Lots Of Stuff

[TGJB]

Back from Saratoga, and there is lots of interesting stuff here which I will touch on briefly for now. Much of it will provoke more discussion, in a good way.
As for Beavis (\"hehehe\")-- you are harmless, so for the meantime we\'re going to continue to let you post, save for unusually obnoxious personal attacks such as the one Paul tells me he deleted (I was away). The Ragozin office sent me a telegram telling me how proud they are to have you as their ambassador here. Your stuff is very effective, don\'t let anybody tell you otherwise-- everyone here is just as dumb and childish as... well, hehehe.

Weight-- I too would like to see the Beyer/Hopkins study. As I said at the Expo (this part they didn\'t cut on the DVD), the way the game is structured makes it very hard to do a study, because horse are assigned more weight when they are in form.

Ground loss-- all else aside (since there is no way to argue with \"I have done well by doing...\"), the issue brings up something else I went into at Vegas. If one horse beats another by 7 lengths, but he ran on the \"better\" part of the track, the 7 lengths could be considered misleading-- but we still want the DRF to tell us how many lengths the horse was beaten. Likewise, it is better to know how far each horse travelled-- if you want to adjust for \"better\" parts of the track, have fun.

Betting strategy-- as I have said before, the trick is to have your bets as closely as possible express your handicapping opinion. This is not nearly as easy or simple as it sounds, and the easiest way to do this is to actually ask yourself the question \"What do I like about this race?\", and articulate an answer in a full sentence. Try it-- if you actually can do it, you will find that the answer should pretty well define your bet. More to come on this, and if I have time I might do what jwo7 asked, for the Saturday Sar card-- not just because I had a big day, but because it was a result more of effective betting than of handicapping. The two things that have improved my game over the last few years are using the new data, and making more focused plays. Meanwhile, anyone interested should go to the betting pointers Alan mentioned.

TimeForm/performance figures/speed figures-- as George touched on, very few races in Europe are truly run-- many are the equivalent of our \"pace\" races. Since they don\'t give fractional times (which would be pretty much meaningless anyway, given the varying configurations), this means pretty much cutting loose every race.

Classhandicapper is right on two fronts-- one is that I do disagree with some of what he is saying. But only as to some of the details-- I do agree that various elements affect the final time (see \"pace\" races, which are extreme cases), especially when you are talking about small differences. We are trying to come up with figures that best reflect the performances of the horses, as we do with the \"pace\" races. As far as I know, I was the one who coined the term \"performance figures\" when I did Post Time in 92-93. But I could be wrong.

[JimP]

Now that you\'re back. Would you mind posting the numbers you assigned for the Arlington Million (at least the top few finishers)? I think seeing how the figures lined up for this one compared to what we had going in, would be very instructive.

[miff]

JB,If you agree there are sometimes better parts/paths of the track, then why don\'t you make an adjustment to the figure instead of assigning a misleading bias aided ground loss number. As far your statement regarding there\'s no way to argue about\"I did this and etc\"Why don\'t you track wide bias aided numbers.I\"ve done it for a long time.You\'ll be very surprised how poorly those figures stand up.

[TGJB]

I guess you didn\'t notice the qotation marks around the word \"better\".

1-- I don\'t believe it. I also don\'t disbelieve it. If it is true at all, I certainly don\'t think the difference in track speeds is extreme, or I would see it in the figures when I work with them-- as I do with dead rails.

2-- Horses don\'t stay in one path throughout a race. This alone makes adjusting the figures for paths a very dicey proposition.

3-- There is no more reason to believe that the relationship between paths stays constant throughout the day than that the track as a whole stays the same speed throughout the day. So even if horses were staying in the same path, you would often have to look at a sampling of one or two horses that ran in a specific path in a given race. Not enough.

4-- SO-- we only note it when it is really clear, which happens when there is a dead rail. And even then, we have found that you can\'t apply a set correction, since what seems to happen is that a lot of horses just don\'t fire. So we mark them, and let you figure out what to do with them.

[TGJB]

Re Arlington Million numbers-- As a matter of course we don\'t give out numbers that horses have run until they are entered back, keeping that information for our horseman clients. I make exceptions for Triple crown races and other situations where there is trmendous interest or we can point out something in particular.

What exactly are you looking for? Why this race?

[Boscar Obarra]

  JB, wasn\'t it many years before the dead rail concept became officially incorporated into your figures, even as a note. And does the competition even do that much these days?

  Of course, it\'s existence is probably as old as racing, and certainly part of common knowledge for 30-40 years, even among the slower witted players.

  As far as \'lanes\', it would be a gargantuan task to make a separate variant for multiple paths.  As JB has said, very small samples.

