HOW FAST are horses getting faster

[miff]

Kev posted,

On TG figs, for the BC.C from 99\' to 2004. The top two horses figs.

99\' = 2.2 and 0.2
00\' = 0 and -1.0
01\' = 1.2 and 1.2
02\' = -2.2 and -0.2
03\' = -3.0 and -1.2
04\' = -4.2 and -4.1


 
Does anyone have the Beyers and Rags for the same races for comparison.Looks like app a 10 length improvement in 6 years on the top fig.

[kev]

Beyer\'s, just for the winner\'s.

90\'  116
91\'  120
92\'  114
93\'  114
94\'  115
95\'  117
96\'  115
97\'  120
98\'  116
99\'  118
00\'  116
01\'  117
02\'  116
03\'  119
04\'  124

Rag\'s
99\' 2+
02\' 0\"
03\' -2
04\' -1
I don\'t have 00\' or 01\'

[TGJB]

I would love to see the actual Beyer quote about adjusting figures retroactively for \"creep\" one way or the other, and the rationale and methodology involved. If anybody knows where that can be found, let me know.

JB,

I also recall that quote and I\'m sure I can find it for you eventually. I just don\'t recall where I saw it. When I find it I will definitely post it and the source.

I believe the problem goes back to before he started publishing figures in the DRF.

It was a problem of shrinking figures.

I think he wasn\'t projecting enough improvement in the figures of some lightly raced young horses and it was very slowly dragging his figures down.



Post Edited (06-02-05 13:16)

Page 114 of \"The Winning Horseplayer\" 1983.

\"The projection method produces extremely accurate figures, but it also causes one subtle problem - a problem I had once thought was due to my personal errors until I found that it troubles almost everyone that makes figures this way. When handicappers make their projections, they tend not to anticipate enough improvement for horses who win races. If the projections are lower than they should be, the resultant figures are lower than they should be. And after a while the handicapper will notice that all his figures seem to be shrinking.......When I first encountered this problem my figures were shrinking so fast that I could barely compare recent races to ones that had been run two months earlier. Fortunately, the remedy for this problem is a simple one. At the end of each month, I review all the results and compare the figure of each race with the par figure for that class. If I find that the races have been run, on the average, one point slower than par, I will retroactively add one point to all my variants and figures for the month\".  1983

[TGJB]

NOTE-- this is what I was talking about when I said pars would affect your figures long term if used ANYWHERE in the process. I don\'t know whether Beyer (or those doing some of his circuits) still do this (there are strong indications they do something like it, having to do with the relationships between circuits), but if you bring your figures back to an artificial \"par\", you will not be able to see whether horses as a group are getting better or worse. I had this conversation with Friedman on the Rag board a few years ago, and touched on it at the DRF Expo.

Time-Form does something like this as well.

TGJB,

I agree with you.

I don\'t know if there is any tying back to pars these days, but I know that the cross circuit check is computerized based on the actual figures.