Last Words?

[Alydar in California]

David: I have exactly 30 seconds. The rest of my reply will come late tonight. For now, \"epistemology\" is the study of knowledge. \"Teleology\" is the study of design or purpose in natural phenomena, hence my Quixote comment. How the hell can someone who keeps talking about creationism and evolution not know that? What do you have to say for yourself, David?

[pmorgan2002]

Alydar, I\'ve since moved from Washington to California and can\'t wait to meet you at the track some day. Going forward, leave me out of your comments.

[Alydar in California]

Why? You want to beg for figures from me, too? Or do you have something else in mind? Do you remember the vulgar term you used to describe this board? Why not repeat it?

[dpatent]

Alydar,

You are 100% correct on the epistemology/teleology issue.  My mistake.  It has been 18 years since my \'Justice\' class with Sandel.

[pmorgan2002]

Because it would be a great pleasure to meet you. Though your definition of begging is much different than most, I do have something else in mind.

Unlike you, I do not memorize what other people on this board have to say. It seems far too unhealthy. I just move and keep my eyes and ears open. You\'ll notice I\'ve not said anything on this board until you brought up my name so, please, find someone else to talk to/about as I have only one use for you and it is not conversation.

[TGJB]

     Alydar did a good job with this, which saves me some work. And of course, again, it’s not really you I’m speaking to, since you have no real interest in getting to the truth—if you did you would have joined  the chorus asking Friedman to post the 13th (which they now have, and which I will be addressing shortly).
     
     Numbered points are mine, not yours, except where indicated.

1- I make no assumptions, other than that prior figures of horses can be used to determine future figures horses run, and outcomes of races. I do this because it’s the premise, and that it works out in practice. Because you don’t make figures, this point eludes you, but Alydar did a pretty good job of explaining it. You can’t assign artificial (incorrect) figures for horses running in different races, at different distances, and different tracks, to have them run in a tight range, and ALSO have them running in a tight range when they come out of those different races and run against each other. It’s physically impossible unless you screw around with the relationships WITHIN the race, and the “tightness” confirms both the numbers you are assigning today and they numbers they are based upon. Spend one week making figures and you’ll get it.

2- Ragozin was first, and has tons of dogmatic rules concerning sprints/routes, changing tracks, etc. I came along, disavowed that stuff, and only make the one assumption, above. Good luck convincing anybody I’m the creationist.

3- (your 1)- Yeah, Bayakoa probably won a FM graded stake with a 19. Do you ever actually listen to yourself?

4- (your 2)- Pure nonsense, and you know it. The point is not the percentage of variability, but of net effect on final time, and re-casting the argument is disingenuous.  The actual difference between variants was 3.6 points (not 5 or 6) for the grass races—a difference of right around 1% of the final time. That’s not a lot, and in this case we’re talking about a course that had been soaked by rain, and was now drying. Regardless of all underlying logic, however, everyone should look at the 5th and 7th races on Preakness day on Ragozin and TG, and draw their own conclusions.

5- (your 3)- Again, disingenuous, especially your granting “graded horses run better”. Not just better—more consistently near their tops. If you actually don’t know this, you’re a very bad handicapper.
    What I said (and BOTH statements are in the 5/24 post) was, “every horse, in a graded stake race, ran at least 6 points off their top”, and “every horse in the field but one ran at least 6 points worse than his previous race.” i.e., bounced. BOTH ARE TRUE ON RAGOZIN—everyone can look at Ragozin’s Schaefer sheets and see for themselves. Again, even if you think any one older stake horse is 50/50 to run 6 points off his top (or bounce 6 points, either way)—and 50% is an astronomically high number considering how close to their tops stake horses run—the chance of all 6 doing it is 2%. Rather than taking a position that can’t be proved (like saying each one had an 80% chance) why not try one that can—take me up on the bet I offered.

6- (your 5)- As has been documented here (my lawyers letter to Ragozin after we taped Ragozin employees lying about us) the “dissing” that takes place is almost all by them, in the field, in private, where we are almost never in a position to respond. The only reason they don’t do it here is because we can respond. And calling my raising questions of figure methodology “dissing” is diss-ingenuous. You know better.

[TGJB]

I\'ll leave most of this to Alydar, and I answered all that needs answering in another post. But to your last point--on the contrary. Because they used a different variant than the previous race (Preakness), they came close to getting it right.

[dpatent]

Actually, Jerry, they used the same variant.  See Friedman\'s post on the subject.

