ThoroTech    ThoroTech Racing    ThoroTech Racing  Hop To Forum Categories  EquiSim - Feature Ideas    Poll - Revising Abilities
Page 1 2 
Go
New
Find
Notify
Tools
Reply
  
Poll - Revising Abilities
 Login/Join 
Steward
Graded Stakes
Picture of Nathan M.
Posted

Question:
Folks,

In preparing the next version I am revising a great deal of the Profiler stuff. At this time, it may also be prudent to eliminate a number of "abilities". I never claim that I have all of the answers, I claim that I have the tools to find out the answers.

It may be that some of the abilities that were introduced in EquiSim about a year ago have not proven their worth in the PROFILER. I say "may be" because there are alternate uses for the data that some may be employing to great effect. I'll explain below.

In general, the abilities in question include any ability that has a typical ratio of starts to races above 2/1. What happens is that the number of starts dilutes the statistics too much. Or does it? Here's an example, from OPX, 6f Sprints:

In the most recent 40 such races, the SPEED ability:

Rating, Races, Starts, Win%, ROI
Dominant: 17, 17, 41.18%, -13.53%
Best:40, 42, 30.95%, +16.67%
Strong: 40, 81, 20.99, +8.89%
Weak: 40, 154, 5.84, -49.94

Other than the fact that Speed looks like a darn good indicator in OP sprints, what this tells us is that you have to have a strong speed (according to ES's style of selecting PLs) to be competitive. Weak speed is really bad. But, by and large, we know this from inference, right? Or do we...?

Its possible to look at the differences between different "weak" (and strong) abilities and get a feel for how important having, or not having, a weak ability is. To add to the example, Weak E2 horses win 5.2% of the time while Weak Late horses win 7.65% of the time. This tells me that early speed is more important than late speed here -- more horses can get away with weak late speed and still win than the number that can get away with weak early speed and win. But, is the difference determinable through Profiler enough - is it valuable (that is, is the information non-trivial and worth spending the time evaluationg)?

To me, the answer is no, its not valuable enough for me. I rarely, if ever, use the abilities that typically have a 2/1 or greater ratio of starts to races. I'd like to hear the opinion of others.

There is one other important aspect about these abilities -- they appear in the FORM view (optionally). That means you can look at a horse in the Form and quickly scan to see how its selected pace line derived abilities match up with the other horses in the field. If Strong/Weak and other "high start/race ratio" abilities were removed, these would no longer be present.

You may wonder -- what's the big deal. Since I can HIDE the abilities I don't want using the Profiler Preferences dialog, why are you making a fuss? The fuss is a technical one - the amount of time ES needs to do a Profiler (analysis) for a given race is exceptionally fast due to the database structure. In geek-speak, it uses 64 bits of information to store all of the possible abilities -- each bit represents an ability. That means 64 is the maximum abilities we can have. While the database can be revised to accomodate more, its much more of a pain and something I don't want to do. If it seems reasonable, I'd prefer to rip out some unused abilities in favor of new, "experiemental" ones.

-Nathan

Choices:
Axe them - I don't use the types of abilities being described
Keep them - I use them in the PROFILER
Keep them - I refer to them in the Form View
Huh? I don't know what you're talking about

 
 
Posts: 1981 | Location: North East OH | Registered: July 23, 2002Reply With QuoteReport This Post
Steward
Graded Stakes
Picture of Nathan M.
Posted Hide Post
I'm really thinking through this here....

It may be worth while to revise the way that Strong and Weak are treated. In the example cited, (and I'll approximate from here on out) knowing that 20% of 80 horses with Strong ability in 40 races won means that there were 16 races won by strong speed ability horses (Best Speed is also Strong Speed). That's 40%. (This does not mean that *A* Strong Speed horse has a 40% chance of winning - a Strong Speed horse has a 20% chance of winning since there are more than 1 strong speed horses per race).

Take this data a bit further. If 30% of 40 BEST speed horses won, that means there were 12 BEST speed winners.

So, we have 16 races where a Strong Speed horse won, of those 12 were won by the BEST speed. That means ONLY 4 horses with STRONG speed and NOT best speed won! This would indicate, if you were keying in on speed in your handicapping, that if you thought the BEST speed horse was not going to win, you could look OUTSIDE of the EquiSim Strong Speed category of horses for the winner and find that winner (in this sample) _most_ of the time _when_ you were right about the best speed horse not winning. Consider that non-strong speed horses usually pay pretty well, this deductive means of finding contenders starts to look, well, profitable!

Essentially, 40 races - Best speed winners = 28 races. 28 races - 4 strong speed winners = 24 races. 24 races / 28 races = 86% of the time Strong Speed horses did NOT win when the BEST speed horse also did NOT win.

