"Fed Up With the CAWs..."

[Poindexter]

Printed out the past performances. It was an 8 horse field with an entry, not a 9 horse field. Ability wise pretty closely matched field. In fact don't like the favorite at all. Not sure I would have pegged Coppola as a lone speed, but it turned out that way I guess. There certainly did not seem to be a lot of speed in the race and Coppola looks pretty good off his Churchill win and even his last race where he faded. Prats horse was probably a pretty easy toss because he doesn't appear to be a any good on the turf. Winner is actually a pretty solid horse too. Was very good Grade 2, 3 runner last year Went sour. Took some time off, last was 3rd off the layoff where he had trouble at the start and still earned a pretty decent figure. Doesn't take a crystal ball to give him a big look in here. So from a handicapping perspective I don't agree with my previous take on the race. So based off of that I will admit there is nothing to investigate in this race. Sharp capping.

[Nitro]

Originally Posted by Dave Schwartz

Before you put a UNTRUE tag on something I post, perhaps you should make sure that you understand the words.

"Models" means they have a database of tote history and the ability to analyze it.

It does not mean they have superior access.

While they can hit the tote system at will, they still get the same information that is available to everyone else.

That information is not updated from the sources any faster; only the dissemination is faster.

Hey Dave. Sorry about my misinterpretation of your one-line comment. However, I think you might agree that the term “Models” can mean any number of things. Apparently, you’re not aware that those of us who rely exclusively on real time O.A. tote data also maintain databases of tote history. So there again some do have what you believe only the CAW’s have.

I don’t to want to misinterpret your take on the “Dissemination” of the tote information, but from what I experience the distribution of that information is as fast as the updating occurs. Perhaps I’m not getting what action you’re actually referring to. Please enlighten me.
.
.

[Dave Schwartz]

Originally Posted by Nitro

Hey Dave. Sorry about my misinterpretation of your one-line comment. However, I think you might agree that the term “Models” can mean any number of things. Apparently, you’re not aware that those of us who rely exclusively on real time O.A. tote data also maintain databases of tote history. So there again some do have what you believe only the CAW’s have.

I don’t to want to misinterpret your take on the “Dissemination” of the tote information, but from what I experience the distribution of that information is as fast as the updating occurs. Perhaps I’m not getting what action you’re actually referring to. Please enlighten me.
.
.

Apology accepted.

Dissemination in this context means that:
They can get the most current odds available with one click.
We have to wait until some website gets their data and puts it up.

IOW, they get their data before BRISnet for example.

So, while we wait for BRIS to get the data up on their site, they get it on demand.

However, that isn't the real bottleneck in the data.

Tracks and ADWs pass their betting data into the tote company. Some are faster than others. Neither are they necessarily consistent in how the do it.

So, the real bottleneck is at SciGames while they wait for data.

In that sense, they face the same challenges as us. What they do get is (basically) one last few seconds flash before the bets are fired and processed.

Truth: It hardly matters.

What DOES matter is that they are far better at projecting the final odds than we are because of the data they've captured - and having the tools to analyze it.

But make no mistake - they get hornswoggled, too.

That's because they cannot predict with much accuracy how the other whales are going to bet.

Know this - when the planet line up and all the whales wind up on the same horses, their horses crash, too.

When there is a day of (say) 7 winners under $5 at a track, you can be assured that ALL the whales had losing days.

[Robert Fischer]

Originally Posted by thaskalos

This is Saratoga, where the tote board doesn't tilt because some guy bet some lucky number. IMO...this sort of exacta combination isn't the type that some sophisticated handicapping algorithm would have isolated on. What if there is something of a more nefarious nature going on...and the CAW action is unintentionally providing convenient cover for it?

When you are a nut like me, I see these short exactas and most cases appear to be completely normal of misread circumstances, a short field and/or some very high longshot non-contenders off-the-board, or normal performances, and randomness, and then there are a minority of these instances where the favorite doesn't perform(w/out excuse other than rating tactics), or the less creative ones where someone trots around on a dream uncontested lead attract a possibility of suspicion.

If you hypothetically had a leading jockey that could occasionally run out of the exacta for you, it would be enough. A couple other teammates of the leading jockey who could run duds or occasionally do some race riding/pace-pressuring, it would be the cherry on top.

This influence could come from positive encouragement like financial kickbacks, and inclusion/lifestyle, or it could come from the natural concentration of super trainers and super owners who provide a large slice of the jockey's earnings and thus loyalties to the team of the jockeys who ride as uncoupled entries in these stakes, preps, MSW, etc..., together.

