Hey, I Gotta Coupla Questions

[TGJB]

As Alydar and I have explained, you can\'t artificially make figures come out in a tight range without either retroactively changing earlier figures, or screwing with the relationships within races. So, I was thinking--Friedman and Ragozin are on record as saying they sometimes go back to change an earlier figure (sometimes a whole day) based upon what the horses do later. Now,

1. I\'m on record as being against it--I can only recall doing it twice in 10 years or so. As I have mentioned before, it becomes a self fulfilling prophecy.

2. First question--what\'s the \"scientific basis\" for the change or the decision in general to base the figures on later figures?

3. What\'s the reason for doing it? Okay, I\'ll help with that one--to get it right, as defined by HAVING THE FIGURES FIT WITH OTHER FIGURES THE HORSES HAVE RUN. In other words, to make the figures fit in a tight range, like we have. The measure of accuracy for all figure makers (including Ragozin) is that tight range--if your figures are dovetailing, you know with confidence you\'re there, and even if you blow an individual race or day (we all do) the overall accuracy of your database will keep you from going off the rails (losing a circuit vs. other circuits, routes vs. sprints etc.).

Listen--all these rules Ragozin has? He didn\'t have them when I was there, when he did all the hard numbers himself--he used his judgement. He instituted rules based on averages (or averages of his judgement, which is even wilder) when other people began doing most of the hard numbers, because he didn\'t trust their judgement (or anyone else\'s). Which is not to say that even in the old days he didn\'t tie together independent events, he did--but no \"rules\", no \"science\".

[nunzio]

With respect to point 1:

Did you ever see how often the Beyers
change in the form ?  They change figures so often that its rediculous.  Thats why I can\'t understand how anyone could take them seriously.

Nunzio

[JRL]

We keep mixing up \"science\" and \"statistics\".  Science, as I am using it, consists of objectively verifiable facts.  A horse going x wide means that it ran x further distance.  A wind of x speed will impede the progress of horse weighing y pounds by z amount.  A horse carrying x pounds is impeded by y amount.  Statistics is a tool of science to take past outcomes and predict future outcomes.  By definition, such analysis cannot be proved \"correct\" it can only be proved to be accurate withing a certain range of confidence.  The more data you have, the more confident you can be.  

Now, nobody makes a variant based upon science.  That is because it is impossible to do so.  You can only make assumptions based on science and engage in a predictive statisitical analsyis.

To the extent that I believe more in Ragozin\'s method than yours, it is that I believe their statistical analysis and assumptions are more rigorously derived. Their analysis is based upon more data points and is not adjusted for any \"anomoly.\"

Based on what you have said, you really do not make a track variant (at least sometimes).  By a track variant, I mean that based on a statistical analysis of the races run that day, compared with past races, you come up with a number that normalizes any figures to some predtertimed baseline.  That variant is than plugged into an algorithm with the raw number to get an adjusted number.  If that variant is to change during the day, which should be rare, it should do so in a consistent manner.

You seem to think that it is unnecessary (at least sometimes), to even have a variant.  That is because you look at the horses who ran in that particular race and determine what number they are most likely to run and then use that baseline to adjust the numbers in that race based on position. When this does not conform to what a consistent variant would tell you, you basically make the number consistent by changing each race based on your \"judgment.\"  You then explain it by variations in the track, as that is the only way it could be explained, therefore it must be true. Math and statistics are thrown out the window.  My quibble with you, is that I do not believe that to be an accurate way of determining numbers in the long run.  In a particular race or a particular day, it certainly may be more accurate, particularly where the statisitical analysis is compromised by limited data (this is one reason why you always attack Ragozin on Triple Crown, Triple Crown prep and BC days because the normal claiming races are not being run and you have races being run at distances that are not normally run at the track, thus the data is usually limited on these days).

As to your issue of changing numbers, I really do not see a problem.  All that is doing is recognizing that based on future events, the statistical analysis now shows that they were probably wrong.  Again, they were probably wrong because of the limited available data, thus more data helps them get it correct.  If it happened a lot, it would concern me because it would demonstrate that they were wrong a lot.  When it happens once in a while, it confirms that the number maker is being honest and rigorous.

