Software Learning Curve

[Dave Schwartz]

Really? You weren't the one that said "whales start out as whales"?

Whales start out as whales means that they begin with a vision, lots of capital and a commitment to bet vast sums of money.

They do not start with a handicapping system that they have already been using because (as Game Theory said) it does not scale.

To generate that kind of handle, one must play about 80% of pools, in 80% of the races, at 80% of the tracks. (Note: It makes no practical difference if the number is actually 85%.)


Being a whale is a difficult endeavor. You must have a systematic approach that takes zero personal subjectivity, that can be used by a worker-bee operator.

Because bets are made as the horses are loading, it takes a staff of agents to pull the trigger in a timely manner. That staff needs managers or supervisors.

To stay ahead of the game (as Traynor has pretty much said) you must be constantly studying and improving. That also takes a staff.

The software is in a constant state of development; another small staff.

The math work takes specialists. More staff.

The staffing takes staff - i.e. to track hours, write checks, replace people, etc.

Suddenly, it is a real business.


How do WE compete? I'm trying but it is very difficult without manpower.

[thaskalos]

Originally Posted by raybo

Price, as a stand alone variable, is also meaningless. It takes the right mix of both to be profitable.

Furthermore, if you are talking about "degrees" of profitability, strike rate takes on additional meaning. Why? Because, the higher the hit rate, the more you can wager. The higher the strike rate, inevitably results in more lower priced horses, and lower priced horses will be able to take more money without reducing those odds. So, in such a case, one could produce much higher capital turnover, and as a result, much higher profits, especially if you are getting decent rebates.

Also, the higher the hit rate, the fewer and shorter the losing streaks become...and the fewer, shorter losing streaks do not demand the same type of bankroll as those longer, more numerous losing streaks which are associated with the lower hit rates.

If I have a higher hit rate, then I can make higher bets with the same bankroll than the player who has a lower hit rate but a higher ROI than me. And those higher bets can often overcome the ROI edge that the other player has.

ROI is not the true measure of profitability. Profit relative to bankroll is the real measure, IMO. At least until the player gets to the point where he can't bet any higher because of mutuel pool-size limitations.

[whodoyoulike]

Originally Posted by raybo

I'm sure Traynor will answer your questions, but I'll fill in the time lag. If you are talking about the type of program that results from thorough, extensive historical research, which most good programs do, then the models it produces are dependent on consistent application, meaning if the program gives you betting selections, then you must bet them, or you destroy the model.

As to your 2nd question, any good program, recreational or professional, should have its own rules contained within it, regarding pass or play situations, and necessarily should tell you when to play and when to pass. Mine does, and it is hardly "professional" quality software, as described by Traynor anyway.

An example which comes to mind, is when a horse has only run sprints 5.5f and 6f in good time and is now going 8f or 8.5f. I've noticed that these types are always identified as contenders. I always have to evaluate whether they fit this race.

[raybo]

Originally Posted by PaceAdvantage

Where does the "daddy's money" come into play again? You kinda lost me with that one a few back, and I was expecting you to mention it again since you are answering a "whale" question, but you curiously left it out of this reply.

I think he probably means that if you are betting with "daddy's money" then you don't necessarily know what the heck you're doing, but are betting huge amounts anyway, because you have access to the money to do so, "daddy's money".

[raybo]

Originally Posted by thaskalos

Also, the higher the hit rate, the fewer and shorter the losing streaks become...and the fewer, shorter losing streaks do not demand the same type of bankroll as those longer, more numerous losing streaks which are associated with the lower hit rates.

If I have a higher hit rate, then I can make higher bets with the same bankroll than the player who has a lower hit rate but a higher ROI than me. And those higher bets can often overcome the ROI edge that the other player has.

ROI is not the true measure of profitability. Profit relative to bankroll is the real measure, IMO. At least until the player gets to the point where he can't bet any higher because of mutuel pool-size limitations.

I'm glad you posted that, I was going to go back and edit my post, to include the bit about fewer, and shorter losing streaks. For someone primarily interested in "degree" of profitability, I agree with everything you said.

[raybo]

Originally Posted by whodoyoulike

An example which comes to mind, is when a horse has only run sprints 5.5f and 6f in good time and is now going 8f or 8.5f. I've noticed that these types are always identified as contenders. I always have to evaluate whether they fit this race.

Good software will do that evaluation for you. It's all part of models creation.

