On the Ragozin site, Michael mentioned seeing Mark Hopkins talk on the OTB channel about the track changing speed 8 Beyer points because of track work, and asked Friedman \"Do you make variant adjustments during the day due to maintenance work\"?
Len\'s reply:
\"If your question is whether track work immediately before or during the card ever produces changes in the variant level from race to race the answer is yes, although in the vast, vast majority of cases the track work is weather related and/or predictable as general maintenance at a particular track. I\'m just guessing, but I think that Hopkins may have been referring to the 2yof stakes opening day at the Spa which went dramatically on the slow side. Even making the kind of adjustment he apparently did would not give those horses anything near what would have been expected (and I\'m still not giving back any of what I cashed on that race) and there\'s no reason not to just give them the numbers that their final time earned. Changing the variant level because the horses in a particular race didn\'t run as you expected is a subjectivity that places the ego of the variant maker over the reality of what the horses actually did\".
Amazing how much there is to work with in a single paragraph. Where to begin:
1-- Those times that the track work produces \"changes in the variant level\"-- how do you know it? Do you have a machine that measures \"physical resiliencey\", as you once put it?
Because if not, you are making the decision the way Hopkins, Beyer, myself and all other figure makers with some relationship to reality do it-- by looking at how fast the horses run relative to their figure histories. And if you are adjusting the \"variant\" based on that judgement, aren\'t you placing your ego over the reality of what the horses \"actually did\"?
2-- \"In the vast majority of cases the track work is weather related and/or predictable as general maintenance practice at a particular track\".
a) Really? For you to know that is true, you would have to be getting complete information about track maintenance at all tracks. Do you? If there are tracks where you are not getting information, do you assume it is being done, or not being done? Do your trackmen report to you-- as ours do wherever possible-- when the track is watered? I ask because there are some tracks where they water the track before every race on some days, and not at all on other days. All permutations of watering take place-- before some races but not others, etc., and I\'m not even getting into the issues that I went into in \"Changing Track Speeds\" (Archives this site) about varying quantities of water, different climatic conditions, etc.
b) Even if you were getting that info, you would still have to correlate the track work with the track speeds. How would you do this without applying your own judgement, as above?
c) \"Predictable\". Boy, is that an interesting word in this context. Please explain what that means, and how the fact that work on the track is \"predictable\" in a certain instance can be applied to making a figure for a race or day. Do you make an assumption, for example, that because specific work was done, the track will be getting faster, and add points as the day goes along, without regard to the results and times of the races?
3-- On the Saratoga filly stake-- that\'s a GREAT example, I think I might have discussed it here already. Races 2 through 5 were dirt races, and after them the main track was sealed. It was reopened after the two grass races, but was noticeably slower (like 6 points) for the final TWO dirt races. This was an example of knowing something had been done to the track-- but even if I had not, I would have had no problem dealing with it because there were 2 races, which made it clear.
Tell you what-- why don\'t you post sheets for the last two dirt races on the day. I\'m going to make a prediction, like I did after the 2001 Breeders Cup (and those who were around then remember how that turned out)-- if you did those races with the rest of the day, at least 75% of the horses in those two races (excluding the winner of the eighth, who ran out of his mind and figures to bounce) will go forward at least 3 points next time out, on YOUR figures.
4-- \"...the numbers that their final time earned\", and \"...the reality of what the horses ACTUALLY DID\" (emphasis added).
Really, what they ACTUALLY did, and earned. How do you know what they actually did? I mean, basing it on the horses would be an excercise in ego, so you must have another way.
Here\'s what you really mean, Len. You mean, make an assumption-- based on absolutely nothing, and one that goes against the science-- that the track stays the same speed. You mean that the right answer is to use a broad average of how fast the track is during the day, and apply it to individual races where it may or may not have been that \"speed\".
In re: Indulto\'s flattering request for my input -- here goes (and I hate myself for this because I am going to throw a tidbit Classhandicapper\'s way).
There is no question that each set of figures reflect certain judgments on the part of the figure maker and his trackmen. I am not intending to be sexist here -- if there are female figure makers and/or trackwomen out there, I would like to meet you and apologize in person.
