I wouldn't be at all surprised to see this one lining up at Ling on good friday.

womenandthosewithnointerestinracing,

What method do you use to store your ratings and therefore how do you get them into the very neat formats above?
n.b I make jorvik prince 3pts (1pt = one length = 0.2s @ 1mile) in wolv 6.15
W6.50 Agree, loads on same top rating
w7.25 Energia top rated by 2pts
w9.00 Sir Gnet top rated by 3pts

And thanks to T2R's thread on Labs Chat for how to post the tables from excel on here.

I'll cut some of that out to fit better.

so you do all the calculations and store in excel? i store my rating within raceform interactive, but cant seem to export them easily to excel

Yes, all calculations in excel, I copy the results in from racingpost.com and sort out all the relevant stuff from that (see that big string in column C in the first example), then my output is the last 4 columns shown here which I c&p values onto another sheet and sort/filter it as I wish.

Plenty cleverer than me I should fink.

113 on my scale which is roughly 78 on your scale. my top absolute top ratings work out at 120-125 normal class 4 handicap about 106.

so did you award 84 from the above then?

my going adjustment was +0.00 so quite fast, maybe thats why mu number is smaller.
my ratings for all races were 98 104 95 113 100 107 98 96

undoubtedly, i cant disagree with you. as you can see quite a bit of head scratching for me:

                                                                                v standard projected    actual

9008x    WOLVERHAMPTON (A.W)    WOLV6F20Y    OLDER    N    6    75.1    -1.43    102    0    98
9009x    WOLVERHAMPTON (A.W)    WOLV1M5F219Y    OLDER    N    6    185.2    -0.32    101    0    104
9010x    WOLVERHAMPTON (A.W)    WOLV7F36Y    2B    N    5    90.4    -1.25    0    0    95
9011x    WOLVERHAMPTON (A.W)    WOLV7F36Y    OLDER    N    5    87.2    1.13    0    0    113
9012x    WOLVERHAMPTON (A.W)    WOLV7F36Y    3B    N    6    89.4    -0.53    99    0    100
9013x    WOLVERHAMPTON (A.W)    WOLV1M4F51Y    OLDER    N    4    159.2    -0.35    100    0    107
9014x    WOLVERHAMPTON (A.W)    WOLV1M1F104Y    3B    N    6    122.2    -1.02    0    0    98
9015x    WOLVERHAMPTON (A.W)    WOLV1M1F104Y    3B    N    6    122.7    -1.43    0    0    96

One factor that may cause a discrepancy is that I don't bother with 1¾+ any more, and I see that your fig for OFIF is higher than the other class 6s.

9008x    WOLVERHAMPTON (A.W)    WOLV6F20Y    OLDER    N    6    75.1    -1.43    102    0    98 ___ 63
9009x    WOLVERHAMPTON (A.W)    WOLV1M5F219Y    OLDER    N    6    185.2    -0.32    101    0    104 ___ N/A
9010x    WOLVERHAMPTON (A.W)    WOLV7F36Y    2B    N    5    90.4    -1.25    0    0    95 ___ 55
9011x    WOLVERHAMPTON (A.W)    WOLV7F36Y    OLDER    N    5    87.2    1.13    0    0    113 ___ 84
9012x    WOLVERHAMPTON (A.W)    WOLV7F36Y    3B    N    6    89.4    -0.53    99    0    100 ___ 64
9013x    WOLVERHAMPTON (A.W)    WOLV1M4F51Y    OLDER    N    4    159.2    -0.35    100    0    107 ___ 70
9014x    WOLVERHAMPTON (A.W)    WOLV1M1F104Y    3B    N    6    122.2    -1.02    0    0    98 ___ 59
9015x    WOLVERHAMPTON (A.W)    WOLV1M1F104Y    3B    N    6    122.7    -1.43    0    0    96 ___  56

i see that your ratings are not purely linear like mine. i.e you differences to my figures vary between 39 - 40 pts. anyway i am in no way saying your rationale is worse than mine btw. its great to have the opportunity to discuss speed ratings with other compilers.

Good to have a 3rd reader of the thread CW.

You doing Southwell as well? As mentioned a few pages up I decided to can it and stick to poly & tapeta.

i may have a small bet on Sir Gnet in 9.00 at wolv tonight, my first speed rating related bet of the year.

5f has always been difficult at Southwell, but I find it unsatisfactory to omit a sprint distance, whereas I'm happy to omit the staying races.
Allez Sir Gnet!
The scale diverges a bit from Mordin. I tailored it specifically to poly/tapeta courses after analysing 2 years worth of collected times by grade.

I developed this particular method about 3 years ago. Had a decent run for a bit, then a less good run after upping the stakes, also I may have been guilty of over-filtering a bit. So this time, having tweaked the method slightly, less filtering and the bets fed in by a bot that stakes bank/200. As I say I'm testing in public, so this started at the beginning of the thread in June.

WD Den