The Handicapper's Cafe

[traynor]

Originally Posted by TrifectaMike

Is it necessary for a "black box" to be universal to be considered a "black box"?

Mike ( Dr Beav has left the building)

Not at all, as long as the black box has a little message box that pops up and says, "Do Not Bet on Races at This Track" for tracks at which wagers will not be profitable (based on real world results in the future, not what happened last week, last month, or last year), and another message box that pops up and says, "Bet Your Mortgage Payment and Every Dollar You Can Beg, Borrow, or Steal at This Track" for tracks at which wagers will be profitable (based on real world results in the future, not what happened last week, last month, or last year).

[TrifectaMike]

Originally Posted by traynor

I think that "hope springs eternal" and can be seriously detrimental to the bank account unless liberally seasoned with good sense and frequent reality checks.

I don't know how many realize how difficult it is to convert regression into future profit in the real world. I bought an application once (many years ago) that showed an ROI of 1.05 or better at "most" tracks (including some I had never even heard of before). Impressive, colorful bar charts showing how much profit had been "earned" at each track. A number of tracks were not represented, but there were so many that it seemed it "worked everywhere all the time."

The problem was that the algorithms essentially "detected" anomalies that might only happen a few times during a racing season, if at all, like "the fastest, highest class entry with a top three trainer, a top three jockey, at a distance it has won in the last month, but only when it goes to post at odds of 3/1 or better." It sounds foolish (like many other software apps for Australian races) but it actually worked in theory. It picked good horses in races with false favorites. On paper, it worked well. In the real world, it didn't work at all.

Why are you so hung up on regression? And what leads you to believe that everyone is using regression?

Oh, and what is a false favorite?

Mike

[traynor]

Originally Posted by TrifectaMike

Why are you so hung up on regression? And what leads you to believe that everyone is using regression?

Oh, and what is a false favorite?

Mike

Are these rhetorical questions?

[TrifectaMike]

Originally Posted by traynor

Are these rhetorical questions?

NO!

[Greyfox]

Originally Posted by TrifectaMike

Oh, and what is a false favorite?

Mike

That is a rhetorical question.

[traynor]

Originally Posted by Greyfox

That is a rhetorical question.

I often sympathize with those who seem to be overcompensating for feelings of inadequacy and inferiority by assuming the mantle of "expert." I suggest it would be more useful (and much less obvious) if a more egalitarian attitude were adopted, and perhaps even a greater willingness to share information rather than alluding to (unproven and possibly completely false) assumptions that manifest the truth of the old saying, "To a carpenter with a hammer, everything looks like a nail."

None of which should be considered ad hominem. That is not my intent. I consider such a boring waste of time that can be used more productively doing other things.

[Greyfox]

Originally Posted by traynor

I often sympathize with those who seem to be overcompensating for feelings of inadequacy and inferiority by assuming the mantle of "expert."

Undoubtedly Trifecta Mike is a top class ace at Mathematics. Certainly an expert in that department.
He may well be a good horse player as well.
I've yet to ever see that though on this board, although he did invite some people to try a project with him. (I haven't seen the results of that posted anywhere.)
The problem with his posts is that he comes across as Socrates and wants to evoke the answer from you. Of course he knows the answer from a math perspective.
But there's a leap from the chalkboard to the race track.
I'm waiting for concrete proof that he can really apply the stuff he theorizes about.

[thaskalos]

I have an idea which I think might prove to be both educational and entertaining at the same time.

Why don't some of the more outspoken among us proceed to handicap a race or two from the upcoming weekend's card...right here in open view.

We don't have to divulge information that we are uncomfortable sharing...nor should we feel the pressure of having to be "right" in our brief handicapping sample.

We all know that this is a marathon rather than a sprint...and that the long run is what counts.

This would effectively push the conversation in a new and more interesting direction...where "real" information is being conveyed...instead of the veiled inferences of theories which remain difficult for most of us to conceptualize.

After all...there is a world of difference between "theory" and "practice"...not only in handicapping, but in everything else as well.

Would it be possible to move over to the "practical" side of things...so we could have a more well-rounded and productive discussion here?

I gladly volunteer to be the first to start this off -- since I was the one who started off this thread -- but I would like to see more people follow my lead, so we can get a "real" exchange of ideas going...instead of just poking holes in my view of things.

What say you?

[JackS]

IMO, The public is onto the best horse (favorite) at least 66% of the time in every race. The fact that they only cash at about 33% has nothing to do about their inability to land on the right horse. Most of the time, when the legitimate favorite loses, we can only try to make excuses for the horse and this can only be done after the race is run.
The remaining 33% of the time, the race can be judged "wide open". In this case, the public is neither right or wrong. This is the type race to be looking for false favorites. In this case, the public will be more willing to accept a single factor i.e- top jock, top trainer most money won etc.