Beyer\'s figures (which I ignore) reflect the judgment that weight and ground are better evaluated by the individual doing the handicapping rather than by the figure maker.
Thorograph and Ragozin -- the two best products I am aware of -- already reflect in their figures certain areas where the figure makers are uncertain about the data or the consequences of the data.
Thorograph uses boxes where a figure cannot be made and also uses \"?\" when there are certain types of doubt (although based on recent posts, I am now unsure about \"time?\" -- this either means, (a) there is doubt as to the accurate timing of the race and we used the time we judge to be the most accurate or (b) the time we used is quite a bit different from the official time, but we know that our time is right and the official time is wrong. I understand TGJB to be saying that the symbol means (b), but if that is the case, I am unsure as to why this needs to be signified).
Ragozin will use the \"~\" in certain cases to reflect various types of uncertainty. Also, I, personally, take all numbers which reflect a track condition of \"..\" or some of the other unusual track conditions symbols ragozin uses with an extra grain of salt.
We also know from last year\'s Breeders Cup that Off Poorly symbols or trouble symbols can be a big deal when it comes to judgments affecting the evaluation of a horse\'s performance. The pace notations (slow, fest, etc) are of the same ilk. All of these are items which could have affected a figure, but the figure maker is saying be aware that the figure may not encapsulate everything relevant.
Now for the issue before us -- numbers which are potentially uncertain as a result of the variant. TGJB has admitted here that there are occasions where a variant was particularly difficult to make and there are rare cases where he will go back and change numbers already awarded because with more information he is able to better assess the variant. I do not know how Ragozin handles such a situation because Ragozin (or Robespierre) does not entertain much discussion of the arcana of variant making. I really do not know if Ragozin would ever say a variant was difficult to make or whether retroactively evaluating variants is something that should be done from time to time.
I understand people having concerns about retroactively making variants and the risks of creating a certain circularity in the numbers. However, I also recognize that there are cases where a variant is extremely difficult to make and a judgment can be wrong no matter how good a figure maker\'s judgment is. I have confidence that TGJB is very aware of these issues and uses his best judgement in this area. I also trust his judgment and note that he is willing to expose his judgment to criticism here, in the marketplace of ideas, and engage in constructive dialog about his figure making which to me gives his figures a certain credibility. As I have mentioned previously, TGJB has a track record of responding well to people who bring to his attention errors in his figures and he will admit a mistake and correct it.
Now, I am not interested in knowing whether a particular race was \"cut loose\" or for some reason the variant differs from other variants used in the same day. If the figure maker uses a different variant and is confident in it, that is good enough for me. When I buy figures, I do not want all the raw data that went into them. What I am paying the figure maker for is to take all the chaotic data, use his judgment, and give me something that reasonably encapsulates the relevant information to evaluate how a horse performed on a given day. If he feels good about it, then so do I. In other words, I do not want to know what is in the sausage or how it was made, all I want to know is that it is delicious and safe to eat.
So, I think CH\'s general idea of informing us about the intricacies of how the particular variant used was arrived at, is gruesome overkill and completely unnecessary (and quite frankly would overload me with information that is either irrelevant or misleading if the final figure is good). However, if a figure maker had a tough time coming up with a particular variant, and he believes there is a 50% or better chance that such instance may end up being one of the rare times that a variant might need to be changed retroactively, then I think some sort of a symbol would be helpful (maybe a v?). I am saying this hoping that it is really a rare case. A sheet full of \"v?\"s would not be very helpful, but I would understand a few sprinkled here or there (and ideally, over time, each \"v?\" would be eliminated one way or the other).
If Indulto is asking about input on the ins and outs of each figure maker\'s variant making, I am not the right person to ask. I really have no special insight. I will say that one very interesting case of differences between the two sets of sheets came up in the context of Musique Toujours. From the sheet reading perspective, I already wrote a detailed analysis comparing the ragozin sheet and the thorograph sheet leading up to the race in which he was claimed. If somebody wanted to get into why the sheets were so different for the same horse, a person who knows more about figuremaking could go behind each figure and determine what went on in the figuremaking. Again, I am not the right person to do that.
