Bris RR vs actual class

[Light]

In my homemade software, I have a category that gives a horse credit for a class drop. So if a horse is in for 10k dropping from 15K,he gets credit.

In the last year, I have been looking closely at the Bris RR which gives a numerical rating for the strength of a race. Usually the higher the number the stronger the class with 1 point being significant enough for separation.

This has really driven the concept of class upside down for my decision to award a horse credit or not for a class drop.Because a lot of times a horse literally dropping in class gets the same or lower RR than a horse who is not.

For example the 4th at Mth yesterday had a real mix of class. Here is how I had them ranked by speed. Their RR is listed next to them, then the actual class they were rated from.This is just an example for discussion purposes. I am not necessarily miffed by the outcome of this race.

4 MTH 6F Clm 7500b 3^

Pgm Horse RR Class

8 CRAFTY CONC 110 Clm7500b
1 ISHIKAWA 109 Clm10000b
3 CITY GENIUS 109 Clm5000n2l
9 DAPPER TAPP 111 Clm10000n3l
2 PROWLER 111 Clm10000n1m
4 STRAIGHT AR 112 sAlw---s
5 SWASHBUCKLE 111 Clm7500n3l
1A PRIVATE RUL 110 Clm16000b
6 BELIEVE IN 112 sAlw---s
7 NICKY RED 113 sAlw---s


So the first question is,would you give a class drop to #1? Normally I would,but the RR is tied for the lowest in the field.That horse finished last as part of the favored entry,so the RR may have been accurate,saying that it was not a real class drop. The other half of the favored entry #1A certainly looked to be dropping but should I award him points? RR says no.The 110 RR is below the average RR of 111 for this field. This horse finished 8th. Was the RR right again or is this a broke down horse? (probably broke down imo). Notice how the #8 got the same RR at 7.5kb for 3yo & up as the #1a for 16kb 3yo & up, both at Mth. Is that believable? Only the #2,#4,#5 were not races from Mth. The order of finish was 9-8-6.

This is just an example of the dilemma of class I am finding today. I'm sure there are many different scenarios of RR's working and not working.

What is your opinion of how to handle RR's. Do you side more with the Bris RR's or the actual class designations to determine if a horse is actually dropping in class?

[PhantomOnTour]

Tricky race conditions.
All horses you mentioned (all 3yr olds)came out of beaten Clm races which, in effect, were n2l races for 4yr olds and up or open to any 3yr old.
This still doesn't explain the BRIS RR question, but I wonder if they base it on the price tag or the n2l condition or what.
In my PP Gen program I don't even view or print the BRIS RR or CR.
Never liked them...good charting homework(and good figs) is the best way of determining the strength of a race, imo.

Sorry, but to answer your question, I check the full race conditions and charts.

[classhandicapper]

Here are some of the problems with class designations:

1. There are an absolutely huge number of them across the country and no one can be familiar with all of them

2. There are strong and weak fields at each class level

3. There are strong and weak groups on any circuit at any given time (claimers, ALW, Stakes, statebreds)

4. There are strong and weak crops by age

5. There are strong and weak sections of the country at any given time

6. There is enormous variance of quality among maidens and other very lightly races horses that have not sorted themselves out yet.

I'm not sure how BRIS calculates their RR, but one of the cores of my handicapping is looking for horses whose form is better than it looks or better than reflected in their speed figures (and vice versa). For me it's a somewhat tedious process of actually looking at the PPs of the horses that were in the field, their trips, the race development, etc... and how horses are running coming out of the race.

If BRIS is calculating their ratings somewhat along those lines (at least looking at the actual horses in the field and whether it was a deep strong group or weak bunch), you are probably better off with their rating than the class designation.

I obviously think you could do WAY better by learning to classify horses on your own as I'm suggesting (at least at one circuit), but it's a lot of work.

[pandy]

I'd side with the RR ratings, but your program should be able to give points for a signficant class drop, such as from an open claimer to a n2l claimer. In that situation the RR's will probably be in line. Moving from a beaten 10 to a beaten 7.5 is not normally a significant drop.