I would guess it would have to be from inside the CAWs themselves or some really sharp presence who could operate in the blind spot or 'cover' of the CAWs big batch watchers in the final cycles.
If it is a sharp private dark syndicate I give them credit in saying I don't think that sharp of a team exists in these pools and that it would in my opinion come from inside the CAWs. Some of the ownerships and ownership teams in the oddly wagered+performed races could be a clue. That's as far as I care to speak of it.

[bks]

Originally Posted by classhandicapper

I love the sport. I enjoy watching the best run against the best. If I find a bet that's great, but if not, I'll still be interested in a great race. We are at the point where a handful of trainers have almost all the best horses, they avoid running against each other and try to keep their own horses apart. So we are getting way too many small or uncompetitive Stakes fields. That makes many of them unbettable too. We aren't going to attract many new fans of the sport that way.

The CAWs don't bother me beyond them having access to pool information I don't have. If I had the technological and statistical skills to create an advanced model and get my bets off later, I'd be a CAW getting bigger rebates. Otherwise, they are just another very skilled player I am competing with. I have to find holes in their game. Keep in mind, we are also competing against insiders that have access to information on the condition and treatment of their horses that we don't have.

But that's just the point. You (and I, and every other individual player) aren't in a position to evaluate the decision-making processes that result in their edge. We only see the results; they themselves may not even understanding the DMPs of a complex algorithm. Long-term, they can't be beaten under these conditions, because of the difference in resources. I know the libertarians on the board like to pretend everything can be conquered through the sheer application of individual ingenuity, but it can't.

My play is now mostly limited to tournaments, where it seems I can still get "lucky" enough to make it fun/profitable - I qualified for the BCBC on the basis of a $20 entry this year. I have no illusions I can win grinding day to day, though.

[bks]

Originally Posted by Dave Schwartz

I agree with your assessment 100%.

The key is to do something that would be antithetical to their approach to the game.

Their approach is a global system with local optima models added.

IMHO, a Topological Data Analysis approach would fit the bill nicely.

Respectfully, Dave, I don't think they HAVE a defined approach beyond the search for exploitable correlations. That's the only unconditional element of their "approach." The search for exploitable situations isn't limited to any specific aspect of handicapping - it's about the continuous identification of exploitable situations and capitalizing on them. From the standpoint of these CAW groups, it doesn't really matter why a situation is profitable. All that matters is:

1) that they can identify them; and
2) they can anticipate new exploitable situations as quickly or more quickly than the general population of bettors, which they usually can, with their greater analytical abilities.

[Dave Schwartz]

Originally Posted by bks

Respectfully, Dave, I don't think they HAVE a defined approach beyond the search for exploitable correlations. That's the only unconditional element of their "approach." The search for exploitable situations isn't limited to any specific aspect of handicapping - it's about the continuous identification of exploitable situations and capitalizing on them. From the standpoint of these CAW groups, it doesn't really matter why a situation is profitable. All that matters is:

1) that they can identify them; and
2) they can anticipate new exploitable situations as quickly or more quickly than the general population of bettors, which they usually can, with their greater analytical abilities.

I don't have a horse in this race.

You said, "You don't think..."
You will believe what you believe.

What you said makes it sound as if there is some "individual judgment" involved in this process.

I assure you that they are 100% formulaic.

Nobody is capable of betting into

• 80+% of the tracks
• 80+% of the races
• 80+% of the pools
• every single day of the year
• Wagering well over $100 million dollars a year

... without a formulaic and systematic approach.

Truth: It is not really Robotic Wagering either.

I'm sure it could be if the gates always opened on time, but (currently) someone has to click the button on every race.

This is not my opinion.
This is fact that I have seen with my own eyes.

[Dave Schwartz]

Originally Posted by bks

But that's just the point. You (and I, and every other individual player) aren't in a position to evaluate the decision-making processes that result in their edge. We only see the results; they themselves may not even understanding the DMPs of a complex algorithm. Long-term, they can't be beaten under these conditions, because of the difference in resources. I know the libertarians on the board like to pretend everything can be conquered through the sheer application of individual ingenuity, but it can't.

My play is now mostly limited to tournaments, where it seems I can still get "lucky" enough to make it fun/profitable - I qualified for the BCBC on the basis of a $20 entry this year. I have no illusions I can win grinding day to day, though.

But now I am confused.

This post tells a different side.

Obviously, I have your meaning wrong.

Lest I make the same folly someone else made recently, please educate me as to your meaning of the other post I responded to.

I guess I simply did not get what you are saying.

[bks]

Originally Posted by Dave Schwartz

But now I am confused.

This post tells a different side.

Obviously, I have your meaning wrong.

Lest I make the same folly someone else made recently, please educate me as to your meaning of the other post I responded to.

I guess I simply did not get what you are saying.

No problem. You seemed to indicate a sharp individual or small group could gain an edge over the CAWs by "do(ing) something that would be antithetical to their approach to the game." My point was that I don't think it's possible over any extended period of time to do that "something" without their analytical tools also identifying it.