SoCal,
I think you are overstating my desire a bit.
I would simply expand the handful of \"V\" notations that you seem comfortable with, with a handful of others where the variant didn\'t generally fit with the rest of the day (a huge break out) or where the figure was based on a single race (like the only 2 turn race or something like that). That would be true even if the figure maker was fairly confident in the figure assigned.
There are \"off topic\" reasons for that.
I think it\'s best that I do not elaborate much.
IMO (and ITO of many others), within the group of races that comes up slower/faster than expected, sometimes the reason is not a change in track speed nor is it a pace so EXTREME that anyone could notice.
Applying a seperate variant in those cases could produce a mildly undesireable result even if the figures seem to make sense given the typical general ability of the horses.
It certainly wouldn\'t involve \"V\"s all over the place.
Plus, I don\'t want to be a pain in the butt about this. It\'s just a suggestion. It\'s not a criticism.
To: CH
I disagree with you.
SCM2
SoCal,
I think if I explainded why I consider those extra \"V\"s important you would agree that they would have some value. It\'s just not a good idea to have that long a conversation here.
CH
I disagree with your first sentence but agree with the second sentence.
This response is based on the 4th edition (original as edited 3 times) of your first post this morning.
SCM2
SCM,
>This response is based on the 4th edition (original as edited 3 times) of your first post this morning. <
I\'m learning to be more careful about how I word things in order to avoid confrontation. Agree to disagree.
I think that it\'s pretty amazing that we are now up to about 20 posts between the two boards in response to my post, and not one has dealt with the subject matter. I get that nobody here feels the need to-- most of you have heard it before, and get it already. But this began with a statement of Friedman\'s, which I disassembled-- and with all the verbiage since, nobody has dealt with the issues, or pressed Len to defend his positions or even explain the \"holes\" in mine that would enable him to have his at all.
One man\'s \"savaging\" is another man\'s \"marketplace of ideas\".
Sorry, but no man can take credit for the \"marketplace of ideas\" -- I got that from the \"Yankee from Olympus\"
I feel I needed to \'fess up before anybody called me on the carpet on this one.
For all of you who did not go to law school, these links give the background.
http://en.wikipedia.org/wiki/Marketplace_of_ideas
http://fact.trib.com/1st.lev.internet.marketplace.html
TGJB,
I have an opinion on what Len said, your response, and on the \"greater issue itself\". If you\'d like to hear it, I\'d be happy to express it, but at this point I\'d rather just lobby for what might help me as an occasional customer of both camps.
If it is really on point-- regarding the question of track maintenance and figure making-- go ahead.
TGJB,
If I\'m off the specifics, I\'m sure you\'ll delete it. :)
a. If a figure maker is willing to concede that the track can change speeds from time to time under extreme circumstances (rain, drying out) AND because of daily track maintenance (both camps believe this if I am understanding their stated positions correctly) I don\'t see why one wouldn\'t use subjective skills (looking at the horses prior figures etc....) in \"both\" cases to obtain an accurate variant when a race doesn\'t seem to fit with the day. It doesn\'t make sense to do it for one and not the other. Of course, you would need to actually know there was rain/track maintenance. It would also be helpful to know how the track tends to react to \"x\" weather/maintenance. I see no major difference between extreme weather situations and track maintenance other than the degree of potential track speed change.
b. IMHO, a variant based on the average of several races would tend to \"limit the degree\" of occasional error. Figures being used as input to a race may not be perfect and interpretation errors are always possible. Larger subjective errors resulting from either of those would/could then be passed on to future races. Therefore, the figure maker should be \"very confident\" that the track changed speeds before breaking a race out. (Just ask Andy Beyer about all the high profile figures he has blown by breaking races out).
I\'m going back to my poker game before the bombs start getting lobbed.
CH-- you don\'t necessarily need to know about the track work, you do need to not make the assumption that nothing was done if you don\'t know.
