Horse Racing Forum - PaceAdvantage.Com - Horse Racing Message Board

Go Back   Horse Racing Forum - PaceAdvantage.Com - Horse Racing Message Board > Thoroughbred Horse Racing Discussion > General Handicapping Discussion


Reply
 
Thread Tools Rate Thread
Old 05-12-2020, 09:58 PM   #76
cj
@TimeformUSfigs
 
cj's Avatar
 
Join Date: Jan 2002
Location: Moore, OK
Posts: 46,773
Quote:
Originally Posted by steveb View Post
DRF is demonstrably wrong then.
My previous post should demonstrate why wrong.
Anybody that knows anything about time, will have expectation in one form or another, or how could they get a reasonable TV?
They are very outdated and were never any good in the first place. But for some reason, they still persist. I think guys like thaskalos demand it in the PPs.
cj is offline   Reply With Quote Reply
Old 05-12-2020, 09:58 PM   #77
steveb
Registered User
 
Join Date: Jan 2008
Location: Melbourne Australia
Posts: 886
Quote:
Originally Posted by CheckMark View Post
I was going to call you out on stating with a base 100 and minus 1 point for each 1/5th slower than 3 Year best time. I tried practicing with what CJ taught me and some numbers do not work?

you can call me out for anything, and that's fine.
if i have strong opinions, then i don't mind them being challenged, because, who knows, it may lead to me learning something new.


but if you can show me where i said to subtract 1 point for each fifth, then......
let me repeat....for winners it's ALL about ratios.



i am confident from reading this board that cj is very smart when it comes to time.
i am also confident, that i am too!!
steveb is offline   Reply With Quote Reply
Old 05-12-2020, 10:00 PM   #78
cj
@TimeformUSfigs
 
cj's Avatar
 
Join Date: Jan 2002
Location: Moore, OK
Posts: 46,773
Quote:
Originally Posted by CheckMark View Post
I was going to call you out on stating with a base 100 and minus 1 point for each 1/5th slower than 3 Year best time. I tried practicing with what CJ taught me and some numbers do not work?
Just so we are all clear here, the variant is a completely separate, very flawed number that is listed along with the speed rating. It isn't added, subtracted, or anything else. Some people do it anyway but it isn't in print.

Show me where the numbers don't work and I'll figure out if you insist on going down this route. It is like deciding you want to build a car and starting out with square wheels.
cj is offline   Reply With Quote Reply
Old 05-13-2020, 10:54 PM   #79
mikesal57
Veteran
 
mikesal57's Avatar
 
Join Date: Sep 2003
Location: NEW YORK CITY
Posts: 3,670
Quote:
Originally Posted by cj View Post
Just so we are all clear here, the variant is a completely separate, very flawed number that is listed along with the speed rating. It isn't added, subtracted, or anything else. Some people do it anyway but it isn't in print.

Show me where the numbers don't work and I'll figure out if you insist on going down this route. It is like deciding you want to build a car and starting out with square wheels.
Why would you say that when back in the 70-80-90's people were making more money then than today....horse racing was going well.....and they were using SR +TV

So with a so called superior figures people are not making as much and the industry is sliding further and further

people say technology makes things better......but in this industry ..it killed it

Back when I was a kid , I went to Aqu & Bel almost every weekend....

It made my day when I caught a few horses that paid over $15 and went home with much more than you came in with

Now a days ., you got the be happy you caught a few $6 horses and almost broke even...
mikesal57 is offline   Reply With Quote Reply
Old 05-14-2020, 12:56 AM   #80
cj
@TimeformUSfigs
 
cj's Avatar
 
Join Date: Jan 2002
Location: Moore, OK
Posts: 46,773
Quote:
Originally Posted by mikesal57 View Post
Why would you say that when back in the 70-80-90's people were making more money then than today....horse racing was going well.....and they were using SR +TV

So with a so called superior figures people are not making as much and the industry is sliding further and further

people say technology makes things better......but in this industry ..it killed it

Back when I was a kid , I went to Aqu & Bel almost every weekend....

It made my day when I caught a few horses that paid over $15 and went home with much more than you came in with

Now a days ., you got the be happy you caught a few $6 horses and almost broke even...
The game is certainly a lot tougher now than it was when I first started playing seriously in the late 80s, but that has nothing to do with the DRF Speed Rating and/or Track Variant falling out of favor. In fact, it is a huge part of why it has gotten tougher. I don't believe people "back in the day", whatever era that was, were making money using those numbers. The people making the money were using better numbers and they weren't publicly available. There have been high quality numbers around for A LONG TIME. They just weren't in the Morning Telegraph.