As for the tight ranges, the entire debate is a red herring.  By definition, if you get rid of \"anomolies\" you are going to have a smoother pattern.  If you want me to find a professor of statistics to prove this, I will.  Now your point is that, if you were wrong, eventually your numbers would fall apart because the \"anomolies\" would turn out to be true and your future numbers would drift back toward the \"amomoly.\"  David\'s point is, maybe that is true, but there would have to be a lot of anomolies and even then, it would take a long time to surface.  

Now, saying that you got one race, the Peter Pan to provide a smooth pattern for all horses does not prove a damn thing about your methodology. At most, it proves you got the past numbers of those particular horses correct, and given that most of those horses had only run a few races, it doesn\'t even do that.  If you turned over all of your sheets over the past decade, identified those races where you adjusted for \"anomolies\" and then had an expert statistician determine whether these number hold up, then maybe I would give your theory some credit.

[nunzio]

Jason,

If you beleive the track speed remains constant through the course of the day,
I beleive you are mistaken.  Take
the time to monitor (at least visually) what the track looks like and its consistency from race to race.  Then watch all the maintenance that gets done.  Do you think changes in temperature and or wind might play a role
in the track speed.  I\'m not personally
involved in the rag-tg debate but the one thing that i am certain on is that racing surfaces do not remain at constant speed
as much as you\'d inititally think.  

when i read that ragozin actually makes variants hold up over multiple days
i found that to be unbeleivable.  it just
cannot be.

Nunzio

[TGJB]

Jason L. wrote:
>
> We keep mixing up \"science\" and \"statistics\".  Science, as I
> am using it, consists of objectively verifiable facts.  A
> horse going x wide means that it ran x further distance.  A
> wind of x speed will impede the progress of horse weighing y
> pounds by z amount.  A horse carrying x pounds is impeded by
> y amount.  Statistics is a tool of science to take past
> outcomes and predict future outcomes.  By definition, such
> analysis cannot be proved \"correct\" it can only be proved to
> be accurate withing a certain range of confidence.  The more
> data you have, the more confident you can be.
>
> Now, nobody makes a variant based upon science.  That is
> because it is impossible to do so.  You can only make
> assumptions based on science and engage in a predictive
> statisitical analsyis.
>
> To the extent that I believe more in Ragozin\'s method than
> yours, it is that I believe their statistical analysis and
> assumptions are more rigorously derived. Their analysis is
> based upon more data points and is not adjusted for any
> \"anomoly.\"
>
> Based on what you have said, you really do not make a track
> variant (at least sometimes).  By a track variant, I mean
> that based on a statistical analysis of the races run that
> day, compared with past races, you come up with a number that
> normalizes any figures to some predtertimed baseline.  That
> variant is than plugged into an algorithm with the raw number
> to get an adjusted number.  If that variant is to change
> during the day, which should be rare, it should do so in a
> consistent manner.
>
> You seem to think that it is unnecessary (at least
> sometimes), to even have a variant.  That is because you look
> at the horses who ran in that particular race and determine
> what number they are most likely to run and then use that
> baseline to adjust the numbers in that race based on
> position. When this does not conform to what a consistent
> variant would tell you, you basically make the number
> consistent by changing each race based on your \"judgment.\"
> You then explain it by variations in the track, as that is
> the only way it could be explained, therefore it must be
> true. Math and statistics are thrown out the window.  My
> quibble with you, is that I do not believe that to be an
> accurate way of determining numbers in the long run.  In a
> particular race or a particular day, it certainly may be more
> accurate, particularly where the statisitical analysis is
> compromised by limited data (this is one reason why you
> always attack Ragozin on Triple Crown, Triple Crown prep and
> BC days because the normal claiming races are not being run
> and you have races being run at distances that are not
> normally run at the track, thus the data is usually limited
> on these days).
>
> As to your issue of changing numbers, I really do not see a
> problem.  All that is doing is recognizing that based on
> future events, the statistical analysis now shows that they
> were probably wrong.  Again, they were probably wrong because
> of the limited available data, thus more data helps them get
> it correct.  If it happened a lot, it would concern me
> because it would demonstrate that they were wrong a lot.
> When it happens once in a while, it confirms that the number
> maker is being honest and rigorous.
>
> As for the tight ranges, the entire debate is a red herring.
> By definition, if you get rid of \"anomolies\" you are going to
> have a smoother pattern.  If you want me to find a professor
> of statistics to prove this, I will.  Now your point is that,
> if you were wrong, eventually your numbers would fall apart
> because the \"anomolies\" would turn out to be true and your
> future numbers would drift back toward the \"amomoly.\"
> David\'s point is, maybe that is true, but there would have to
> be a lot of anomolies and even then, it would take a long
> time to surface.
>
> Now, saying that you got one race, the Peter Pan to provide a
> smooth pattern for all horses does not prove a damn thing
> about your methodology. At most, it proves you got the past
> numbers of those particular horses correct, and given that
> most of those horses had only run a few races, it doesn\'t
> even do that.  If you turned over all of your sheets over the
> past decade, identified those races where you adjusted for
> \"anomolies\" and then had an expert statistician determine
> whether these number hold up, then maybe I would give your
> theory some credit.