Maybe I'm kidding myself, but that seems very useful. The problem right now is only that the presentation of the information requires you to do the above manually to get down to the more useful info. Is it worthwhile to employ ES to do this type of calculation, worthwhile to spend the time determining a useful way of presenting this information? Hmmm.....


-N
 
Posts: 1981 | Location: North East OH | Registered: July 23, 2002Reply With QuoteReport This Post
Graded Stakes
Picture of JustRalph
Posted Hide Post
"Maybe I'm kidding myself, but that seems very useful. The problem right now is only that the presentation of the information requires you to do the above manually to get down to the more useful info. Is it worthwhile to employ ES to do this type of calculation, worthwhile to spend the time determining a useful way of presenting this information? Hmmm....."

after watching the way you cleaned up in that contest yesterday.....I would say "it might be worth using all these idea's" . Nathan, the nice thing is that you have already figured most of this stuff out for us. But I am glad you are taking us along for the ride. I can tell the guys on this board that haven't met Nathan, that all of this stuff goes thru his head in about a second and a half. He is hard to keep up with (at least for an old country boy like me) but his mind is always working. I for one am enjoying the ride here. This is fun, but I am a little weird. Keep coming with the information Nathan. It can only make me a better user in the end. Being a better user of ES results in profits. Cool

Just Another Horse player
 
Posts: 503 | Location: Mid Atlantic Region | Registered: July 25, 2002Reply With QuoteReport This Post
Graded Stakes
Posted Hide Post
Nathan..

I thought you would still be partying; however, here you are in dee-eeep thought.

Unless it's a turf race I never pay any attention to the "weak" designations, and, then, only at the weak E1. When looking at a turf race I will look for the "weak" E1 only when I have found a "best" or "dominant" late pace horse. Completely axeing all designations except "best" and "dominant" would be just fine with me. Also, on the rare occasion that I print the form view, it would be nice to omit the printing of those weak designations.

I concur with JustRalph...thanks for giving us a glimpse of the roadmap.
 
Posts: 70 | Location: Baton Rouge | Registered: July 25, 2002Reply With QuoteReport This Post
Steward
Graded Stakes
Picture of Nathan M.
Posted Hide Post
Hey George,

Thanks for the feedback.


Don't forget you can go to the Profiler Preferences dialog, Display tab and hide the abilities you are not interested in.
A) They won't appear when you view (or print)
B) ES won't waste processing power calculating stats for hidden abilities.

-N
 
Posts: 1981 | Location: North East OH | Registered: July 23, 2002Reply With QuoteReport This Post
Graded Stakes
Posted Hide Post
N...

I've tried everything to keep the "weak" and "strong" from printing or even showing in the form view, but those suckers are everpresent. I can omit both in the Profiler view, however. I am rather sure that I have properly "hidden" all "weak" and "strong" designations.

I'm obviously overlooking something. Give me an assist when you have time.

Thanks, George
 
Posts: 70 | Location: Baton Rouge | Registered: July 25, 2002Reply With QuoteReport This Post
Steward
Graded Stakes
Picture of Nathan M.
Posted Hide Post
Ahh, you are right George. They are everpresent in the Form View. The only way to rid them is to use the Form View menu and turn off the ALL the abilities from being displayed.

-N
 
Posts: 1981 | Location: North East OH | Registered: July 23, 2002Reply With QuoteReport This Post
Graded Stakes
Posted Hide Post
N...

Don't worry about this at all. It's just another of those small things that we old men insist on fretting about. If I wasn't frettin' about this I'd be frettin' about something else.

I seldom print the PP's, and you've got much better things to do. Thanks anyway, but fergit it!!
 
Posts: 70 | Location: Baton Rouge | Registered: July 25, 2002Reply With QuoteReport This Post
ALWN1X
Posted Hide Post
I have been reading everyone's comments about which settings they use.
Nathan,(our resident expert), which settings do you think would produce the best indications for the winner?

Personally, I use the Speed settings, plus the Strong E1 and E2, along with the strong Late for long higher class races.
I also use the E-Distance Cutter because I have picked some horses that have been competitive in its last race and are shortening up 1/2 length in today's race.
Anyone care to add to this??

MyHorse1
 
Posts: 30 | Location: Tulsa, Okla. 74145, USA | Registered: July 25, 2002Reply With QuoteReport This Post
SoCal Handicapper
Graded Stakes
Picture of JimG
Posted Hide Post
Myhorse1,

I use track, dist category. surface, and no off-tracks.

Good question, I'm looking forward to the responses.

Jim
 
Posts: 581 | Location: Virginia | Registered: July 25, 2002Reply With QuoteReport This Post
  Powered by Social Strata Page 1 2  
 

ThoroTech    ThoroTech Racing    ThoroTech Racing  Hop To Forum Categories  EquiSim - Feature Ideas    Poll - Revising Abilities

© 2002