[Light]

Do you agree or disagree with the Bris RR's I mentioned that gives a 7.5KB race at Mth the same RR as a 16KB race at Mth for the same distance, age and sex and conditions.Here are the races in question:

Mth 6/26/11 Race#7 6f 16KB

https://www.brisnet.com/secure-bin/brisclub/archives_pdf.cgi?type=arc&country=USA&track=MTH&date=2011-06-26&race=7

Mth 7/8/11 Race #7 6f 7.5Kb

https://www.brisnet.com/secure-bin/brisclub/archives_pdf.cgi?type=arc&country=USA&track=MTH&date=2011-07-08&race=7

I've checked how several of the horses in each race did subsequently since their performance in their respective races and the jury is still out.

[pandy]

I think Bris has it right but there are going to be times when the RR ratings are not going to be as accurate, that's the way it is with all ratings including speed ratings. But the RR ratings are very good. I use them all the time to help me pick longshots and they are very helpful. Are they flawless? No, no numbers are, but they are helpful.

[classhandicapper]

I don't use Bris. So I can't access your files.

I had the PPs for the 6/26 race, looked at that, and reviewed the charts of both.

It's pretty rare for a 7.5 race to be equal to a 16K race on the same circuit (all else being equal). The 16K race did not look particularly weak on the surface, but I am not overly familiar with those conditions at MTH. What I can tell you is that the 16K race was on the slow side and the 7.5K earned the faster Beyer. So it's possible that the Bris Rating is actually a combined speed/class designation rating.

They may start with a kind of PAR for the class designation and then tweak it up or down depending on whether the race was fast or slow for the class.

That would result in a situation where if two races were equally fast, the one with the higher class designation would get the better rating. If two races had the same class designation, the faster one would get the better rating. At the extremes, there would crossovers where a lower designation would get a higher rating.

I would suggest that you try to find out as much as you can about how they are calculating the ratings and decide if you think they makes sense and fit into whatever else you are doing. If the already include speed, that would have implications for your own weight of speed figures.

[Light]

I think Bris has it right

Why do you think Bris has it right in this case?

[cj]

Does BRIS assign a class rating before the race, or after? In other words, is one listed in the race heading prior to the running? If there is, does it get altered after the running?

[Capper Al]

Does BRIS assign a class rating before the race, or after? In other words, is one listed in the race heading prior to the running? If there is, does it get altered after the running?

I use BRIS class ratings. I agree with Pandy that they are like any other generated figure that will at times produce erroneous ratings. To answer CJ's question, BRIS does not give today's race a class rating until after the race. This is something I wish BRIS would do.

[Longshot]

For those that have the EquiSim program and use Multicap PPs there is a race
CR written in blue in the headed for each race. I'm not sure what Nathan's formula is for determining the CR for the race.

[Johnny V]

I use BRIS class ratings. I agree with Pandy that they are like any other generated figure that will at times produce erroneous ratings. To answer CJ's question, BRIS does not give today's race a class rating until after the race. This is something I wish BRIS would do.
Capper Al,
The Class ratings are a measurement of the actual performance of the horse in the past races and I would think that BRIS would be unable to give a class rating before the race. It would be like giving the speed rating of the horse before the race.

As far as the Race Rating goes it is kind of the same thing but measuring the performance of all the horses and then giving the rating of the race. I have talked to the people at BRIS sometime back about the ratings and to be honest I could not get anything worth of value from them about how the ratings are calculated. Apparently they do have range of Race Ratings for certain class levels that maybe could be used before the race if published but that was about it. What they are I have no idea and I don't know if giving a range before the race would really be of that much value.

[EJXD2]

Hi, all,
Good discussion on the ratings and their use in handicapping. Here (http://bit.ly/BRISratings) is a primer on BRIS Race and Class ratings.

Some highlights: The ratings are generated irrespective of the conditions of the race and final time. The program uses the "who beat whom and by how much" methodology and assigns that number to the race after the fact. These are revised daily, so race ratings can change as horses from that race continue to compete.

EdD

[cj]

I use BRIS class ratings. I agree with Pandy that they are like any other generated figure that will at times produce erroneous ratings. To answer CJ's question, BRIS does not give today's race a class rating until after the race. This is something I wish BRIS would do.