In other words, if YOU noticed that something, and you noticed it with your eyes, then either:

1. It will soon show up in their analyses and be built into the next model iteration, and the edge will be gone; or

2. your eyes are lyin' to you, and whatever edge you believe yourself to have will reveal itself over time to be something else.

I think we broadly agree here, except possibly on this point.

[Dave Schwartz]

Originally Posted by bks

No problem. You seemed to indicate a sharp individual or small group could gain an edge over the CAWs by "do(ing) something that would be antithetical to their approach to the game." My point was that I don't think it's possible over any extended period of time to do that "something" without their analytical tools also identifying it.

In other words, if YOU noticed that something, and you noticed it with your eyes, then either:

1. It will soon show up in their analyses and be built into the next model iteration, and the edge will be gone; or

2. your eyes are lyin' to you, and whatever edge you believe yourself to have will reveal itself over time to be something else.

I think we broadly agree here, except possibly on this point.

Oh, IMHO, the eyeball scan approach has zero chance of playing at a whale level.

The whales use a global, formulaic approach.
Can we agree on that?

If we agree on this, I suspect we will agree on most everything.
_______
What I am advocating is multiple systems - each designed to attack a subset of races data.

The problem with doing this manually is that the data segments are just too many to create because when you break them down using logic, you wind up with 10s of thousands.

Literally.

Aside from this being nearly impossible to build, it would be equally difficult to maintain.

In addition, the sample sizes of these data sets would be too small.

However, an AI engine could do this.

What I am specifically advocating is a Topological Data approach in which the system builds some number of systems then fits the system to a data set which it has decided fits best.

If you read that again, the AI is not creating data sets, it is creating systems and fitting them to data sets via rules about each race.

IOW, the AI engine manages the system creation/development and decides on a race-by-race basis which system works best in this situation.

[classhandicapper]

Originally Posted by bks

No problem. You seemed to indicate a sharp individual or small group could gain an edge over the CAWs by "do(ing) something that would be antithetical to their approach to the game." My point was that I don't think it's possible over any extended period of time to do that "something" without their analytical tools also identifying it.

In other words, if YOU noticed that something, and you noticed it with your eyes, then either:

1. It will soon show up in their analyses and be built into the next model iteration, and the edge will be gone; or

2. your eyes are lyin' to you, and whatever edge you believe yourself to have will reveal itself over time to be something else.

I think we broadly agree here, except possibly on this point.

You seem to be trying to talk yourself into believing it's impossible to win as long as CRWs are playing. I disagree.

When you are handicapping, there will be times you think a horse is a big overlay or underlay. Over time, you're going to see patterns in where you are doing consistently well and where you are not. If you have a database or are willing to do a lot of manual work, you can then test those situations (and virtually anything else you can think of) to see what the long term data says.

If you find something with some value, try to exploit it.

If you don't, don't play.

If others catch on to something that was working, so be it. Those horses will stop looking like overlays to you and should stop betting them.

If they catch on to everything that has worked for you, stop playing or keep looking.

[Al Gobbi]

Twitter

Twitter

[Poindexter]

Originally Posted by Al Gobbi

https://twitter.com/PatCummingsTIF/s...75386351538450

https://twitter.com/PatCummingsTIF/s...75387723084127

I guess that sort of dispels the myth that if you eliminated rebates and brought takeout to proper levels the caw would be gone and racing would not be able to survive.

[ScottJ]

Originally Posted by Poindexter

I guess that sort of dispels the myth that if you eliminated rebates and brought takeout to proper levels the caw would be gone and racing would not be able to survive.

Exactly what does Pat achieve by "leaking" this information?

First, he does not help the small bettor by letting you know at which track this is occurring.

Second, he does not give insight on the data to let you know exactly how serious the net-negative on the small bettor might be.

Third, he effectively keeps this information close so that he gives a clear message that he can trade on data which is not available to the public.

Quite frankly, while Pat's intentions might have been pure here, he does nothing to help anyone other than of course pushing his own name forward as being right through a self-serving "I'm hearing big things" post.

[ScottJ]

Originally Posted by Al Gobbi

https://twitter.com/PatCummingsTIF/s...75386351538450

https://twitter.com/PatCummingsTIF/s...75387723084127

Name the track, Pat. You believe that the TIF is interested in protecting the industry? Step up to the plate and show it here.

Let me give you another, darker, interpretation of Pat's leaked phone call. "Learned today via a racetrack executive that a group of CAW players have infiltrated their racetrack so completely that they are clearly able to beat the takeout by gaining private advantages from the backstretch employees and we have not cracked the situation."

Why leave this doubt? Better to have said nothing.