JCapper Pace Ratings

[socantra]

I just pulled down the documentation for Pace Factors from Jeff's site. I'll browse them and try to post with more knowledge after bird festivities.

Til then, I thought we might move this to another thread, so as not to stomp on Shoeless' TPR thread, which is producing useful info on the Total Pace Ratings, at least to Synergism users.

There are 13 pace factors in JCapper, out of a total of 113 or so factors. All of their pace calculations appear to be based on the BRIS pace figures rather than on running lines, so they would appear to have little direct relevance to the Total Pace Ratings from Pace Makes the Race.

Dick...

[socantra]

JCapper is certainly powerful software. It is essentially a database set up for handicappers. As both partsnut and JimG mentioned, it has a steep learning curve and is not for everyone.

I am one of those it is not for. I don't believe that database handicapping is the way to maximum profits. I think most factors will regress to the mean when subjected to rigorous statistical analysis. I am of the school that believes the proper approach to handicapping is that of early 20th century baseball player Willie Keeler, who is remembered for his advice to "Keep a clear eye and hit em where they ain't".

All that to say, I am probably not the proper person to review or criticize JCapper. What I feel I can look at is the pace portion of JCapper, which is only 13 factors out of a total of 113 or so. Those I think are weak compared to the tools we are used to as students of the Sartin methodology. They are simply not up to our standards.

At the heart of the problem is that JCapper, like Equisim and a number of other programs is almost totally dependent upon the BRIS speed and pace ratings built into their data file.

I have no real problem with their speed ratings. I've taken apart BRIS, Beyer, TSN, and Trackmaster, and I believe all are reasonably competent efforts at producing variable point per length speed ratings. After looking at the above mentioned products for several years, I can't find a whole lot of difference except that Beyer massages their data by hand much more so than the others and there is considerable disagreement over whether that is good or not.

The BRIS pace ratings though, the E1, E2 and Late Pace are, in my opinion, almost totally useless. I have tried to make logical sense out of them for the last year or so, with no luck at all. They just don't come together in any coherrent pattern. I'm sure others will disagree, but mine is not an isolated opinion. C. J. Milkowski of PaceFigures.com says that he was driven to create his program because the BRIS figures were "worse than useless".

Doc Sartin used to say that handicappers are the most adjustment happy creatures on the planet, and I firmly beleve that the BRIS pace numbers are an example of this. I think one would be much better off using raw paceline numbers with minimal adjustment for track and distance changes.

The Total Pace rating in JCapper has absolutely nothing to do with the Total Pace Ratings from Phase I or Pace Makes the Race. It consists of multplying the BRIS Early Pace Rating (E1 or E2, depending on distance) by two, adding the Late Pace Rating (which has been modified by subtracting 2 for each 2nd call beaten length) and dividing the whole thing by three. According to the docs, the last pace line is used, regardless of surface or distance. I'm not sure how the program decides whether to use E1 or E2. That is not documented.

The program's other pace ratings are all similar efforts. They involve the massaging of the BRIS pace ratings, essentially combining numbers and applying adjustments to these already adjusted numbers. Then you top it off with a few more adjustments just for good measure. Garbage in, garbage out.

My apologies to Jeff and the users of JCapper. It is an amazing piece of programming and an astounding program. Maybe I'm wrong about database handicapping and maybe it is the wave of the future.

I don't though, believe I'm wrong about the pace portion of the program. Even the world's best builder can't build a good house on a shaky foundation and the BRIS Pace Ratings are as shaky as any foundation I've ever seen.

Dick...

[Fast4522]

Dick,
Linear regression will eliminate more runners than wager capping could ever do by parameters that a given runner does not have at all different odds, and also tell you if that darn chalk that you loathe so much is good. I have been up to the very same stuff like Jcapper for about three years now, and will admit that the pace figures E1 E2 and so on are crap compared to good old phase 3. That being said when you combine Sartin with linear regression that is not Sartin, it becomes quite powerful. You may have two tiers that are very close (tightly grouped) but one may not even be considered as having the parameters that can win.

[socantra]

I'm not sure what you are talking about. I am not opposed to linear regrassion, as long as it does not get out of hand. I said not one single word about wagercapping. Do you classify anything that is not statistically oriented as wagercapping?

I am not prepared to argue about the merits of database handicapping. I merely said it was not for me. If it works for you, that's great. It doesn't work for me and doesn't interest me. If it works for 500 others, that's fine. We will be selecting different winners. I do not loathe chalk. It simply does not allow me the leeway to handicap as I choose. I would assume that some can make money from betting chalk. It is not for me.

My only contention about JCapper is that the pace section is fatally flawed from the use of the BRIS pace numbers. I am a pace handicapper. That is enough for me to loose all interest in the program. It is a very impressive piece of work. I'm sure that many people are and will be very happy with it.

Whatever you are doing with your personal program is not available to me. If it's another pace approach that works for you, I congratulate you. Since itr is not something I have access to, it is of very little interest to me on a practical basis. May you have good skill and double digit mutuels.

Dick...

[Fast4522]

Dick,
I agree with everything you just said, just pointing to what it has besides the crappy bris pace figures. And its true I mentioned wagercapping in the dated past in my view, and did not suggest that you were a chalk player or not one. You are correct when I mentioned "loathe chalk", to be honest I think we all do, but have understand when its good. I am not trying to be personal, but just interesting in posting.
Steve

[socantra]

No personal offense taken, tho I must confess to being somewhat puzzled by the occasional hostility of your posts in general.