TG--Jason, I will give you a detailed response to this, but I probably won\'t get to it before Sunday (at the earliest). Alyday, if you want to do some of the heavy lifting, be my guest.

[Alydar in California]

Jason wrote: \"The more data you have, the more confident you can be.\"

That depends on how accurate it is and how well you interpret it. Examples:

1: TG and Rags both estimate the wind. Since TG cuts loose more races than Rags, Ragozin relies more on wind estimates. If you\'re doing a race in isolation, wind doesn\'t matter. In his book, Ragozin says he adjusts the raw times based on the wind. Wind estimates are based on observations from people who are a good distance from the horses. Observers watch the flags right before the race goes off. Then they watch the race and hope the wind doesn\'t change. Does anyone really trust this process? (Has it changed?) Wouldn\'t you prefer to minimize your reliance on it?

2: Ragozin makes use of claiming race averages when trying to determine how fast horses were EXPECTED to run today. Do you trust these averages?

3: Ragozin relies more heavily on time charts, which are based on averages. At SA, 109.60 might equal 143.00. These times enable him to compare horses who run at different distances. But relationships between distances can change from meet to meet. They can also change within a meet, from day to day, and within a day. When these relationships change, things can get treacherous. Wouldn\'t you prefer to minimize their use?

4: Ragozin ignores bad rails when making variants. TG takes them into consideration. Which way do you prefer?

5: Go to the archives and read Bruno De Julio\'s comments about work done on tracks and how this changes their speed. Also, go to the Del Mar site and read the interview with Steve Wood, who is the track superintendent at DM and SA. Read about what he does to the track before AND between the races.

Jason wrote: \"By definition, if you get rid of \'anomolies\' you are going to have a smoother pattern.\"

No one is arguing that you won\'t have a smoother pattern today. The problem is that these horses, who will have INACCURATE numbers for their last race, will run back against different competition. And when they do, the man making the figures is constrained by the results charts. Patterns that are artificially smooth will become rougher because your figures will show slower horses beating faster horses. When this happens, you have no choice but to give the winner a better number, which will tend to make the pattern rougher.

Jason wrote: \"David\'s point is, maybe that is true, but there would have to be a lot of anomolies [before your numbers fall apart].\"

Any sentence that begins with \"David\'s point is\" is in serious trouble. See above. No one is arguing that a bad number now and then will make the figures fall apart. Patent threw that nonsense in there after he decided he had to back off his TG-has-no-use-to-him BS. But keep in mind that a bad number for one horse means it was a bad number for EVERY horse in that race. And bad numbers stay on the sheet and can lead to more bad numbers.

[TGJB]

Good work. I\'m going to be posting a lengthy response to Jason this afternoon, and obviously there is some overlap with your response. I\'m going to create a separate string so we can point people to it in the future.