Yes, if you are going to have a rating, have something to compare it to is my thinking. My guess is that they are using the results though, not the way the field looked going in.

[classhandicapper]

The problem with looking at the results only becomes apparent after you classify horses for long enough.

In deep and competitive fields a lot of horses wind up getting trips they aren't comfortable with. Some of the speeds get outrun, some of the closers wind up further back than they usually are, the moves into contention are tougher and faster so some good horse don't finish as well as they usually do etc... So what happens is that a bunch of fit and good horses get totally buried and look way worse (and slower) than their actual ability.

The same thing happens in reverse in especially weak fields. Cheaper horses that normally get outrun get into contention, the moves are less competitive etc... and the horses look better and faster than they typically look in the same class under tougher conditions.

That's where the value is in class handicapping.

Everyone is looking at speed figures and finish positions, but they often have nothing to do with how well or fast a horse is capable of running when spotted against a field of different strength. Plus, since everyone is a speed handicapper these days they are all dealing with the same occasional errors.

To classify horses properly, you have to look at the quality of the field going in, the trips, race development, pace etc... (all of it) and then interpret the result. When the interpretation is inconclusive or there isn't enough information to draw a conclusion, you often have to see how the horses come back and then adjust your thinking after the fact.

What you sort of want is a "Race Rating" that tells you how strong the field was and "Performance Rating" that measure how well a horse ran within it.

[classhandicapper]

Hi, all,
Good discussion on the ratings and their use in handicapping. Here (http://bit.ly/BRISratings) is a primer on BRIS Race and Class ratings.

Some highlights: The ratings are generated irrespective of the conditions of the race and final time. The program uses the "who beat whom and by how much" methodology and assigns that number to the race after the fact. These are revised daily, so race ratings can change as horses from that race continue to compete.

EdD

In the documentation they say they also include a RACE RATING which a way of telling you how strong the field was. Perhaps that's what the original poster is really looking for.

If you are going to try to automate Class Horse Ratings, that comparative approach is probably best, but without including trips you are going to be way off base a lot of the time. That's one thing that has improved in my game over the years. The better I understand the trips, the better I can classify the horses and understand the strength of the field.

The other issue they are going to have is that many groups of horses do not cross over very often so there aren't many common opponents to compare. I think Class Pars help in that regard. Then you can rate fields relative to PAR and compare horses across groups that don't usually race against each other.

[JustCoolGene]

How BRIS Class and Race ratings are assigned:

This was explained to me maybe 12 years ago when BRIS was much more open to questions regarding their datafiles, reports, and figures.

In each race, the Bris computer looks at the recent Class and Race figures of the top finishers in that race along with the beaten lengths of each horse. The computer then calculates the Race Rating for that race along with the Class Rating for each horse. To check the validity of figures for that race, subsequent figures for each horse are then compared after they come out of said race to make sure they are in-line with the previous race. If figures for the top finishers dramatically change in subsequent races, then the computer re-adjusts the precious figures accordingly.

Before each race is run, BRIS assigns a class rating to each horse in each race. This is simply an average of last three class ratings. By averaging the last three, the class rating gives more weight to consistent horses. The theory is that if two horses each have an identical top class figure within the last three races, the consistent horse is better than a less consistent horse.

Bris supplies a Class Par for each race in the MultiCaps file. To get an idea how strong a field is, look at the average Class Ratings (last 3) of the top horses in the race and you get a clear idea how strong the race is. Then compare these figures to the par and you will know if the race is above or below Par according to Class.

Gene

[classhandicapper]

Sorry to be sort of dominating the thread, but this is my favorite topic. In reading the original question again and after reviewing the documentation, if you are trying to get at field strength and whether a horse is dropping in class or moving up in class you should be using the RACE RATING. That should be superior to any official class designation. If you are trying to measure a horse's performance within a race in a class oriented way (instead of using speed figures for example) you should be using the Class Rating. That's the way it appears to me.

[pandy]

Why do you think Bris has it right in this case?

Because the average class/speed of the horses, it was a weak field. I prefer the Race Ratings to the Class Ratings.