To your other points:

1) I have made figures before so I don\'t need a lecture on the subject.  Your first sentence reveals exactly what I have been saying all along -- that the past determines the future; and that\'s why you get into trouble messing with variants, making them faster, slower, faster to so that the \'tight ranges\' are not violated.

2) The dead horse is beaten.

3) Jerry, all I know is what the number was -- it was a 19.  It was early in the year 1990 or 1991 I think.  I\'m sure you think Ragozin got the number wrong, so what\'s the point of discussing this?

4) However you want to characterize the % change -- and I still believe that it is more meaningful to deal with realistic ranges -- I asked for some historical data to back up your adjustment.  You gave none.  Zippo.  Zilch.  And we both know why, don\'t we?

5) You are still failing to engage on the specific horses involved.  Yes, graded horses tend to run better, yes they tend to run more close to their tops.  But my point was, and is still unrefuted by you that the specific horses in the Schafer all figured to bounce and all but 1 of them (Bowman\'s Band) were not typical graded stakes horses but allowance-types who happened to be entered in a graded stake race whose particular patterns were awful going into the race.  Now, you can always claim that Ragozin\'s numbers in the prior races for these horses were all wrong, but if I look at the Rag. sheets and try to say what percent chance those horses had of running well, it was about 80/90% to run bad for most of them.  You arbitrarily pick 50/50 but 1) look at the percentage of good numbers most of those horses were running and 2) the % of the time they bounced off big efforts.  That will give you a clue that 50/50 was wildly optimistic for that bunch.  And even if the chance of all running bad was 2% that\'s still 1 race in 50, which happens 2-3 times a day in America.

6)  Fair enough.  However, I think it\'s fair to say that the content on this board is more confrontational to the competitor than vice versa, which I don\'t have a problem with in theory, though the substance of most of the posts is, shall we say, less than enlightened.

[JimP]

In the ongoing dissing contest, David wrote:
\'Your first sentence reveals exactly what I have been saying all along -- that the past determines the future; and that\'s why you get into trouble messing with variants,...\'

Hey guys, as best I can understand the TG and Rags processes, they both depend on the ASSUMPTION that past is predictive of the future. TG uses the \'past\' of the specific horses in a particular upcoming race to predict the time of that race and then builds a variant from that prediction. Rags uses the \'past\' of groups of similar horses running at the distance at the track to predict the time for that upcoming race and then builds the variant from that prediction. They both rely on the past to predict the future in much the same way. As I understand the major difference in approach is that TG adjusts the variant between races on a single card and Rags does not do that (except for rare instances). The reason that one adjusts and the other does not is based on unproven assumptions in each case. One assumes that the track and other conditions can change in unobserved ways and the other assumes it can\'t.

I think I have summed up the basics of the two views. Both views seem to be deeply held. Both views are based on assumptions. Both parties seem to have equal disdain for the approach of the other. Until one party or the other offers some factual basis for their assumptions, one has to choose which assumptions to accept on faith and then  decide with personal experience whether those assumptions lead to a product that works.

[TGJB]

They used the same variant (roughly) as the other sprints--not as the Preakness. See my reply to Friedman.

1) You don\'t read very carefully, or to be more precise, are very selective about choosing what points to reply to. Your posts say a lot more about you than they do about me or the subject.

2) You ain\'t kiddin\'.

3) Please think. Do you think it\'s likely she really won with a 19? If you blindly accept their figures as right, what\'s the point of discussing figure making at all?

4) Historical data on what? Until someone comes up with a machine to measure moisture content, rate of drying, and its effect on time of racehorses, the only way you can do it is by looking at the past figures of the horses. That\'s why we all do it this way. By the way, it looks like they actually have  the grass course getting SLOWER that day. What do you make of that?

5) It\'s not about running well or not (as you again fail to answer directly, and attempt to recast the argument), it\'s about their chances of going back 6 POINTS OR MORE, or RUNNING 6 POINTS WORSE THAN THEIR TOPS. And you know it. I repeat--take me up on my bet. In any event, now that you know they were sliding the variant, do you still think they should have tied it to the Preakness?

6) As I just said, in this forum they clam up--they don\'t want an open dialogue. As you have seen in the posts from the lunatic fringe (Jim, Howard Dennis, etc., who unlike the Ragozin office don\'t realize how this stuff plays) there is a lot of really poisonous stuff out there about us, and me. Guess where it came from? I\'ll give you a hint--who would have a motive?