I had not seen that Steve Wood interview before, and there is some very good stuff in there--effect of tides, the chute getting different maintenance, that it\'s hard to keep moisture content constant throughout the day, and the effect of shade on moisture at different parts of the track.

Thanks for pointing it out.

[JRL]

Yes, I believe that averages used correctly are the most accurate way of determining variants.  That is the core of this debate.  If you don\'t believe in them, then you are only left with the TG methodology, so obviously you should use TG.

I think of \"bad rails\" the same way I think of changing track speeds on the same day, they happen, but nearly as often as people think.

I can\'t argue that any of these \"anomolies\" don\'t exist.  But if I take seriously everything you say, then I would buy nobody\'s sheets because they really just made up.  The real question is how often does this stuff happen and how do you know it happened.  When JB adjusts for anomolies, I would guess that sometimes he is right, sometimes he is wrong, and the same with Ragozin.

One inaccurate number for all horses in a race may or may not show up through examination. It depends on the types of horses, how they run the next time, and how \"inaccurate\" the number is.  Further, what if the \"real\" number is somewhere in between the Ragozin and Thorograph number.  More importantly, we always have these debates about a particular day that just occurred, thus it is impossible to see if what you are saying is true.  That is why it wouled be necessary to look at a number of sheets over the last three years and have JB point out where he adjusted for \"anomolies\" and then maybe we could see what happened.

I\'ll close by asking you and JB a question.  I live in So. Cal. and own a couple of horses.  I can tell you that horse racing today is not the same as it was even 5 years ago.  More horses are souped up on steroids and god knows what elese (mostly legally).  This makes horses run big efforts, sustain them for a small period of time and then fall apart.  They cannot fill any races here other than maiden races.  Thus, seeing sheets that show horses doing crazy things at the major tracks corresponds with I see happening to racing.  (I won\'t name anybody on this board, but I would be interested in seeing the TG sheets for a couple of trainers in So. Cal.)

Why is it that Ragozin\'s sheets for mid-level claiming races at small tracks where horses run every two weeks have much more consistent running lines?  Are only the big tracks \"variable\" in major cities?  Why is that when I started betting in the early \'90s did the Ragozin sheets look much smoother?  Did these track issues only surface recently?

[TGJB]

Clarify your last paragraph--if you are talking Ragozin vs. Ragozin, you are asking the wrong people. If it\'s TG vs. Ragozin, you are asking me to swallow a whole lot of assumptions.

[JRL]

My point is that you and Alydar keep saying if you are wrong about the numbers adjusted for anomolies, you will not get tight patterns.  Thus, as you accuse Ragozin of failing to adjust for anomolies, the same would be true.  My point is that there are several tracks and types of races where Ragozin\'s patterns look very \"tight.\"  Further, though I can only say this through my own experience, the patterns on Ragozin have generally gotten less \"tight\" over the last 10 years.  This to me is easily explained by the changes in horse racing.  But if Ragozin\'s numbers being all over the place is due to their inability to get the variant correct, and not due to the horses, then why is this phenomenon seem recent and why does it only affect certain tracks?

[TGJB]

I have no idea whether it is happening at some tracks and not others--that\'s what I meant by asking me to swallow your assumptions. Leaving that aside, Ragozin used to do more tracks himself, and some small tracks are (or were) done mechanically. All this is beside the point--if you understood my long post you understand why. Tight figures CAN\'T mean anything other than accurate figures, when it applies to a data base as a whole (as opposed to one horse or one circuit).

[superfreakicus]

I guess I\'ve got a question, too.

how the hell can I bet 10 races w/a $1000 bankroll and end up w/$1400 using wildly divergent and inaccurate #\'s?
meanwhile, a whole lot of dudes using accurate #\'s couldn\'t touch me.
that\'s not even the tip of the iceberg --- go check my posted contest results and see how many tri\'s I just missed which included price horses.
either those #\'s kick ass, or I must be a friggin\' genius.

[TGJB]

Yeah, that\'s it. And it applies to everyone else who was on line with you to cash, no matter how they handicapped. You\'re right--you are a genius.