[sammy the sage]

Over 65% of the time I IGNORE class ratings...as in REALITY they are AFFECTED/EFFECTED more by trainer's smarts (some of which are not)...and owner's CASH pocket's (some ARE deep and some ARE not) & EGO...

For example today at "Toga' alone several dropping from 250k or 500k purchase price to a 50k or 25k claimer or even just won a 50k and dropping...

Trainer's take...owner unloading...horse just NOT stakes quality...get rid of 'em...I'll buy another one...I wanna see MY name in THE spotlight...

[Light]

RR's are suggesting to me that there is an even wider range of interchangeability in the various class designations than classical beliefs would lead you to think. This is what my point is about the 7.5kb race being equal to the 16kb race. It is breaking the stereotypes.

Here is another example also from Mth on Sat of the usefulness of RR's.Again these are ranked by speed.

2 MTH 1 1/16 Alw 50000n1x 3^

NAME RR CLASS

8 ITSAGOODTEN 115 Clm25000
6 BLUEMBER 110 MdSpWt
2 MARQUET REB 114 Clm14000
1 HUNG JURY 115 lw68200n1x
5 FIFTY SENSE 115 OC10000n1x
4 KING HENRY 115 lw54000n1x
7 JERSEY KISS 115 lw68200n1x
3 MAGICAL MYT 113 Clm25000b



#6 was the favorite with a huge RR inferiority to the field, finishing 5th. #8 who looked "outclassed" won at 7-1. He looked outclassed but the RR said he was on par with the allowance horses and the favorite was actually the outclassed one. To me the odds in this race were based on the public thinking en mass in the old classical way we were all brought up to think about class. The way Quinn's book about class said things were supposed to work.

I know the Bris RR's have their own flaws,but in this particular race they seemed indispensable.It was an easy toss of the favorite using them.


I have no idea of how Bris comes up with a numerical rating for class via RR's. But if someone has a formula for making homegrown ones it would be great.

[sammy the sage]

Light...a clue...horses CAN'T read #'s...they CERTAINLY have NO clue about WHAT class they're ABOUT to run...

oh...ps...and SOME jockey's ARE riding DRUNK or otherwise...

ps..ps..and I'm quite certain THAT some races selected TO BE entered were DONE at 10'pm w/a JUST liguid refreshment to help...

much LIKE the Author of THIS reply :lol: :D :bang:

[BIG49010]

When you look at class, I look at it from the Trainer's perspective, and being he is going from 10k to 7.5, I would say he thinks he is dropping slightly.

You are correct, that you have some tough calls at Monmouth, when you get these races that let multiple 3 yr old winners in them. Jim Quinn's book on Class Handicapping started me 25 years ago looking at results in these types of races, and where the winners came from. After a few thousand you'll get the hang of it.

[thaskalos]

RR's are suggesting to me that there is an even wider range of interchangeability in the various class designations than classical beliefs would lead you to think. This is what my point is about the 7.5kb race being equal to the 16kb race. It is breaking the stereotypes.

Here is another example also from Mth on Sat of the usefulness of RR's.Again these are ranked by speed.

2 MTH 1 1/16 Alw 50000n1x 3^

NAME RR CLASS

8 ITSAGOODTEN 115 Clm25000
6 BLUEMBER 110 MdSpWt
2 MARQUET REB 114 Clm14000
1 HUNG JURY 115 lw68200n1x
5 FIFTY SENSE 115 OC10000n1x
4 KING HENRY 115 lw54000n1x
7 JERSEY KISS 115 lw68200n1x
3 MAGICAL MYT 113 Clm25000b



#6 was the favorite with a huge RR inferiority to the field, finishing 5th. #8 who looked "outclassed" won at 7-1. He looked outclassed but the RR said he was on par with the allowance horses and the favorite was actually the outclassed one. To me the odds in this race were based on the public thinking en mass in the old classical way we were all brought up to think about class. The way Quinn's book about class said things were supposed to work.

I know the Bris RR's have their own flaws, but in this particular race they seemed indispensible.It was an easy toss of the favorite using them.


I have no idea of how Bris comes up with a numerical rating for class via RR's. But if someone has a formula for making homegrown ones it would be great.
As Tom Ainslie would say..."after every race, another 'system' is born".