I remember one that came out of the blue saying that wagercapping was crap and didn't work. I have personal experience that conflicts with that statement and know several others that do also. It is not a matter of debate and I see no reason to engage in one.

I also don't understand your contention that the board does not have coverage of the older manual entry programs. It does have that coverage. Its just that very few seem to care. When Bill first set up the board, I asked that he set up a section for programming, because I was interested in learning to do some things in Excel, and have done a lot of work with speed figures and such. He did, but after a few months of deafening silence, he took it off. I couldn't blame him for that. He gave it a shot and nobody showed up. He can't drag people in or force them to get involved in conversations they don't care about.

I also don't miss the endless threads of screaming at PA. They produce a great deal of heat, but very rarely any light at all. I don't appreciate the CNN style of debate, where two people take opposite ridiculous extremes and scream at each other til its time for the next commercial. I believe that kind of discussion is best held in neighborhood taverns, where it can be immediately referreed by the entire crowd.

I come here to learn, help and discuss handicapping. I find my time is wasted here less than it is at other forums. You are obviously a talented handicapper. I would be pleased to see you more involved in the discussions, rather than just dropping by to throw grenades.

No personal attack or hostility intended. I'd like to learn more about some of the things you do with your personal program. I will gleefully pick your brain, but will try to offer something of quality in return.

Dick...

[partsnut]

socantra;25780]JCapper is certainly powerful software. It is essentially a database set up for handicappers. As both partsnut and JimG mentioned, it has a steep learning curve and is not for everyone.

I am one of those it is not for. I don't believe that database handicapping is the way to maximum profits.

Dick, from what I've seen and have been shown, I am of the opinion
that this progam is a very formidable handicapping tool. It could indeed be the best of all available. There is no one element that would be used as a predictor. This is not a "black box" It's all in what and how you put the elements together and how you interpret them. It is a true handicapping program.
It can be extremely accurate with some very contentious races where you would not be able to use "the old school and the new school software" because you would not be able to get the correct pacelines and naturally not be able to land on a single pick winner.

socantra;25780: Im not prepared to argue about the merits of database handicapping. I merely said it was not for me. If it works for you, that's great. It doesn't work for me and doesn't interest me

Dick, from what I know about you, you're well read and well written and have lots of ability. Someone as knowledgeable such as yourself would have a field day with this software. You are computer literate and you could master this software in a short amount of time.

socantra;25780: The Total Pace rating in JCapper has absolutely nothing to do with the Total Pace Ratings from Phase I or Pace Makes the Race. It consists of multplying the BRIS Early Pace Rating (E1 or E2, depending on distance) by two, adding the Late Pace Rating (which has been modified by subtracting 2 for each 2nd call beaten length) and dividing the whole thing by three.

Thank you for the correction. The original TPR was good but in my humble opinion, the JCappers total pace is better. I was sent a recent query on WOX (Polytrack) and saw that the top 4 Total Pace numbers won 50% of all the races run at the track. This was a 13-14 day sample. The ability to query with the program tells you what's winning and whats not. From these queries you have the ability to create your own winning methods.

It is not important to me who buys the program, who likes it and who hates it.
If I could afford it I would have it.
Maybe Santa Clause will grace me with a copy for Christmas.

[Fast4522]

Dick,
Thanks for posting, if you want to not agree I like that too, honest. There is nothing I prefer better than a intelligent opinion that differs rather than a flower story. I sure wish I had it in me to play with the manual enter programs because it really looks like some are having alot of fun. I think the reason they are doing well is because they are doing a end run against alot of bad varients. Linear regression will point out when the chalk is good or bad or inconclusive as I posted before. Which to me says those who employ it are in fact picking up moneys from folks who duch bet and wager cap even if the dutch or wager capping is the result of a win.
On a different note I have always liked your posts even when its counter to what I posted and so on. I have deleted all trackmaster programs from my Laptop (what I am always on) because I find the propriety figures/varients to out with the moon at times. I say that to give you insight as to why I think and say some things. I have always checked data head to head (trackmaster & bris) and in this month I said enough. I look forward to your posts but like some I may back off to let others say what they feel. Right now its all about sales and money just like it all started. The BBS is doing well, it has many more folks reading its messages than it ever did, and I commend you for keeping a thread going.

Steve

[Turbulator]

Originally Posted by partsnut

I was sent a recent query on WOX (Polytrack) and saw that the top 4 Total Pace numbers won 50% of all the races run at the track. This was a 13-14 day sample.

OK, so TPR has the winner 50% of the time. Now what?

I can find similar statistics for a host of different factors.

How do you make money with this information?

[Turbulator]

Originally Posted by socantra

I am one of those it is not for. I don't believe that database handicapping is the way to maximum profits. I think most factors will regress to the mean when subjected to rigorous statistical analysis.

I concur. In my opinion back testing is bunk. After running NUMEROUS tests over a very large database with a similar database software program, I could find NO factor or combination of factors that showed a profit going forward in multiple tests.

If on the rare occasion, I did find one that tested positive going forward (and never much positive, usually somewhere between $2.01 and $2.15 dollar net), it would never test positive going forward on further tests. The only time they did test positive going forward was in short tests. In long tests failure was guaranteed. They always crapped out over time.

The conclusion I came to is the one I've held all along...in order to make money at horse racing you have to handicap. There's no other way.