Quote:
Originally Posted by Bill Lyster
Mitch or Ted:
Which of the three workout numbers do you consider the controling piece of info or most important? The %, the SR or the WR?
Attachment 50367
For this horse, the winner, the last 4 w/o show improvement over the works 66 and 86 days back.
the place horse in this race:
Attachment 50368
The last 4 works are within 30 days of todays race. Two strong works 90-88 and then two works with the dogs up (which might hide possibly better final times). i've seen higher rated works followed by slower works, where one might consider the last two in this case to be maintenance works as long as the 90-88 were within 30 days of todays race.
[note] I can see that the process of researching this issue is going to take a lot of recording and reviewing, just from this one race. We need to come up with the most important factors to study in order to get meaningful info.
bill |
Bill, Mitch is adding an emphasis to a readout which has been in RDSS forever - the Rank% - and giving spot-play examples of using it, so to the extent that it is another useful tool, and knowing examples of HOW to employ it, I am grateful for the contribution!
Essentially, Mitch has suggested extending the Rank% measurement to
ALL Workouts, not just the last 5 maximum and only since the last race, as it was before. He has also suggested marking the BEST Rank% (or collection of BESTs within a tolerance range e.g. 4%) since last raced (or BEST ever, in the case of FTS or 2TS) to see improvement or readiness.
For horses for whom we have
less other info to go by (FTS/2TS, other Maidens, longer layoffs, etc), and coupled with observations about
changes today (changed equipment, Trainer, distance/surface) this represents
additional information to judge potential fitness. Maybe get a few more races at longer prices, take on a few more horizontal bets, maybe downgrade a 'hollow' favourite.
It's another 'angle', IMO with situational use, not necessarily broad use like BLBL/VDC/CSR/PP/PL. I have seen Mitch's examples publicly and privately focusing on longer priced horses and Maidens (though not exclusively), as he has written here. Some good hits, indeed. Many times, there are too many potential horses in a race with this angle so it gets confusing. A proper study would include all the hits AND MISSES and tally the ROI from legitimate and consistent inclusion of the angle in candidate races at candidate minimum Odds thresholds. Until then, it's anecdotal - which is not nothing! - and you don't have to give the money back when you hit. But you also don't get a refund if you lose by betting indiscriminately!
I don't think we have found the final presentation of this information about uses of Rank%, but we're working on it. I've added a Summary column on the Entries screen in the Workout section to highlight this Rank% angle, and filter it by other more well-known composite figure ratings (CR, CSR, Prime Power, ProfitLine, etc). Also to indicate degrees of effort against more horses -
for further investigation. Screen shots to follow soon; testers to get an update soon
One note about
Work Rank (WR) which Mitch says he does not like/use (no problem): per
the FAQ on this, this is a COMPOSITE factor, not something pertinent to each Workout, hence why it only shows on the top line beside the 1st workout, like all the other Form Patterns. It measures average Workout Rank% PLUS cumulative distance worked and raced within the past 30 day. Horses working slower, finishing behind more horses, working fewer or shorter distances will get a lower WR than the converse. All the detail is in the inputs to this factor - this is a Summary, with the shortcomings that a summary often entails. It is what it is, but again it's not nothing.
Also FWIW, though thanks for the opinions
, I think I'll stick with the long-standing formula for Rank% I gave earlier. We already know that the bullet work or
1/nn will be 100%, so the suggested alternate formula delivers no NEW information and while producing different numbers, tends to also be quite co-related to the ups and downs of the existing formula, so not really any lost info.
Also, the existing formula
does give extra info related to how many horses were beaten (work time was faster than), in that due to the higher denominator in the formula when more horses were beaten (1/35 versus 1/11) the resulting higher Rank% will tend to indicate that the effort was better than more other horses the same day and distance.
I also added a few tweaks to Rank% to respect efforts from longer distances (4f and up) and from a standing Gate start or around 'Dogs'.
Let's see what we see during subsequent testing rounds.
Thanks Mitch for your suggestions and feedback, and thanks Bill and others for the useful questions and challenges. Nice teamwork!
Ted