Modeling Do You Use ?

[pktruckdriver]

Modeling a track's many variants and many classess and ages and sex of the horses.


I was what wodering what percentage uses modeling as a form of handicapping, and if they let there software do it for them or do it the old fashioned way.

Now-a-days there are many out there with there homegrown track modeling programs, are they working and worth the effort?

Some of the major programs provide that info in them, others may be harder to get it out of , but then others don't, and as I get back into this again, I was this time wishing to be a little more track specific, instead of nationwide as before.

I guess I am asking if track modeling works for you, but any advice you would share here, yeah on PA , I know, but then who actually listens to it anyway, most are dead set in there way, and will ignore it anyway, right?

I will not hold my breath here, but then again you guys been helpful in the past, very much so actually, so thanks.

[acorn54]

the software "equisim", provides a track modeling. you can set the model for any class level,number of races you want to got back, off or fast tracks only and so on and so forth. the rating it comes up with is called the "profile" rating.
the best it was able to do is get you in general is a -10 percent roi.
it is based on all the factors you chose, that are available in the bris data files.

[Tom]

The HTR Robot lets me model almost anything in a short time frame.
I can see how early speed is holding, what tracks are shipping in successfully, how certain trainers do in various situations, what factors are hot, what ones are cold......Donnie posted a small MaxVel model in the software thread a couple of weeks ago that was eye-opening. I hit a stretch last year at Del Mar where the top 2 SP horse were hitting at 70% for a couple of weeks.

[kenwoodallpromos]

IMHO, (1) the distance as well as the other track variations (modeling) determine the winner more than any other factor except (2) far superior quality of a individial animal who is rested enough; And the shorter the field the more importantof those 2 major factors.

[Houndog]

Many people will not keep a model or a track profile because of the work involved. As Tom said in his post the HTR Robot makes this as painless as possible. It is very user friendly and requires no special programming skills.

I tried the trial download a while ago and now the Robot can model the MAXVEL readouts which is an added plus.

[upset]

I think its more important to kept track of any bias for handicapping. If a horse lead to the eigth pole and stopprd 10 days ago its nice to know that only one front runner won that day. As far as modeling? You should inherently know the tracks you bet and how they usually run.

[acorn54]

i don't know how you detect a track bias
if the horses are front runners early and win is it really an early speed bias, if all the early speed winners were low odds horses?
anyway the bias can change on a moments notice and put you off a good odds horse that you would have, except for the perceived bias at the moment.
i talked to someone who tried following track bias and he told me all week long it was a closers bias at the track he was following, his computer selections for the day were two early speed types and he passed on betting them and they both won at long odds.
an anecdotal story for sure, but if only nine races are run for the day on dirt and all your winners are low odds early speed types its questionable if early speed alone was the contributing factor or if it was the overall credentials of the horses. just my opinion.

[clore1030]

I use a handicapping model based on the rankings as provided by RDSS. The program does have an export feature, but I prefer to do it by hand for a select criteria.

I have it broken down by distance and separate spreadsheets for each track and meeting.

Yes, it works for me. The manual input ensures that I notice certain factors. For example, in yesterday's Withers I used my model for the Aqueduct mile. The top-rated FW horse has been winning at a 67% clip and the top FX horse at a 71% clip.

The FX horse won and I got a $51.00 mutuel. The top FW horse ran second, the favorite was out of the money.

I handicap every day there's a fast track and just sit and wait for my spots. I've made only six wagers at AQU during the whole meet which ends today.

Now my question is whether I'll play the model horse in today's eighth. It's a 7F race and the top-ranked CPR and BL horses have been winning at a 67% clip. The same horse comes up in both categories (the 4 horse) as well as tops in FW and FX which has been winning at a 50% rate according to my model for 7F.

The issue is whether I'll get required odds after multiple scratches and whether I'm willing to apply dry track figs to today's sloppy track.

I'm pretty much mechanical in how I select pace lines, but I do leave some latitude for going beyond a "best of the last three comparable" based on whether a horse is approximating its max potential.

The key is patience and will power.

[pktruckdriver]

Originally Posted by clore1030

I use a handicapping model based on the rankings as provided by RDSS. The program does have an export feature, but I prefer to do it by hand for a select criteria.

I have it broken down by distance and separate spreadsheets for each track and meeting.

Yes, it works for me. The manual input ensures that I notice certain factors. For example, in yesterday's Withers I used my model for the Aqueduct mile. The top-rated FW horse has been winning at a 67% clip and the top FX horse at a 71% clip.

The FX horse won and I got a $51.00 mutuel. The top FW horse ran second, the favorite was out of the money.

I handicap every day there's a fast track and just sit and wait for my spots. I've made only six wagers at AQU during the whole meet which ends today.

Now my question is whether I'll play the model horse in today's eighth. It's a 7F race and the top-ranked CPR and BL horses have been winning at a 67% clip. The same horse comes up in both categories (the 4 horse) as well as tops in FW and FX which has been winning at a 50% rate according to my model for 7F.

The issue is whether I'll get required odds after multiple scratches and whether I'm willing to apply dry track figs to today's sloppy track.

I'm pretty much mechanical in how I select pace lines, but I do leave some latitude for going beyond a "best of the last three comparable" based on whether a horse is approximating its max potential.

The key is patience and will power.

Thank you for the response, great info, and great discipline too, as my discipline is still being learned, but when it is mastered, my winning percentage will go up dramatically too, no doubt, as I will bet only on the races I should, and pass on the ones where there is doubt.

patrick

[clore1030]

The forum table of contents says that the last post here was made by Patrick at 12:16AM today. I see the last post as being from three days ago and there's no indication that it was edited.

I only bring this up as I made a response to Patrick a few days ago that does not seem to be here either.

Is this thread closed because it's a survey thread? I guess I'll get an inkling if this post also disappears.

[Red Knave]

Originally Posted by clore1030

The forum table of contents says that the last post here was made by Patrick at 12:16AM today. I see the last post as being from three days ago and there's no indication that it was edited.

I only bring this up as I made a response to Patrick a few days ago that does not seem to be here either.

People responded to the poll without adding a reply. That changes the update date/time.
Not sure why your reply is missing.

[clore1030]

Originally Posted by Red Knave

People responded to the poll without adding a reply. That changes the update date/time.
Not sure why your reply is missing.

OK, thanks for the info. My response was mostly some additional tips for Patrick and the recommendation that he spend as much time on interpreting the data he yields as he would on handicapping, if not more.

I threw in some more stats from my own models just to show how one for 6 furlongs could vary greatly than one for 7 furlongs.

For example, at AQU, I had more winners at 6F coming from the top rated LPR horse (44%) than the EPR one (31%).

But at 7F, the top TPR rating delivered more than the top EPR and LPR combined.

I also pointed out that as the lowest two ranks of claimers yielded the most inconsistent results, I began to avoid charting them and the remaining figures became much more stable and thus my wagers much more profitable.

But one only uncovers such things by doing the grunge work. I had 30 years in analyzing and predicting audience levels in the broadcast TV industry. I was creating spreadsheets for tracking data about as soon as such were available. Those who moved up the ladder from initial data entry positions were those who interpreted as they were doing.

Anyone can compile stats, it's what you do with it that counts. As my research of the class structure confirmed, one can make the data more meaningful by realizing what is "noise" and what is valid.