  When I was playing, I\'d note W+ to denote a wide aided trip, W- for wide hindered.

[Lucy]

\"As for Beavis (\"hehehe\")-- you are harmless, so for the meantime we\'re going to continue to let you post, save for unusually obnoxious personal attacks such as the one Paul tells me he deleted (I was away). The Ragozin office sent me a telegram telling me how proud they are to have you as their ambassador here. Your stuff is very effective, don\'t let anybody tell you otherwise-- everyone here is just as dumb and childish as... well, hehehe.\"


always nice to get top billing.

and I will have to apologize for the deleted post --- if I had known you were such a hothouse orchid, I never would have made that joke at your expense.

As someone that does try to incorporate bias into his analysis, IMO JB\'s summary couldn\'t possibly be more correct. IMO, it would be  a huge mistake to try to incorporate bias into his figures for all the reasons he stated.  

He does note the extremely dead rails. On those days you can look at the notes as to how wide the horse was. Then you can decide if you think a figure is a little suspect in terms of measuring the horse\'s actual performace.

However, all this stuff gets into very subjective evaluations of the horses\' abilities going into a race vs. how they ran in the race and why. 10 fairly competent guys are going to have 20 different opinions.

So things like bias are going to have to  remain a subjective area of the game. They can\'t be incorporated into a figure to the same degree of accuracy as time, ground lost and weight.

[miff]

Classhandicapper,

Bias may be more subjective, but a strong bias is far more relevant to a figure than ground loss or weight. So, you are left with the situation where a very important factor(bias) is not incorporated in the figure whereas lesser factors are.That makes for very weak numbers on severe bias days.

[JJP]

CH-

I do know of one handicapper who not only incorporates bias into his numbers but pace as well.  He uses those 2; not weight or wind.  And in all honesty, I believe pace and bias are far more significant than weight and wind.  It is time consuming; I\'ve tried doing it myself and it is difficult.  I do know he rates the track by \"path\"; (i.e. 3 path track, 4 path etc.)  I know one thing: judging by his results, its pretty effective.

I agree that pace and bias are hugely important.

I just believe it is an error to try to incorporate them into a figure. I think it would be even worse if JB sold a product like that. IMO, he is doing the right thing. The only thing I would ask for is a seperate number without ground loss so I don\'t have to back it out and perhaps a few more pace comments.

I have also dabbled with adjusting speed figures with pace/final time formulas. I\'ve also tried to quantify bias in terms of lengths of impact etc.. I\'ve pretty much stopped.

Making pace figures that are as accurate as JB\'s speed figures is close to impossible. If you think you can, you probably don\'t understand all the problems. Second, IMO there is no formula that works for all horses. IMO, the impact of pace is dependent on things like stamina, brilliance and other individual qualities. So when you combine a \"reasonably accurate\" speed figure with a \"somewhat inaccurate\" pace figure and put them into a \"somewhat inaccurate\" formula you often get a mess. Same with bias.

IMO, what you want to do is look at the horse\'s figures and know the trips.

Using Beyers in an extreme example:

If a horse ran 100, 97, 101, 98, and 77  in his last 5 races and the 77 was in a very fast paced duel on a deal rail, you really don\'t have much of problem here. You can view him as a high 90s/100 horse.  If you plug slightly innacurate pace and bias figures into a slightly inaccurate formula and all the errors are in the same direction, you might wind up with 90 or 110. IMO, that\'s worse.

I rate paces as neutral, fast, very fast, slow, or very slow (paying attention to specific 1/4s and using observation as well at fractions). Rate the bias as none, strong, very strong etc... and then get a \"feel\" for how a horse ran based on all evidence available.    

Granted, not all horses have such clear and consistent figures. However, IMO, adding complexity to an evaluation with \"suspect\" formulas etc... is not a good solution. It may be better than ignoring the factor, but it is not a solution.



Post Edited (08-25-04 12:07)

[miff]

CH,

Generally,I agree overall. A new product re pace figs only is out, it\'s better than anything I\'ve ever seen.I wish TGJB would get more involved in pace figs. Many of the factors you allude to are surely within the TG data base to make pace figs.

[JimP]

To TGJB re my Arlington Million question: I\'ve probably pushed this enough. I understand why you wouldn\'t want to publish the numbers. I just thought the race was interesting from the standpoint of Euro figs. Really I was only looking for an assessment of the running. Do you think the Euro figures held up? If so, it appeared to me that the US horses must have Xed for the most part. If the US contingent ran to prior form, then several of the Euros must have topped. I was having trouble aligning the performances relaive to prior form. But that was just my non-expert view. My interest was in hearing how you saw it. Maybe you can address that without publishing the figures.