[thaskalos]

IMO, there is no way of really knowing if there are indeed definitive "class" barriers which separate some of these horses...or if these class barriers only exist in the trainers' minds.

In every par chart that I have ever seen, the "real" difference between a $7,500 and a $10,000 claimer is, on average, about one fifth of a second...going 6 furlongs.

And yet...the vast majority of those $7,500 claimers - who dominate their own races in fast times and make the seemingly logical move one step higher up on the claiming ladder - not only fail to reproduce their usual performances at this slightly higher level...they are repeatedly beaten soundly, as if they were facing horses of much greater ability than themselves.

Two weeks later, they are again entered in their "normal" $7,500 lever...and their superior "form" is resurrected.

What was it that beat them so soundly in the $10,000 level? Why weren't they able to run their normal race -- pace-wise or final time-wise -- at a level only one step higher?

Was it the theoretical one-fifth of a second difference which separates the two groups?

Was it those blazing "in-between" tenths of a second, often found in the "classier" races, which beat those "outclassed" horses into submission?

In my opinion...it was NEITHER.

Those "cheaper" horses lost because their trainers didn't think that they could win!

They were entered not to win...but to be better prepared for the next "winnable" race.

That's why I am mystified by those bettors who claim that..."class doesn't matter."

CLASS MATTERS GREATLY...even if it exists only in the trainer's mind...

[Light]

As Tom Ainslie would say..."after every race, another 'system' is born".

Very true. I'm not that naive. Been playing over 20 years. I am still not totally sold on RR's and that's why I brought up the subject. But they are growing on me.

In my computer analysis the single most win producing factor is speed. The second major factor is class. (Pretty obvious). When Beyer came out with his book in the 70's about making speed figs it revolutionized the way we handicap.There was a big flurry of new ideas into the early 90's but since then its been the doldrums.

I think a numerical rating for class is a revolutionary concept.Didn't think it could work but RR's are showing me there is something there. Not between a 10K horse and a 7.5K horse but between a 16K horse and a 7.5K horse. This is just one example. Quinn would surely disagree. I've disagreed with Quinn's concepts of class ever since I read his books and today his concepts are even less relevant. This industry has changed from 20 years ago and I think RR's can become a valuable tool in a handicappers arsenal considering the state of the industry today. If someone wrote a book about creating them like Beyer did with speed figs back in the day, it would become a best seller.

[classhandicapper]

RR's are suggesting to me that there is an even wider range of interchangeability in the various class designations than classical beliefs would lead you to think. This is what my point is about the 7.5kb race being equal to the 16kb race. It is breaking the stereotypes.

Here is another example also from Mth on Sat of the usefulness of RR's.Again these are ranked by speed.

2 MTH 1 1/16 Alw 50000n1x 3^

NAME RR CLASS

8 ITSAGOODTEN 115 Clm25000
6 BLUEMBER 110 MdSpWt
2 MARQUET REB 114 Clm14000
1 HUNG JURY 115 lw68200n1x
5 FIFTY SENSE 115 OC10000n1x
4 KING HENRY 115 lw54000n1x
7 JERSEY KISS 115 lw68200n1x
3 MAGICAL MYT 113 Clm25000b



#6 was the favorite with a huge RR inferiority to the field, finishing 5th. #8 who looked "outclassed" won at 7-1. He looked outclassed but the RR said he was on par with the allowance horses and the favorite was actually the outclassed one. To me the odds in this race were based on the public thinking en mass in the old classical way we were all brought up to think about class. The way Quinn's book about class said things were supposed to work.

I know the Bris RR's have their own flaws,but in this particular race they seemed indispensable.It was an easy toss of the favorite using them.


I have no idea of how Bris comes up with a numerical rating for class via RR's. But if someone has a formula for making homegrown ones it would be great.


I may be misunderstanding you, but if you are trying to classify horses, there are two things going on.

1. Measuring the overall quality of the field compared to the norms for that class designation and other fields.

2. Measuring how individual horses ran within that race and what their present class is. The second is dependent on the first.

