the $1million.

Given the pool restrictions you find in most American racing, no matter what your bankroll size and handicapping acumen it would be a long tough road to win back the million lost through everday betting.

Or perhaps the big players can make this happen more easily than I would imagine. It would be interesting to hear how they could do it.

[LemonSoupKid]

people act like they knew what was going to happen in the Derby as if they "knew it all along" when the result predicted comes true. Seriously, who, over 20 years, consistently does well on the Derby? No one. Why?

Races worth handicapping don't have 20 starters. If you think you can predict reliably when there is only 1 race a year like that, you're crazy. If you predict correctly, more power to you, but don't act like it's obvious that this "was or wasn't going to happen"

Again, I'm way more impressed with prognosticating the scenarios of the Preakness and Belmont. I'll give credit where it's due there and won't mind braggin after the fact ...

LSKid

[PaceAdvantage]

Again, I'm way more impressed with prognosticating the scenarios of the Preakness and Belmont. I'll give credit where it's due there and won't mind braggin after the fact ...
LSKid

Why? You yourself just said how it is most difficult to handicap a race like the Derby. Why would success in handicapping an easier race (Preakness and Belmont) impress you more? That line of reasoning makes little sense.

[Hank]

Assume you are reacting to my post.How am I supposed to ACT, I risked my hard earned cash on what I thought was MOST likely to occur.Derby or 2nd race at lone star whats the diff?I mean this is a hanicapper's forum,most cappers understand were the other is coming from when they critique a race.I mean without CONVICTION how can you bet YOUR money?:ThmbDown:

[LemonSoupKid]

Why? You yourself just said how it is most difficult to handicap a race like the Derby. Why would success in handicapping an easier race (Preakness and Belmont) impress you more? That line of reasoning makes little sense.

Pace, in response, allow me to ask a few questions (and then provide my own answers):

How does anyone develop a skill? They practice it.

The derby is equivalent to MLB making a promo day, something like, "Pitchers dominate today" and putting the hitters up against a pitcher throwing from 50 feet 5 inches. On that one day, you think people have any skill at actually predicting how a hitter will fare against the pitcher? Of course not. It would be far more dealing with random chance. And I wouldn't give a guy props if he said, for example, Jim Thome is going to hit the best that day, or whomever. As such, my point lies in that I just don't buy that by picking the winner you REALLY knew what was going on, compared to the other races. There can be some knowledge about who you think does and doesn't have a chance to do well, but it's way more random than anything. If you do it every year at the Derby windows, I'll buy it. But I don't think anyone does. I do it every year at the Pimlico and Belmont windows --- so I know that profits on those races are VERY REAL possibilities.

The last two are more like races we bet on ALL year. The Derby has MANY factors that make it speculative at BEST.

LS Kid

[PaceAdvantage]

The last two are more like races we bet on ALL year.

Do they run the Belmont at a different distance in your world? How does 1 1/2 miles at Belmont park equate to ANYTHING else I bet on all year?

[Stevie Belmont]

Pace, in response, allow me to ask a few questions (and then provide my own answers):

How does anyone develop a skill? They practice it.

The derby is equivalent to MLB making a promo day, something like, "Pitchers dominate today" and putting the hitters up against a pitcher throwing from 50 feet 5 inches. On that one day, you think people have any skill at actually predicting how a hitter will fare against the pitcher? Of course not. It would be far more dealing with random chance. And I wouldn't give a guy props if he said, for example, Jim Thome is going to hit the best that day, or whomever. As such, my point lies in that I just don't buy that by picking the winner you REALLY knew what was going on, compared to the other races. There can be some knowledge about who you think does and doesn't have a chance to do well, but it's way more random than anything. If you do it every year at the Derby windows, I'll buy it. But I don't think anyone does. I do it every year at the Pimlico and Belmont windows --- so I know that profits on those races are VERY REAL possibilities.

The last two are more like races we bet on ALL year. The Derby has MANY factors that make it speculative at BEST.

LS Kid

Quite possibly the worst post I have ever r ead any where. Senseless is the only word I can think of.

[Tom]

Pace, in response, allow me to ask a few questions (and then provide my own answers):


The last two are more like races we bet on ALL year. The Derby has MANY factors that make it speculative at BEST.

LS Kid

Hello. ALL races are speculative. The Derby is no voodo race, it is a race for 3 yos at a mile and a quarter, a distance that actually makes it more easy than many 8.5 - 9.0 3yo stakes races. ALL races have many factors - not just the Derby. The key is to focus on the ones that really matter.
My hit rate - WP, exactas, is higher on the Derby than my overall record. Other than turf marathons, the Belmont is the only mile and a half race I bet all year - if I bet it at all.