There are a lot of reasons the game is tougher, and at the very least a small part of that is the advent of better speed figures. You have to look a lot deeper now to find those $15 horses than just playing the top speed figure.
cj is offline   Reply With Quote Reply
Old 05-14-2020, 09:12 AM   #81
mikesal57
Veteran
 
mikesal57's Avatar
 
Join Date: Sep 2003
Location: NEW YORK CITY
Posts: 3,670
Quote:
Originally Posted by cj View Post
The game is certainly a lot tougher now than it was when I first started playing seriously in the late 80s, but that has nothing to do with the DRF Speed Rating and/or Track Variant falling out of favor. In fact, it is a huge part of why it has gotten tougher. I don't believe people "back in the day", whatever era that was, were making money using those numbers. The people making the money were using better numbers and they weren't publicly available. There have been high quality numbers around for A LONG TIME. They just weren't in the Morning Telegraph.

There are a lot of reasons the game is tougher, and at the very least a small part of that is the advent of better speed figures. You have to look a lot deeper now to find those $15 horses than just playing the top speed figure.
AMEN ...CJ

what ticks me a little...is that no body publishes the traditional SR + TV in their PP's.....Unless DRF still does .....so that the old timers can relate to stuff that they grew up with...maybe they are inferior but dont eliminate because they think so...a lot of people grew up with those numbers and should still access it.
mikesal57 is offline   Reply With Quote Reply
Old 05-14-2020, 10:28 AM   #82
Tom
The Voice of Reason!
 
Tom's Avatar
 
Join Date: Mar 2001
Location: Canandaigua, New york
Posts: 112,110
DRF published them in the paper version.
BRIS has them as an option in the data files.

What other PPs were available to you back in the day?
__________________
Who does the Racing Form Detective like in this one?
Tom is offline   Reply With Quote Reply
Old 05-14-2020, 12:02 PM   #83
mikesal57
Veteran
 
mikesal57's Avatar
 
Join Date: Sep 2003
Location: NEW YORK CITY
Posts: 3,670
Quote:
Originally Posted by Tom View Post
DRF published them in the paper version.
BRIS has them as an option in the data files.

What other PPs were available to you back in the day?
Back then I only got the DRF....and it was $3

I have a sub with HDW and didnt find it...
mikesal57 is offline   Reply With Quote Reply
Old 05-31-2020, 08:59 PM   #84
MJC922
Registered User
 
Join Date: Nov 2012
Posts: 1,478
Quote:
Originally Posted by classhandicapper View Post
I've experimented with using the PAR combined with various adjustments for lengths between the horses, but some erroneous conclusions are inevitable if you try to automate it. Sometimes a couple of average horses dominate a terrible field and sometimes they were actually monsters.

It helps if a few of the horses in the race have race records so you have some idea of the quality of the field, but if they are lightly raced, form can change dramatically from race to race

It also probably makes sense to look at the connections. If some million dollar first time starter from Baffert wins by 8 and there was a gap of 5 from 2nd to 3rd, you can probably be pretty sure the horse ran huge. You might come to a different conclusion if it was a 30-1 shot from a barn that doesn't handle top horses.

I think in these situations you almost have to look at the times of the race even though you know the speed figures will be less reliable than usual also.

Finally implemented a fix for this today. It won't be perfect but it should be solid enough for these types of races. I don't chase after these races myself but being way off (even once in awhile) is never a good thing. It may even help iron out any regional differences for some of the more isolated populations of 2-3 year olds over the winter when there might have been limited cross-competition.
__________________
North American Class Rankings

Last edited by MJC922; 05-31-2020 at 09:02 PM.
MJC922 is offline   Reply With Quote Reply
Old 06-01-2020, 09:28 AM   #85
classhandicapper
Registered User
 
classhandicapper's Avatar
 
Join Date: Mar 2005
Location: Queens, NY
Posts: 20,439
Quote:
Originally Posted by MJC922 View Post
Finally implemented a fix for this today. It won't be perfect but it should be solid enough for these types of races. I don't chase after these races myself but being way off (even once in awhile) is never a good thing. It may even help iron out any regional differences for some of the more isolated populations of 2-3 year olds over the winter when there might have been limited cross-competition.
I'd love to hear approximately what you are doing, but I don't expect you to give away any of your magic formulas.

