Jason L. wrote:
>
> In the interest of full disclosure, I am a Ragozin user.
> However, I can assure you that I have never spoken to anyone
> in the Ragozin and am not making this post in an attempt to
> secretly tout their product. I do have a few things to say
> about this exchange.
>
> When you take as a compliment the fact that your winners
> "look better" you are entirely missing the point. All of
> your horses look "better" because your lines are smoother.
> That is only a benefit to the unsophisticated sheet user.
>
> In my mind, the sheets are extremely valuable for two things
> (1) identifying horses to throw out as too slow or bounce
> candidates and (2) identifying horses with explosive patterns
> that are likely to move forward. When you have smoother
> looking lines, it may make the horses look better, but it
> makes both of these taks much more difficult to accomplish.
>
> The fact is when I look at your sheets, several of the horses
> in every race look more or less the same and each horse tends
> to have a smooth looking pattern. Nobody is arguing that you
> go back and tweak individual horses numbers to make them look
> better. However, when your overriding philosophy is that
> groups of horses in a single race do not deviate from the
> norm -- a proposition that does not comport to my own
> experience -- then by definition your numbers are going to
> have far less deviation.
>
> In my view -- and I do not purport to be a variant expert --
> there are just too few data points in a single race to make
> the conclusion that the track must be the reason all horses
> ran slower or faster than you would expect compared to
> different race on the same card. While I am willing to accept
> the premise that sometimes the track changes significantly
> during the day, I suspect that is far less often than you
> articulate (mostly it seems to me, in an attempt to
> distinguish yourself from Ragozin).
>
> I do not see any valid scientific manner in which to change
> the variant for a races on the same day based on what you
> would have expected to occur. In fact, the process of doing
> so makes the entire statistical analysis suspect. On a
> particular race, you may be right and you may be wrong, but
> this process moves too far into the realm of guesswork for my
> taste. Though all variant making includes statistical
> analysis with a little guesswork, the idea is to minimize the
> guesswork, not maximize it. If that means some of the
> numbers turn out to be "wrong" -- I have certainly seen some
> suspicious looking numbers in the past -- so be it. But at
> least with Ragozin, I have some comfort that the numbers are
> based on a consistent scientific foundation and not one
> person's opinion.

TG--See my replies to David Patent, also my posts Changing Track Speeds (11/17/01), and Figure Making Methodology (5/2/00). You are welcome to post here any time with comments or questions.



TGJB