[raybo]

Originally Posted by traynor

It makes sense, but I think a more appropriate question might be, "Are there tracks or situations in which extended play using your application over a period of time results in losses?" That is not intended to be a snarky question. I have seen way too many people fasten on one set of results that show what they want to see, and completely ignore another set of results that is equally compelling evidence of things they don't want to see.

Yes, there are tracks that are unprofitable using my program. That is why we do track testing before playing them.

You keep using the term "regression", I have heard to term before, "logistic regression", for example, but haven't a clue about it, not having the required higher math education such a method would require.

I simply look at all the winners of all the races in my database, and what types of pace pressure races they came from and grab the data that comes from the majority of the winners in each of those pace pressure groupings (not quite that simple but that's the basics). I use that data to eliminate horses from win contention, in like pace pressure races, then apply other calculations for further eliminations, and finally rank the remaining contenders by a proprietary total velocity formula and bet any of the top 3 ranked horses, to win, if their odds are acceptable.

To my knowledge, nobody else does exactly what I do in the program, so the picks it produces are unique, in that aspect. That is the reason we get both lower odds picks as well as boxcar odds picks, and everything in between. Most programs pick either low odds horses, or longshots, not both. We concentrate on tracks that have higher average win payouts, in order to bet those higher odds winners. That is the reason some tracks are unprofitable and others are profitable, the average win payouts combined with the program's hit rate on it's picks.

Aqueduct, for example has a great hit rate with the program, but the average win payout at that track is one of the lowest in the country, thus the program produces a net ROI of about .90 there, because we don't bet program picks below our minimum odds requirement, while betting those picks that are at or above our minimum odds requirement. So, if we bet 2 horses, of the possible 3 produced by the program, and the pick we didn't bet, due to it's odds being too low, won the race, then even though the program picked the winner, we did not bet it, but rather the other 2 picks, and we lost the money we bet on those other 2 picks. Again, do the math, some tracks are profitable and others are not.

[traynor]

Originally Posted by raybo

Yes, there are tracks that are unprofitable using my program. That is why we do track testing before playing them.

You keep using the term "regression", I have heard to term before, "logistic regression", for example, but haven't a clue about it, not having the required higher math education such a method would require.

I simply look at all the winners of all the races in my database, and what types of pace pressure races they came from and grab the data that comes from the majority of the winners in each of those pace pressure groupings (not quite that simple but that's the basics). I use that data to eliminate horses from win contention, in like pace pressure races, then apply other calculations for further eliminations, and finally rank the remaining contenders by a proprietary total velocity formula and bet any of the top 3 ranked horses, to win, if their odds are acceptable.

To my knowledge, nobody else does exactly what I do in the program, so the picks it produces are unique, in that aspect. That is the reason we get both lower odds picks as well as boxcar odds picks, and everything in between. Most programs pick either low odds horses, or longshots, not both. We concentrate on tracks that have higher average win payouts, in order to bet those higher odds winners. That is the reason some tracks are unprofitable and others are profitable, the average win payouts combined with the program's hit rate on it's picks.

Aqueduct, for example has a great hit rate with the program, but the average win payout at that track is one of the lowest in the country, thus the program produces a net ROI of about .90 there, because we don't bet program picks below our minimum odds requirement, while betting those picks that are at or above our minimum odds requirement. So, if we bet 2 horses, of the possible 3 produced by the program, and the pick we didn't bet, due to it's odds being too low, won the race, then even though the program picked the winner, we did not bet it, but rather the other 2 picks, and we lost the money we bet on those other 2 picks. Again, do the math, some tracks are profitable and others are not.

Sorry about the confusion in terminology. "Regression" generically means "studying what happened in the past."

It sounds like you could really be on to something. I wish you the best of luck with it. Many of the same caveats would still apply about past results not traslating into the future, but you seem to have a lot of it covered. From what you have said, it would seem more likely that results could repeat using your approach because the selections are not obvious. That is, it is unlikely that other bettors would detect a pattern and exploit it.

I appreciate you taking the time to explain what you are doing. It makes a lot more sense now. Thank you.

[Capper Al]

Originally Posted by thaskalos

I have an idea which I think might prove to be both educational and entertaining at the same time.

Why don't some of the more outspoken among us proceed to handicap a race or two from the upcoming weekend's card...right here in open view.

We don't have to divulge information that we are uncomfortable sharing...nor should we feel the pressure of having to be "right" in our brief handicapping sample.

We all know that this is a marathon rather than a sprint...and that the long run is what counts.

This would effectively push the conversation in a new and more interesting direction...where "real" information is being conveyed...instead of the veiled inferences of theories which remain difficult for most of us to conceptualize.

After all...there is a world of difference between "theory" and "practice"...not only in handicapping, but in everything else as well.

Would it be possible to move over to the "practical" side of things...so we could have a more well-rounded and productive discussion here?

