CSR / Composite Speed Rating
 

One of many questions.
Has anyone did a serious look (model) of this factor?
From Ted's writings he considers it a strong factor.
Thanks.,
gillb

Richie Mathie [ R.Math] is the guy to check with. That's his baby and he has great stats on it. Shoot him a p.m. and he'll fill you in.
Tim

CSR / Composite Speed Rating
 

Can someone please help with this RDSS2 factor.
Matbe r.math
Thanks
gillb

Hey gillb send him a pm. I talked with him today he will help you.
Tim

Csr
 

Hi Tim … thanks
I can't find his email/contact.
Bernie

Bernie I'll send you a pm in a few.
Tim

Bernie, you can read about the Composite Speed Rating (CSR), new to RDSS2 (i.e. not in older Sartin programs) in the RDSS Documentation PDF found in the RDSS Documentation section.

Here is the link to the RDSS Documentation PDF, which you can download and read for handy reference:

http://www.rdss2.com/pubs/RDSS2.1.Glossary.pdf

CSR is first mentioned on Page 3 of the Documentation. In the PDF, you can search for Composite Speed Rating or CSR to find other references to this factor.

You can also search the P&C site using the Site Search tool (http://www.sartinmethodology.com/googlesearch) to find existing discussions about CSR and its usage.

Still, I can recommend you talk to Dick M (rmath) about the work he has done with this and other factors. He has done that a lot in other discussion threads.

Ted

Non Paceline
 

Hello

Here are some stats from my SoCal RDSS model
DMR and SA - All races

Factors APV, CR ,CSR, CR+, PL, RX1,2,3

Attachment 48626

Bill,

Just a question. Is the CSR stat shown in your post, the original CSR that you get when you open RDSS or is it the adjusted CSR (if required)?

The reason I ask is when, say, a horse is running today at 9F but its last or second last race was an 11F or 12F race the usually slow adjusted speed rating from the marathon makes the CSR rank vs the other horses so low as to not be in the top 5 CSR. The CSR can also be low if the last or second to last race involved a significant trouble notation, in which case you might want to have the program not use the troubled race.

Regards,

Hi Bill

Lovely weather we are having here in Northern San Diego. If I could I would be building an Ark:o

Bill What I posted from my model is out of the box RDSS. I am not sure how to adjust the CSR numbers

What I can do is post my SoCal model sorted to eliminate any race longer than 8.5 furlongs. This way in the few instances were a horse had been running marathons, and now is cutting back to 8.5 or less, I don't think the percentages would be much different.

Here is the chart from above but with no races longer than 8.5 furlongs

Attachment 48627

Bill,

I assume that you know that CSR gives the most recent race the most weight and successive races less weight (last 3 or 4 - I'm not remembering what Ted told me.) Because of this, if the horse has an unusual or poor running line in races 1 or 2 back, the CSR number computed using a bad race automatically downgrades the CSR rank.

Here is a horse from last Friday at DMR. The first screen shot shows the 3 with a top 3 APV (in yellow) but a CSR rank of 5. When you put your mouse arrow on the CSR and click it, it eliminates the worst adjust SR and recalculates, in this case, the CSR rank compared to this field is now 3rd best instead of fifth best. Due to how little lines from 3 races back or more are weighted, the rank has the best chance of changing if lines 1 or 2 are not reflective of what the horse normally does.

This is essentially an NW2L race and the last two races were a little higher class, so you can use the maiden win line or the 2nd back if you ignore beaten lengths and rely on total energy. The chart for the last race does not indicate any problems other than perhaps the fact that better horses were entered.

Recalculating CSR should be done sparingly, much the same way you would analyze a past race for forgivable problems. Sometimes the recalculation does not change the rank, so if CSR rank is a significant factor to your handicapping, be mindful of that fact.




Attachment 48629

Attachment 48630

Csr
 

Thanks Bill

Now I know how to adjust the CSR. Thanks a million
Good skill to you

Bill L.
My pp's show the #3 with 7.00CSR and adjusting to 4.00CSR
Any idea why the difference?

