Stats oddities, again...

[JohnTChance]

I apologize for bringing up the following oddity. Maybe I\'m crazy or stupid,
or I\'m not understanding what I\'m reading. But things like this startle me
to no end because of the idea that serious dollars are being wagered,
based, at least in part, on the statistics you provide in your product.

In the 9th race at Pimlico on Friday, in the sheet for FAMILY BUSINESS,
the jockey profile for Jerry Bailey AT PIMLICO says that he\'s had 12 starts there.
12 starts and 3 wins (in the last 12 months). However in the sheet of HIGH FLY
who\'s running in the Preakness on Saturday, THAT sheet\'s profile of Bailey
reveals that he\'s only had 5 starts AT PIMLICO (in the last 12 months).
5 starts and 2 wins.

Not that this specific oddity is of  earth-shattering importance here, but...
what gives here? And by the way, which stat is correct? Please explain,
if you can.

JohnTChance

[TGAB]

John, both data are correct.

Briefly, we processed the 5/20/5 Pim entry file on 5/16/5. The jockey backups lag a day, as do the trainer backups, which means through 5/14/5. Preakness day 2004 was 5/15/4 and qualifies as last 12 months. The Preakness entry file 2005 was processed 5/18/5 with jockey data updated through 5/16/5, which means the last 12 months starts from 5/17/4. Bailey had seven mounts on Preakness day 2004.
 
For more information on the timing of trainer, sire and jockey updates see my note from 9/9/03--Re: \"A similar question about sire & trainer stats\".

[JohnTChance]

Thanks for straightening that out. But another question...

In any sheet for trainer Michael Dickinson on Friday\'s Pimlico card, such as
FOREVER DAYLIGHT in the 3rd race... in the RUNS based stats, which I think you\'ve
told us are \"real time,\" and are \"from 2001 on,\" Dickinson is 0 - 0 - 0 in LAST RACE
UNDER 11 days. That is, NO starts. But in the FIGURE-based stats, he has as
many as 9 starts in this category.

Since the FIGURE-based stats [have a heavy query but...] are taken
from the year 1999 to the present, am I correct in assuming that the
discrepancy is explained in that all his starts in this category came in
the years 1999 and 2000? I\'ll assume this is true. But it seems a bit odd.
And makes me wish both the runs-based and figure-based stats are
taken from the same timespan - without overlap.

JohnTChance

[TGAB]

There is an error in the runs-based stat for Dickinson in this category. Dickinson had at least 3 runners come back under 11 days from 2001 to the present. We are fixing the error. Thanks for bringing it up. We try to do the best we can but occasionally we do make errors so keep looking and keep asking questions.