"Clear As Mud" (735 Views)
Posted by:
TGJB (IP Logged)
Date: April 19, 2004 02:16PM
Some of you probably noticed Friedman's response to my recent post about the problems with Ragozin's 1 and 2 turn figure relationships in Southern California (done by Friedman). I'm pretty busy right now with the Derby and other projects, but briefly--
Almost exactly a year ago, I thoroughly discussed the subject here, pointing out lots of examples showing that Rag had the So Cal sprints way fast and the 2 turn races way slow-- the SA Derby winner getting a way better figure for winning a turf sprint than winning the two graded races going long, etc. You can find the post and LOTS of examples in the archives.
Aside from what Friedman was doing not holding up from a common sense point of view, I knew it didn't because there was a big disparity between what they were doing and what we were-- we tend to run roughly 3 points faster than they do, but in So Cal our sprint figures were only about a point faster, while our routes were averaging about 5 faster. So one of us (at least) was wrong.
Well, there are ways to check this stuff-- Ragozin used to have a statistical check while I was there, we do one a little more accurate here, and as part of some really innovative stuff we are putting in place now we will be doing something even more sophisticated. But we ran the check as recently as this past February for not just So Cal but all tracks, and it works like this-- you look at the winning figure for all the older claimers at a track over the last couple of years, break them out and average them at every distance (to see if you have a distance out of line), and combine all the one turn races and compare them to the two turn races. At the circuits where I do the figures myself (about 10 now) EVERY SINGLE SPRINT/ROUTE RELATIONSHIP CAME UP WITHIN HALF A POINT. A couple of the smaller tracks done by others came up with a slightly bigger spread (about 1 1/2 points), and we are fixing those up.
So Len-- what I suggest you do is use your computer the same way, or better yet, give the raw data to DRF and have them run an impartial test. Run the same test we did for Southern California (remember the mile at Hol is 1 turn), for the 12 months ending with the end of the 03 Santa Anita meet, and take a look at how it comes out-- my guess (not really a guess) is that you will come up with a spread of between 3 to 4 points. The reason I say to do it ending back then is, as I said in the earlier post, that your recent numbers seem more in line. Now, that could be because I've only seen a few Derby contenders recently, but there are only two possibilities-- either you fixed what you are doing, or the recent figures I saw are an aberration and you are still getting it wrong.
By the way, nice spin on Vegas. As you know, we put my Vegas presentation up on the home page of this site, and it was downloaded over 2,000 times in the first two weeks-- so I'm not too worried about the information getting out, or the spin.
TGJB