"Good" software, IMO, is absolutely filled with "decision trees", lines of code or logic constructs, that ask questions and depending on the answers, take a specific path that ultimately results in all, or almost all possible scenarios being accounted for. As someone said a long time ago (several years ago somewhere on this forum), if you can think it, computers can be programmed for it.

[PaceAdvantage]

Originally Posted by raybo

I think he probably means that if you are betting with "daddy's money" then you don't necessarily know what the heck you're doing, but are betting huge amounts anyway, because you have access to the money to do so, "daddy's money".

And here I took it to mean that the only reason there are "whales" is because they were all wealthy to begin with, primarily from inheritance...I was under the impression that Traynor believes there are no true whales making serious, serious money.

[thaskalos]

Originally Posted by PaceAdvantage

And here I took it to mean that the only reason there are "whales" is because they were all wealthy to begin with, primarily from inheritance...I was under the impression that Traynor believes there are no true whales making serious, serious money.

Well...according to Dave, they MUST be wealthy to begin with...no?

How else could they set up such a costly enterprise?

[AndyC]

Originally Posted by Dave Schwartz

Whales start out as whales means that they begin with a vision, lots of capital and a commitment to bet vast sums of money.

There is a vast difference in having a vision and realizing a vision.


[QUOTE=Dave Schwartz]They do not start with a handicapping system that they have already been using because (as Game Theory said) it does not scale.

Nonsense. A handicapping system is merely an estimation of actual probabilities. Scale applies to the act of betting while maintaining
a statistical edge. There is no need to scrap a solid handicapping system.

[raybo]

Originally Posted by thaskalos

Well...according to Dave, they MUST be wealthy to begin with...no?

How else could they set up such a costly enterprise?

I think Dave accounted for that when he mentioned that the software creator could be offered a percentage of future profits, which would offset the upfront costs somewhat. He also established that one wouldn't necessarily need to have gobs of money, oneself, when he mentioned other investors being included in the project. There are probably more than a few people, with lots of money, willing to invest in such a project.

[raybo]

Originally Posted by PaceAdvantage

And here I took it to mean that the only reason there are "whales" is because they were all wealthy to begin with, primarily from inheritance...I was under the impression that Traynor believes there are no true whales making serious, serious money.

I have no idea what Traynor believes or doesn't believe, because I've seen many examples of his taking both sides of discussions, even within the same thread.

[thaskalos]

According to Dave, there are 6 Whales operating in this game...and these 6 Whales were "Whales to begin with".

How can we be SURE that they were "Whales to begin with"? Isn't it possible that they started smaller...and they grew in size as their profit and experience increased?

If I bet $100-million a year, then I am a Whale. Do I bet this to begin with?

[AndyC]

Originally Posted by raybo
Price, as a stand alone variable, is also meaningless. It takes the right mix of both to be profitable.

Furthermore, if you are talking about "degrees" of profitability, strike rate takes on additional meaning. Why? Because, the higher the hit rate, the more you can wager. The higher the strike rate, inevitably results in more lower priced horses, and lower priced horses will be able to take more money without reducing those odds. So, in such a case, one could produce much higher capital turnover, and as a result, much higher profits, especially if you are getting decent rebates.

Originally Posted by thaskalos

Also, the higher the hit rate, the fewer and shorter the losing streaks become...and the fewer, shorter losing streaks do not demand the same type of bankroll as those longer, more numerous losing streaks which are associated with the lower hit rates.

If I have a higher hit rate, then I can make higher bets with the same bankroll than the player who has a lower hit rate but a higher ROI than me. And those higher bets can often overcome the ROI edge that the other player has.

ROI is not the true measure of profitability. Profit relative to bankroll is the real measure, IMO. At least until the player gets to the point where he can't bet any higher because of mutuel pool-size limitations.

Good points.

[whodoyoulike]

Originally Posted by raybo

Good software will do that evaluation for you. It's all part of models creation.

"Good" software, IMO, is absolutely filled with "decision trees", lines of code or logic constructs, that ask questions and depending on the answers, take a specific path that ultimately results in all, or almost all possible scenarios being accounted for. As someone said a long time ago (several years ago somewhere on this forum), if you can think it, computers can be programmed for it.

Is your software capable of evaluating when a sprinter can go longer? I've find that sometimes they can and other times they aren't suited to the pace.

[raybo]

Originally Posted by whodoyoulike

Is your software capable of evaluating when a sprinter can go longer? I've find that sometimes they can and other times they aren't suited to the pace.

If your decision process is not based on "intuition", but concrete data, then yes it could be programmed.