This is a good time for everyone to take a look at the Expo presentation, even if you have done so before. You will see why you have to be careful of adding to increase sample size-- it\'s only a good idea if it\'s apples and apples.
http://www.thorograph.com/archive/files/multimedia/Vegas2004/Vegas2004lo.ram
By the way, one of the points that some people don\'t understand is that unlike the early figure makers who used pars (and therefore only the winners), those using the projection method use all the horses in a race, and therefore are increasing the sample size from just one.
We get as much information about track maintenance as we can-- almost certainly a lot more than Ragozin (like, when they water the track, aside from sealing/harrowing etc). But I still make as few assumptions about cause and effect as possible-- it\'s all just informational grist for the figure making mill.
TGJB,
\"CH-- you don\'t necessarily need to know about the track work, you do need to not make the assumption that nothing was done if you don\'t know.\"
I understand the distinction.
I\'m heading up to Saratoga Saturday for the Whitney and Test. I\'m on the bit. :)
SCM2,
Thanks for the intellectual integrity in identifying the source of the marketplace phrase(and, that was an interesting site). You\'re in a peculiar place to be doing that. I always think of the 50\'s Russia/US media battles in which one of the jokes was, \"Russians invent televinsky\" in the PR battles to show national superiority. I don\'t know, maybe they did, but it seemed they were forever trying to establish some intellectual property for whatever the popular invention of the day was. I suppose to do any less than you did is to be no more than a cold war propagandist. raz
JB is up in Saratoga this weekend and I read him Robespierre\'s musing. He found it amusing and will respond when he returns.
This should get interesting.
The last time this got stirred up Marc @ DRF got banned for life from the Board. Alydar in Calif was tossed off also. I later was given a two week suspension.
The Ragozin Board software was changed to stop Graphite hackers from asking attacking questions. Jake H started issuing threats at TGJB on both the Ragozin Board and this one too.
After reading Lens longwinded attack, an obvious blindsided ambush, while the TG general was away for some R & R. This one has the potential to go nuclear.
Okay, first things first--
I would suggest to anyone who is going to take the time to read this that they first read the top of this string, where I quote Friedman\'s original post, and reply to it. Then you should go to the Ragozin board and read his 8/5 2:51 post \"Re RE RE Question for Len\", which went up while I was gone. I\'ll deal with that one quickly, if I have time I\'ll deal with the long boring one at a later time.
1-- First of all, you will note that in my post I asked several questions, none of which Len addressed. No surprise there.
2-- \"...the truth remains that none of this can be settled by clever argument-- the only test is whether the numbers produce predictable results and that test takes place at the betting windows and not on BBs...\"
a) Hence my repeated offers to have a handicapping contest over a long period of time with Len, and hence his declining repeatedly.
b) Hence my prediction of what the horses will do next time out-- it\'s a test, just like you said, and if I\'m right you will be able to see the results.
c) Actually, you can settle some things by debate-- that\'s why people do it. Like whether an approach makes logical sense, and whether a position is defensible or just dogma-- that\'s what my QUESTIONS were all about. And all \"blathering\" aside, Len talked an awful lot without answering them. Again.
d) Len, please-- you keep talking about \"success at the windows\". Our guys have done astonishingly well in handicapping contests, and have won the last 3 NTRA National Championships. What the hell are you talking about? My guys are creaming yours in the only game where the the results are made public.
3-- You gotta love Len\'s Derek Jeter analogy. Except I wasn\'t talking about just one \"player\", I was talking about a group, and batting averages are purely statistical. There\'s no judgement-- UNLESS the conditions change.
Len, pay attention-- here\'s the correct baseball analogy:
A group of 10 players (not just one) plays in Colorado, and are traded to other teams. I say, conditions in Colorado are such that batting averages are inflated, I predict that at least 7 will hit at least 30 points worse next season. Get it? I am saying you need to apply a different variant to Colorado than Shea, for example. (And if you don\'t believe me, talk to Bill James).
My prediction was that 75% of the horses in the last two races on the day in question (with the exception of one horse I specified) would go forward at least 3 points, which is an awful lot, and an awful high percentage, ON RAGOZIN. I felt pretty safe, because if they did those races with the others, as Len says, they got them wrong by a lot more than that.