[Valuist]

Again, I'm way more impressed with prognosticating the scenarios of the Preakness and Belmont.
LSKid

2 words describe the Preakness as a betting race the majority of the time:

total shit

You can take a stand in the Derby AND get value. By the time the Preakness comes around, the value is completely shot. Want to bet Barbaro at 4-5? Go ahead. What was Smarty Jones, 1-2? Want to bet Brother Derek off his wide trip? Guess what, he won't be any higher than 3-1. And the winners of the Preakness who were off the board at CD? There were no bargains on Point Given, Pine Bluff or Snow Chief. Afleet Alex, off a 3rd at CD, was the favorite.

There's only one reason to bet Pimlico on Preakness day and its because the undercard is usually very strong.

The Derby as a betting race? No question it is the best.

[Tom]

Best thinkg about the Preakness is the Pimlico Special, the day before, and the undercard the day of.

[the_fat_man]

Chickenhead-


The Fat Man-

Actually, I did cash a couple matchup plays on the race (including one that was posted). I missed on tris and exactas, even though I had Barbaro on top in a few tickets and had Steppenwolfer in the run-up slots. But I'm not upset; it wasn't like Steppenwolfer was close to beating BGC. And I was the first person that I know of who said SNS would be one of the favorites and I took criticism for it on here.



Actually got off my lazy ass and searched the archives.

It turns out that ToeToe (among others) was the one having a man/beast love thing with SNS. :lol:

My apologies.

We were in the same camp for this one.

[Valuist]

Thats what I thought. I remember you didn't like him.

[Houndog]

The Kentucky Derby is a race that is most certainly one of the most challenging, but also offers the potential for great rewards. I have to agree with the other posters that the time to have Barbaro was in this race. With a reduced field in the Preakness it may be a good race to watch. Although I did not have Barbaro as my win bet, I used him with several horses in my exacta plays, but alas no Bluegrass Cat.

[Red Knave]

Even if you think you have a 20% edge, is it rational to be $1million on the Derby with a 20% chance of cashing and an 80% chance of blowing the $1million.I know this is just semantics, but, to me, a 20% edge means that I have 100% chance, given enough opportunities to wager, to get back $2.40 for every $2.00 I bet.
Perhaps you mean, a 20% chance to win this one race based on some probability line. So that should mean that they need to get back $10.00 per $2 wager 20% of the time in order to break even. With the SNS wager they would have received more than that had the horse actually won. Is it rational? Not to me, but then I don't have the ability to wager $1mil on one race. Perhaps with a long enough timeframe on which to judge performance and a large enough bankroll ...

Given the pool restrictions you find in most American racing, no matter what your bankroll size and handicapping acumen it would be a long tough road to win back the million lost through everday betting.Yeah. The KD is one of the few NA races where you actually could bet $1mil and not get 1/9.

[sjk]

Red Knave,

I mean the 20% edge in the same way that you would. You have an expectation that if you bet 1000 races, at the end of the play you will have a profit of 20% of the amount bet.

If your bet on one of the races is far, far greater than the bet on the other 999 races (because the size of the pool allows you to do so) your assurance of being in the black is much less (in fact less than 50-50) even though your expected value would still be 20%.

I don't think it is a rational thing to make that outsized bet but was looking for other points of view.

[highnote]

I don't think it is a rational thing to make that outsized bet but was looking for other points of view.

It is rational if you have an edge. It is possibly irrational if you don't.

There are some people who bet 1 million but don't have an edge because for them it is entertainment. Some people like expensive wine. Some people like yachts. etc etc etc

But if you have an edge and you have a 50 or 100 million bankroll, then 1 million is not a big wager.

[Red Knave]

If your bet on one of the races is far, far greater than the bet on the other 999 races (because the size of the pool allows you to do so) your assurance of being in the black is much less (in fact less than 50-50) even though your expected value would still be 20%.Okay, I get your question now.
I guess it's easy to see that I can't answer from experience. :confused: ;)

[sjk]

What if you did have that large a bankroll. How much could you really bet on the run of the mill North American race. It seems to me that if you are shooting for 10-1 type winners you can't bet much more than 1% of the pool before the effect on the payoff prevents you from betting more.

Probably someone out there does this calculation as a matter of their normal betting and might post the correct answer.

That is one real difference between ROI and rebates. Betting larger amounts erodes your ROI but not your rebate.

[fastCow]

What if you did have that large a bankroll. How much could you really bet on the run of the mill North American race. It seems to me that if you are shooting for 10-1 type winners you can't bet much more than 1% of the pool before the effect on the payoff prevents you from betting more.
.

Reminds me of the super-hedge fund created by nobel-prize-winning economists about a decade ago. After some initial successes, they somehow ended up being over-leveraged and under-capitalized and just sunk under some cascading bad luck. Among the pools they were betting in were small ones like the Dutch real estate market, which they were the whale in.

Like Mandelbroit says, if you want to last the trip make sure that your craft can stand up to bad weather, even though it's usually clear and sunny.