Charles Carroll - Handicapping Speed

[kingfin66]

I had heard of Carroll's book in the past, but have not read it. In the excellent handicapping anthology, The Best of Thoroughbred Handicapping, there is a brief chapter about Carroll's speed figures. The chapter outlines the basic idea behind them and the basic process for making the figures.

I ran a few cards using the basic figures and have to say that I am very, very impressed. As with any methodology, you have to use the information correctly and insist on value. These figures seem to find some hidden value.

Does anybody on the board use these figures on a regular basis? If so, do you make your own, or do you have the Speed Handicapper software? So far, I have just done up a quickie spreadsheet to help me test them out a little bit. I have seen Karlskorner post (yes, I have done searches) regarding the use. Karl also advocates keeping it simple, which I like to do also.

I would love to hear feedback from anybody willing to share their opinions, success/failure, and expertise in using Carroll's methodology.

Thanks.

[Tom]

I think Karlskorner uses a Carrol program to genreate the speed numbers. He has spoken highly of them -try a search for Carrol in posts by Karl.

[kingfin66]

From my post:

I have seen Karlskorner post (yes, I have done searches) regarding the use. Karl also advocates keeping it simple, which I like to do also.

I would love to hear feedback from anybody willing to share their opinions, success/failure, and expertise in using Carroll's methodology.



I'm looking for feedback from Karl and anybody else who uses it. I bought the book last night - I love handicapping books and like to look at different ideas.

[Light]

King


It amazes me that a well rounded handicapper like you is just getting around to this guys book that came out 14 years ago.

His book got me obssessed with the length of a horse cause he is right that what drf uses for the length of a horse is wrong and that is the strength of his figs.I went around measuring horses back in the early 90's and found they were averaging 8-1/2 feet,not the 10 feet drf convenietly uses. This means there are more lengths per furlong and therefore a length is equal to a different measure of time than we normally use.

However he seems to blend this insight with the same stubborness your man Pizzolla does. He ignores track variants and worse,he obtains his figures using the world record for a distance(not even the track record).

I had some degree of success with my personal adjustments on Carrolls work but I eventually came back to using 10ft per length again (even though I knew it was wrong) because it is mathematically easier to calculate figs . But that was back before computers were commonplace.With the power of Excel and Access today,it may be a different story for you.

I wonder if this is why CJ's figs are so successful here: From his web site:
the horse is adjusted at a rate of 8 feet per length

[kingfin66]

I don't know about being a well-rounded handicapper. In fact, I try not to be as odd as that may sound. Too many variables makes for sensory overload.

Carroll was at Pizzolla's 2003 seminar, so I've actually known who he was and of his book for awhile (even before that seminar). I just never gave much thought to the book as it wasn't one of the more commonly discussed pieces of equine literature.

The fact that he uses the world record doesn't bother me. That just means that there will be lower figures for some tracks than others. The lack of variant doesn't even bother me that much as his figures get into fairly extended decimals. What I am curious to know is what, if any, track-to- track adjustments are used. This does matter greatly. One quick and easy solution that I thought of and have been using in my test is the ATM track equilization chart. It does basically what I need it to do for a simple spreadsheet.

One more note about using WR. It would likely be a nightmare for him to use track records...so many tracks, so many records. In the case of a track like GP, the records fall frequently with the new surface. WR makes sense to me.

What were your personal adjustments and what is "some degree of success," if you don't mind my asking?

[kenwoodallpromos]

I like his online articles.

[Kreed]

Is the WR for sprints in some sort of relationship to the WR for a Mile or 9fs?
Distance adjs + Track2Track adjs + Daily Variants ---- i still don't think anyone
can work w/out them, but if you play one circuit the work would seem much
easier: putting AQU & AQU-inner & BEL & SAR into some sort of relationship
isn't easy but it seems doable.

[joeyspicks]

I like the software....Speed handicapper 4.5. Has no data base capabilities or automatic pace line method (selection). I have found if you take the time to pick a good paceline (he has sort method.....sort by Caroll number, sort by date, sort by distance).....the program produces excellant results. I wish it had an auto paceline selection method.....just to compare my pick with the suto pick. I find the software very accurate IF you are good at paceline selection. A bit cumbersome to use and you cannot load mor than one track at a time.

[kingfin66]

Paceline selection has long been the bane of handicappers. His article at desertsea that explains how he uses the program has a good explanation of paceline selection. He also defers to Pizzola's Form Cycle Windows. I am very well versed in that methodology but don't think that it applies to this type of speed/pace handicapping.

Thanks for the feedback Joey.

Derek, WR for each distance are used to calculate a speed figure for the distance being looked at. Each distance is broken down into lengths, then you take the WR time divided by lengths to geta time/length standard for the distance. That standard is then applied to what the horse actually did in the race. Oh yeah, and the winner's time per length factors in there too.

The beauty of these figures looks to be that you get a much more accurate time, based on beaten length adjustment, for the horses you are evaluating, and you are comparing all horses to the same standard...less subjectiveness is involved.

I tend to agree with you that this would work better within a particular circuit; Florida, The Rust Belt, EmD, SoCal, NoCal...

Thanks for the feedback. Please keep it coming.

[karlskorner]

I use the original program from Carroll's book, which ignores variants and speed ratings, written some 20 odd years ago, since then I believe he was influnced by other programmers and inserted SR/Var in a later edition of his book. I have discussed this with him numerous times. My original program was written on a C64 ( with much heartache ) and included a paralel chart which took many months to get right for CRC/GP/HIA ( I doubt if it will work elsewhere ). A friend converted it to PC when I bought a 486 and now 98. He died 10 years ago and I have had several programers try to break into the program, but failed. Hence, I keep a copy in a bank vault in case of hurricane, tidal waves etc. in S. Florida. The combined progams are simple, probably too simple for most handicappers, but it has worked for me for over 25 years. I know it's shortcomings and work around them.

[Light]

What were your personal adjustments and what is "some degree of success," if you don't mind my asking?

I used 2 adjustments. In step 1 of the calculations, I adjusted the winners time by the drf track variant according to the method Brohamer describes in "Modern Pace Handicapping",which incidently came out the same year.

Secondly,(I believe step 5),I divided the horse's time into the local track record rather than the world record.My feeling was horses coming from slower surfaces will have a better,more fairer rating instead of being penalized for running on a slower surface.Adjusting a Suffolk Downs horse to the world record was insane.

By "some degree of success",I mean it seemed as good as what I was using at that time,but since I had to convert everything the drf form told me,it was too much work,so I dropped it. But you have rekindled my interest now since I can probably have a macro do the math. The laborious part now would be creating a database and studying its results to see if its worth it. However I would doubt it is much better than Sartin,Beyer,Pizzolla,Prime Power,(etc) figs.And some of those are alot easier

[kingfin66]

Thanks for the feedback Karl. I saw your previous posts when searching Carroll on this site. Your sticking with the original program is very telling. More elaborate is not always better. From the short chapter I read in Quinn's book, it is apparent that the numbers themselves are actually pretty elaborate.

When you use your program what data do you have to input? My guess is winner's time, beaten lengths, and distance? How do you adjust for the changes in the GP track surface and world record times if you can't modify the program?

Hopefully, my copy of the book will arrive in the next day or two. I almost went for the original 1991 hardback edition, but decided to go for the latest edition.

[karlskorner]

Sent you a PM with my personal thoughts