I gladly volunteer to be the first to start this off -- since I was the one who started off this thread -- but I would like to see more people follow my lead, so we can get a "real" exchange of ideas going...instead of just poking holes in my view of things.

What say you?

I do most weekends in another Capper's Corner on another website. It's a contest of WPS bets on our top selections. I don't wager that way, but haven't done too bad.

[traynor]

Originally Posted by Greyfox

Undoubtedly Trifecta Mike is a top class ace at Mathematics. Certainly an expert in that department.
He may well be a good horse player as well.
I've yet to ever see that though on this board, although he did invite some people to try a project with him. (I haven't seen the results of that posted anywhere.)
The problem with his posts is that he comes across as Socrates and wants to evoke the answer from you. Of course he knows the answer from a math perspective.
But there's a leap from the chalkboard to the race track.
I'm waiting for concrete proof that he can really apply the stuff he theorizes about.

I meant no personal affront, nor was a personal attack intended, on Mike or anyone else. I have a character defect developed over many long hours listening to professors pontificate on topics they have to teach because no one will pay them to apply the content of that topic in the real world. When I read things that look like "osmosis of the cosmosis" I tend to blank it out and a little voice in the back of my head asks, "Will this be on the final?"

[raybo]

Originally Posted by thaskalos

I have an idea which I think might prove to be both educational and entertaining at the same time.

Why don't some of the more outspoken among us proceed to handicap a race or two from the upcoming weekend's card...right here in open view.

We don't have to divulge information that we are uncomfortable sharing...nor should we feel the pressure of having to be "right" in our brief handicapping sample.

We all know that this is a marathon rather than a sprint...and that the long run is what counts.

This would effectively push the conversation in a new and more interesting direction...where "real" information is being conveyed...instead of the veiled inferences of theories which remain difficult for most of us to conceptualize.

After all...there is a world of difference between "theory" and "practice"...not only in handicapping, but in everything else as well.

Would it be possible to move over to the "practical" side of things...so we could have a more well-rounded and productive discussion here?

I gladly volunteer to be the first to start this off -- since I was the one who started off this thread -- but I would like to see more people follow my lead, so we can get a "real" exchange of ideas going...instead of just poking holes in my view of things.

What say you?

I'd be willing to do this, depending on the track you choose to play. Of the ones currently running I would choose to play Calder, but that's up to you.

And, I hope everyone who follows the selections realizes that such a small sample of plays means very little, long term, winners or losers. Not an advance "excuse for poor performance" but, simply, the truth.

I have gone through and posted picks for races, from several different tracks, before the fact, here and on my forum, many times in the past, posting pre-race screenshots of the program's picks, so I have no qualms about doing it again.

Who else is in, Traynor, Mike, Al, etc..?

Oh, Traynor, if you think I'm "stroking my ego", and don't like to share my knowledge and experience, do some checking in the archives, you could start with the "Black Box Challenge we had here (although I was using a different "black box" then, I did quite well), or better yet check out the 2 "free" workbooks I have made available to anyone who wants to download them. Some of us, have already become profitable in racing and have decided to help the industry by helping others become profitable. As long as most potential new players think that horse racing cannot be beaten, yes, even with a black box, the industry will continue to decline, and other games involving gambling, like poker, publicize the multi-million dollar winners of tournaments around the world will continue to steal those potential new horse racing players from our industry.

I don't know about you, but I would like for horse racing to keep existing for a while longer (at least until I kick the bucket, anyway).

[traynor]

Originally Posted by thaskalos

I have an idea which I think might prove to be both educational and entertaining at the same time.

Why don't some of the more outspoken among us proceed to handicap a race or two from the upcoming weekend's card...right here in open view.

We don't have to divulge information that we are uncomfortable sharing...nor should we feel the pressure of having to be "right" in our brief handicapping sample.

We all know that this is a marathon rather than a sprint...and that the long run is what counts.

This would effectively push the conversation in a new and more interesting direction...where "real" information is being conveyed...instead of the veiled inferences of theories which remain difficult for most of us to conceptualize.

After all...there is a world of difference between "theory" and "practice"...not only in handicapping, but in everything else as well.

Would it be possible to move over to the "practical" side of things...so we could have a more well-rounded and productive discussion here?

I gladly volunteer to be the first to start this off -- since I was the one who started off this thread -- but I would like to see more people follow my lead, so we can get a "real" exchange of ideas going...instead of just poking holes in my view of things.

What say you?

I seriously hope you did not misconstrue my posting about subjective opinions as being "poking holes in your view of things." I think you are one of the most literate, knowledgeable people on this forum, and I gain a fresh perspective from virtually every one of your postings. I may disagree with some of the things you write, as I am quite certain you disagree with some (or most) of the things I write. Fortunately for me, I have reached a position in life that agreeement is not a prerequisite for learning. I learn from the things you write whether or not I agree with them. I hope to continue doing so.