But tell you what, Len-- since you are basically taking the position that ANY horse who runs badly is more likely to go forward, let\'s do this-- lets also look at the horses in the first four dirt races that day that ran at least 3 points off their top, and compare. I will bet you that at least TWICE AS HIGH a percentage of the ones who did it in the last 2 races go forward at least 3 points on yours. That\'s absolutely apples to apples, and if you are right, they should not. Right?
4-- Tell you what-- let\'s both respect the intelligence of our customers. I\'m going to post the dirt races from the day in question, with the numers they ran. The last two took place after they sealed the track, and I used about a six point different variant than for the earlier races.
Len, why don\'t you do the same-- post the sheets for the day, with the numbers they ran, and let everyone take a look at just what we\'re talking about. And not to sound like John McEnroe, why don\'t you just answer the damn questions?
Okay, first things first--
I would suggest to anyone who is going to take the time to read this that they first read the top of this string, where I quoted Friedman\'s original post, and replied to it. Then you should go to the Ragozin board and read his 8/5 2:51 post \"Re RE RE Question for Len\", which went up while I was gone. I\'ll deal with that one quickly, if I have time I\'ll deal with the long boring one at a later time.
1-- First of all, you will note that in my post I asked several questions, none of which Len addressed. No surprise there.
2-- \"...the truth remains that none of this can be settled by clever argument-- the only test is whether the numbers produce predictable results and that test takes place at the betting windows and not on BBs...\"
a) Hence my repeated offers to have a handicapping contest over a long period of time with Len, and hence his declining repeatedly.
b) Hence my prediction of what the horses will do next time out-- it\'s a test, just like Len said, and if I\'m right you will be able to see the results.
c) Actually, you can settle some things by debate-- that\'s why people do it. Like whether an approach makes logical sense, and whether a position is defensible, or just dogma-- that\'s what my QUESTIONS were all about. And all \"blathering\" aside, Len talked an awful lot without answering them. Again.
d) Len, please-- you keep talking about \"success at the windows\". Our guys have done astonishingly well in handicapping contests, and have won the last 3 NTRA National Championships. What the hell are you talking about? My guys are creaming yours in the only game where the the results are made public.
3-- You gotta love Len\'s Derek Jeter analogy. Except I wasn\'t talking about just one \"player\", I was talking about a group, and batting averages are purely statistical. There\'s no judgement involved-- UNLESS the conditions change.
Len, pay attention-- here\'s the correct baseball analogy:
A group of 10 players (not just one) plays in Colorado, and are traded to other teams. I say, conditions in Colorado are such that batting averages are inflated, I predict that at least seven of the 10 will hit at least 30 points worse next season. Get it? I am saying you need to apply a different variant to Colorado than Shea, for example. (And if you don\'t believe me, talk to Bill James).
My prediction was that 75% of the horses in the last two races on the day in question (with the exception of one horse I specified) would go forward at least 3 points next time out ON RAGOZIN, which is an awful lot, and an awful high percentage. I felt pretty safe, because if they did those races with the others, as Len says, they got them wrong by a lot more than that.
But tell you what, Len-- since you are basically taking the position that ANY horse that runs badly is likely to go forward next time (\"regress to the norm\"), let\'s do this-- lets also look at all the horses in the FIRST FOUR dirt races that day that ALSO ran at least 3 points off their tops, and compare the two groups. I will bet you that at least TWICE AS HIGH a percentage of the ones that ran in the last 2 races go forward at least 3 points on your figures. That\'s absolutely apples to apples, and if you are right, they should not. Right?
4-- Tell you what-- let\'s both respect the intelligence of our customers. I\'m going to post the dirt races from the day in question, with the numers they ran. The last two took place after they sealed the track, and I used about a six point different variant than for the earlier races.
Len, why don\'t you do the same-- post the sheets for the day, with the numbers they ran, and let everyone take a look at just what we\'re talking about. And, not to sound like John McEnroe, why don\'t you just answer the damn questions?