I spent the whole weekend researching, tinkering, and testing my own formulas just for stakes races again. I now have 5 1/5 years of data.

The best I result I could generate came when I combined my class rating with the horse's speed figure if the horses was coming out of stakes race, but using just the horse's speed figure if it was coming out of any other type of race.

To really optimize it, I lowered the speed figure of the horse if he was moving up in class from ALW, CLM, MSW etc.. into a stake by approximate 1 length. I'm not sure why that exact amount worked best. It could partially be my class scale and part just the move up in class.

I'm certain the complexities and subjective errors in speed figures can be improved upon by looking at "who beat who by how much" in a class type analysis "and vice versa", but some races are easier to class than others in an automated way because there are literally thousands of conditions, purses, and combinations.
__________________
"Unlearning is the highest form of learning"
classhandicapper is offline   Reply With Quote Reply
Old 06-01-2020, 10:15 AM   #86
Tom
The Voice of Reason!
 
Tom's Avatar
 
Join Date: Mar 2001
Location: Canandaigua, New york
Posts: 112,110
5-1/2 year of data - times or Beyer speed figures?
__________________
Who does the Racing Form Detective like in this one?
Tom is offline   Reply With Quote Reply
Old 06-01-2020, 11:09 AM   #87
classhandicapper
Registered User
 
classhandicapper's Avatar
 
Join Date: Mar 2005
Location: Queens, NY
Posts: 20,439
Quote:
Originally Posted by Tom View Post
5-1/2 year of data - times or Beyer speed figures?
Formulator PPs, results charts, and all my own metrics and data for all the major tracks and any race card that included a Graded Stake if it wasn't held at one of the majors I cover.

The Beyer figures are part of the Formulator files.
__________________
"Unlearning is the highest form of learning"

Last edited by classhandicapper; 06-01-2020 at 11:13 AM.
classhandicapper is offline   Reply With Quote Reply
Old 06-01-2020, 06:30 PM   #88
MJC922
Registered User
 
Join Date: Nov 2012
Posts: 1,478
Quote:
Originally Posted by classhandicapper View Post
I'd love to hear approximately what you are doing, but I don't expect you to give away any of your magic formulas.

I spent the whole weekend researching, tinkering, and testing my own formulas just for stakes races again. I now have 5 1/5 years of data.

The best I result I could generate came when I combined my class rating with the horse's speed figure if the horses was coming out of stakes race, but using just the horse's speed figure if it was coming out of any other type of race.

To really optimize it, I lowered the speed figure of the horse if he was moving up in class from ALW, CLM, MSW etc.. into a stake by approximate 1 length. I'm not sure why that exact amount worked best. It could partially be my class scale and part just the move up in class.

I'm certain the complexities and subjective errors in speed figures can be improved upon by looking at "who beat who by how much" in a class type analysis "and vice versa", but some races are easier to class than others in an automated way because there are literally thousands of conditions, purses, and combinations.
The fix isn't much but it's ok, it amounted to three SQL passthrough queries, an injection of the par only for firsters and only on the first iteration. The first query populates a table with up to date maiden pars, second query neutralizes the maiden pars where the sample size is just too small, (yes even with a 13 year db there are still like 1.5% of these races) because you know I don't like shortcutting so everything is by track, seasonal, age, price, statebred etc, the one corner I cut was distance sprint/ route would've given a better early pace par but as rare as these races were it wasn't worth it IMO. The third query just matches up the firster with its par.

I'd been planning to do this for several years of mulling it over and honestly just avoiding the issue due to everything else life was throwing at me.

The purist would leave the par out of it and avoid lightly raced altogether, I can't really argue with that and to be honest in the beginning I had no solid pars then, only the usual estimated BS most people end up having to work with, but the data processing has gone on for so long now I can just query for the par and it's there. The problem with not doing anything was the computer requires several more races to finally 'see' enough data to get a good line on what really went down. When races were filled almost entirely with either very good maidens or very bad maidens it was a problem to not do anything, subscribers certainly deserved a better estimate than just wait-and-see. So I hope this helps. It should allow the class estimate to settle at a point closer to reality much sooner in the process whenever there's limited cross-competition.
__________________
North American Class Rankings

Last edited by MJC922; 06-01-2020 at 06:43 PM.
MJC922 is offline   Reply With Quote Reply
Old 06-01-2020, 09:03 PM   #89
classhandicapper
Registered User
 
classhandicapper's Avatar
 
Join Date: Mar 2005
Location: Queens, NY
Posts: 20,439
Quote:
Originally Posted by MJC922 View Post
The purist would leave the par out of it and avoid lightly raced altogether, I can't really argue with that and to be honest in the beginning I had no solid pars then, only the usual estimated BS most people end up having to work with, but the data processing has gone on for so long now I can just query for the par and it's there.
I love what you are doing.