Same APV, CR, PL, BPP

No clue on your question Dogkatcher; could be the version of RDSS, but that is just a guess.

One thing to note. If you decide that horse A needs to have its CSR adjusted, and you go ahead and adjust it, you will see the revised rank after doing so. But if there is another horse in the race that needs a recalculation, the program will revert horse A to its original rank.

So once you recalculate, note the new rank in case it gets superseded by a subsequent "other horse" recalculation. Then when you look at the Rx3 screen mentaily know that horse A's CSR rank is whatever it changed to after recalculation.

dogcatcher it could be because of different settings for your variant & trk. to trk. adjustment.


Bill Lyster can you do this more than once to adjust the CSR if it has two ugly races ?


Mitch44

Thanks Mitch, Variant adj. made the difference.

Mitch:

I don't think so. As I recall the program uses 100% of last race; 80% of second back, approx 64% of 3rd back and about 51% of 4th back,so if horse has two subpar (your visual) adjusted speed ratings in a row, the rank rarely changes if you have to rely on race 3 or 4. Better to just mentally take the two bad races into consideration (if there are legit reasons to excuse them) and use later lines knowing the horse is prone to racing poorly (or perhaps WAY over its class) several times in a row.

Thanks Bill. I usually do this mentally but it would be nice if we could select the correct 3.


Thanks,
Mitch

Hi guys, CSR uses Adjusted Speed Ratings (ASR), so any adjustment settings you have in the Config screen (ITV, DTV, DTV range, Default/Val Mode) will result in slightly different ASRs, and thus slightly different CSRs.

It was easy enough to skip over the single worst of the recent ASRs, but much more work to build some kind of interactive means to individually select which SRs should be included (and to make sure there were enough selections made and not too many, etc, etc). So I went with just excluding the single worst and tacking on the next one down the list, if there was one.

I agree it would be nice to be able select specifically the components of the CSR, but then you start to drift away from one of its strong points - that it's an AUTOMATED calculation.

Getting rid of just one bad Speed Ratingis better than nothing, IF you need that adjustment, IMO.

Maybe someday ... :)

Ted

"Maybe someday. . . " I believe it would greatly produce a better rating and at the same time strengthens the line scores of the RX+ screen.

As is, it does with its automated calculation / black box approach provide the layman a better approach to nothing however its not an acceptable approach for me and I work around it with my own modifications. It would be a nice update, one can only hope. Sartin has always provided the users with choices.

Mitch44

That is absolutely the case. Part of the race prep process has to (still, of course) be looking very carefully at the horse's record. Using or not using a race line in it's (CSR) computation is crucial. Absolutely crucial. I think Ted and I talked once about being able to select which races to use for this number, in RDSS. I always use adjusted numbers (which come from TrackMaster) in my calculation of this, (and related numbers which in my models go into an entirely separate set of "further" data .. ie further numbers that can flow from just one set of numbers. I calculate my own CSR, but it's virtually the same formula that is in RDSS). They are powerful. I would defer to others who have a lot of data on these, in terms of stats, but in general, TrackMaster's adjusted numbers, with further statistical manipulation, (ie a correctly calculated CSR), are quite powerful. There is a short YouTube presentation of how TM does it's speed ratings, if anyone is interested.

Also just a brief note about the "#num!" (error message) you see in Bill V.'s screen shots. The Excel "min" formula, ... the row with the "blue" 0.85 (which if you're using the actual worksheet Bill is, means that 85 % of the data in that column, falls within the range of "1" to whatever the "min number" is showing as. The "min" is adjustable ... he could change it to 0.50, or 0.75 or whatever he wants to see. That error message means (almost certainly) somewhere down in the column of data (that it is calculating), there is either a blank or a "non-number". If that gets removed or fixed, the field with the "#num!" error message will display an actual calculated number.

Very glad to see the way that data is summarized was useful, Bill ! Thanks.

Happy and blessed holidays to all !!

Tom
(Dorianmode)