[mandown]

Two questions, David:

(i) If figure-makers shouldn\'t use \'the past to determine the future,\' why should figure-users? If TG\'s approach is so wrong then more fool anyone who has ever bought the Sheets, DRF, ThoroGraph or any other horse-racing data-source. I presume that includes you?

(ii) If \'the past doesn\'t determine the future,\' then why did you \'look at the Rag. sheets and try to say what percent chance those horses had of running well\' and decide \'it was about 80/90% to run bad for most of them?\' I guess not all history is bunk, is it?

[dpatent]

Mandown:

I was a bit imprecise with my language re: the past and the future.  The difference I see b/wn Rag. and TG is that with Rag. the past gives you a probabilistic range of outcomes.  With TG, that range is highly constrained by a particular viewpoint about what a horse will or will not do on that day, and that view is constrained only by the mind of Jerry Brown and is not tied to any measurable data.

[dpatent]

Jerry,

So as not to be accused of ignoring your questions again:

4) I don\'t know what they did with the variant for race 10.  Friedman can address that.  If they had the turf getting slower, then that would be interesting, now wouldn\'t it?

5) All I can say is that when I look at the probable ranges I assigned to the horses in Race 11 I had everyone running between a 7 and 12 -- most of them a 9 or worse -- except Bowman\'s Band.  That put them all at 6 points or more below their top.  BB was a huge question mark for me. I figured he\'d bounce 2-5 points.  He bounced 6.  That kind of result doesn\'t make me want to tear up the variant.

[mandown]

David,

\'Imprecise with your language??!!\' - you just gotta be a politician.

And is there that much difference between a \'probalistic range of outcomes\' - another magnificent fence-sitting phrase that would do a politician proud - and a viewpoint? Are you saying Jerry has only one view? If so you haven\'t read his posts. The whole point of his approach is that he \'fits\' a race to reflect the \'probalistic range of outcomes.\'

Does Ragozin do any different? He says the track was drying out on Preakness Day and so they adjusted the variant. Was there \'measurable data\' for that (scientific analysis of a track sample?) and a rigid formula based on the measured moisture content - or was it a viewpoint based on the times recorded and the horses that recorded them? And whatever vague terms - \'a light use in exotics\' etc - Friedman might use in previewing a race as far as I\'m aware
they only put one figure into their database after the race, not a \'probalistic range of outcomes.\' In other words they form a viewpoint just the same as Jerry does.

And what\'s the difference between a viewpoint that sprint and route variants should be split and one that says they shouldn\'t? Forget the dogma. Logic says there\'s an argument for both. Which you subscribe to reflects only your opinion.

In your reply to Jerry you refer to how you used a \'probalistic range of outcomes\' based on your interpretation of the Sheets for Preakness Day Race 11. You formed the viewpoint(?) that Bowman\'s Band would bounce 2-5 points. Doubtless the fact that you were wrong - according to the Sheets he bounced 6 you say - in your mind proves Jerry is also wrong.

Now how about this for a thought? By your own admissions (in the HP challenge string) you don\'t get much time for handicapping these days. Possibly it was operator error? Jerry makes figures seven days a week, 52 weeks a year and has done so for the past 20 years. To equate your judgment with his would not, I suggest, encompass a logical \'probalistic range of outcomes.\'

Horses aren\'t machines and tracks aren\'t uniform. Figure-making is subjective. That\'s the whole point. To quote an old saw, it\'s the difference of opinion that makes horse-racing.

You may not subscribe to Jerry\'s methodology but why do you want to dissuade others? Surely the more people being \'misled\' the less the Ragozin product is being \'diluted?\' That is what you said, isn\'t it? And if they really would be happier with lower sales why do they censor all of Jerry\'s posts? But then are they a \'bunch of communists/socialists,\' to quote one of your posts, or \'Coke not talking about Pepsi\' to paraphrase another?

No wonder you have trouble with the concept of horses being consistent.

[dpatent1]

Mandown:

I agree with almost everything you say.  The big difference is that JB assigns a probability of 0% to a number of outcomes that I think have a significantly higher probability of occuring.

BTW, I\'m not a politician, though I used to be an attorney.  How\'s that for leading with my chin?

Finally, I am gratified to see that the level of discourse on these boards has improved dramatically in the last week.  Whereas before a high percentage of comments here were on the order of: \"Ragizin sux,\" now we are debating the meaning of tautology, teleology, and epistomology; and uncovering some interesting assumptions about figure-making.