One problem I have is with the data. I don't have a class field with clear cut consistent condition information.

One simple example.

If all basic NW1 ALW races were coded with track in one field, ALW in another field, NW1 in another field, and the purse in another field, it would be easy to know what I am dealing with.

The problem is the condition field contains the purse and other details. That creates thousands of conditions for a simple NW1 race even at the same track because the purses vary even for the same condition.

I've tried parsing out the purse, but there are still so many combinations it's a mess. Then I tried using just track/purse ranges, but even that has issues. It's just a monumental project for what seems like very little gain. It's not worth it when I can just look at the speed figure for those non stakes races and get better results than I've been able to get with my attempts so far.

In my real handicapping, that's a different story. I'll read the conditions, look back at the PPs of the horses in the field to gauge it's strength and so on. But for this automated testing I think I'm ready to throw in the towel going any further than I am now.
__________________
"Unlearning is the highest form of learning"

Last edited by classhandicapper; 06-01-2020 at 09:10 PM.
classhandicapper is offline   Reply With Quote Reply
Old 06-02-2020, 07:17 AM   #90
MJC922
Registered User
 
Join Date: Nov 2012
Posts: 1,478
Quote:
Originally Posted by classhandicapper View Post
I love what you are doing.

One problem I have is with the data. I don't have a class field with clear cut consistent condition information.

One simple example.

If all basic NW1 ALW races were coded with track in one field, ALW in another field, NW1 in another field, and the purse in another field, it would be easy to know what I am dealing with.

The problem is the condition field contains the purse and other details. That creates thousands of conditions for a simple NW1 race even at the same track because the purses vary even for the same condition.

I've tried parsing out the purse, but there are still so many combinations it's a mess. Then I tried using just track/purse ranges, but even that has issues. It's just a monumental project for what seems like very little gain. It's not worth it when I can just look at the speed figure for those non stakes races and get better results than I've been able to get with my attempts so far.

In my real handicapping, that's a different story. I'll read the conditions, look back at the PPs of the horses in the field to gauge it's strength and so on. But for this automated testing I think I'm ready to throw in the towel going any further than I am now.
Yea that's tough. I ended up having to write my own conditions parser. So basically I send the entire conditions into a parser multiple times and get back an answer for the different fields, example a male or female function, send the entire conditions to the function as a parameter, inside the function it will look for every possible combination of the key words i.e. FOR FILLES AND MARES, if that is found the function returns FEMALES, otherwise MALES.

So there's a function for race type, age, sex, statebred, claiming price etc. then there's one that builds a condition 'string' so that might return, NW1X or whatever it may be. I leave purse out of my par query. If it's AQU, WINTER, NW1X, FILLIES, 3YO, STATEBRED, the par should be solid so I don't want purse in there, that would just get in the way.

One problem I did have is the charts I use have a bug in them (or at least had a bug) which caused typos to appear in the text of the conditions so one of the first things I had to do was write a find and replace type of spell checker. lol
__________________
North American Class Rankings
MJC922 is offline   Reply With Quote Reply
Reply




Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

» Advertisement
» Current Polls
How do you primarily find bettable horses?
Make a formal odds line - 21.70%
23 Votes
Rank the horses in order of preference & compare it to the odds - 21.70%
23 Votes
Feel - 12.26%
13 Votes
Only play against short priced horses I dislike - 5.66%
6 Votes
Look for value oriented trips, angles, situations - 33.02%
35 Votes
Other (describe) - 5.66%
6 Votes
Total Votes: 106
This poll is closed.
Powered by vBadvanced CMPS v3.2.3

All times are GMT -4. The time now is 11:09 PM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright 1999 - 2023 -- PaceAdvantage.Com -- All Rights Reserved
We are a participant in the Amazon Services LLC Associates Program, an affiliate advertising program
designed to provide a means for us to earn fees by linking to Amazon.com and affiliated sites.