In other words, a Grade 1 race can be extremely deep and strong (like a Breeder's Cup race) and get a very high rating, but some of the horses in that race might be Grade 2 or Grade 3 horses and others could be above average Grade 1 horses. In either case, they would all be dropping in class if they faced an average Grade 1 field next, but the good ones would be standouts and the bad ones losers.

Looking at a Race Rating is telling you if a horse is moving up or down in class. That can be significant because changes in class mean changes in race conditions that can sometimes/often impact form.

Looking at a Class Rating is telling you how well the horse supposedly actually ran and what his current classification is.

These Bris Ratings look somewhat suspect to me on both counts, but I don't know them well.

I recently started a blog where I am reviewing Grade 1 races on dirt. I give each race a Class Rating (race rating) and compare it to the PAR for that race. So you can see if it was average or above/below average for the race and compare the rating to other Grade 1 races I have reviewed. I don't use a purely mechanical method that would be easy to program. At this point I am not giving out Ratings for individual horses, but you can usually infer what I am thinking from my comments. I started in May and have been adjusting/improving the ratings slightly to better reflect my thinking. You may enjoy looking at it. Feel free to ask questions.

http://www.classhandicapping.blogspot.com/

[sammy the sage]

IMO, there is no way of really knowing if there are indeed definitive "class" barriers which separate some of these horses...or if these class barriers only exist in the trainers' minds.

In every par chart that I have ever seen, the "real" difference between a $7,500 and a $10,000 claimer is, on average, about one fifth of a second...going 6 furlongs.

And yet...the vast majority of those $7,500 claimers - who dominate their own races in fast times and make the seemingly logical move one step higher up on the claiming ladder - not only fail to reproduce their usual performances at this slightly higher level...they are repeatedly beaten soundly, as if they were facing horses of much greater ability than themselves.

Two weeks later, they are again entered in their "normal" $7,500 lever...and their superior "form" is resurrected.

What was it that beat them so soundly in the $10,000 level? Why weren't they able to run their normal race -- pace-wise or final time-wise -- at a level only one step higher?

Was it the theoretical one-fifth of a second difference which separates the two groups?

Was it those blazing "in-between" tenths of a second, often found in the "classier" races, which beat those "outclassed" horses into submission?

In my opinion...it was NEITHER.

Those "cheaper" horses lost because their trainers didn't think that they could win!

They were entered not to win...but to be better prepared for the next "winnable" race.

That's why I am mystified by those bettors who claim that..."class doesn't matter."

CLASS MATTERS GREATLY...even if it exists only in the trainer's mind...

For EVERY example you post of the ABOVE...I can post the EXACT opposite :rolleyes:
That's why I am mystified by those bettors who claim that..."class doesn't matter." :sleeping: :lol:

[Capper Al]

"What is your opinion of how to handle RR's. Do you side more with the Bris RR's or the actual class designations to determine if a horse is actually dropping in class?"

Getting back to your question, my answer is yes! I look at the RR's like speed figures, and I look at the class designations in the traditional sense for class movements. If both methods confirm each other then it's looking pretty good that there is a class horse in the field. Or both methods might eliminate a horse as a pretender.

[Light]

I may be misunderstanding you, but if you are trying to classify horses, there are two things going on.

1. Measuring the overall quality of the field compared to the norms for that class designation and other fields.

2. Measuring how individual horses ran within that race and what their present class is. The second is dependent on the first.

To clarify my little chart of the horses with their RR's: That's how I handicapped the RR's before the race occurred.What you present in your blog is your numeric assessment (RR) of the horses after they run.

[InControlX]

[QUOTE=thaskalos]

And yet...the vast majority of those $7,500 claimers - who dominate their own races in fast times and make the seemingly logical move one step higher up on the claiming ladder - not only fail to reproduce their usual performances at this slightly higher level...they are repeatedly beaten soundly, as if they were facing horses of much greater ability than themselves.

QUOTE]

Well put! For my 2 cents I believe far too much emphasis is placed upon absolute race times and far too little on the "herd decisions" of pace in the race. Often, higher class horses produce mediocre times because the "herd pace" decided it. These are not inferior entries, just naturally slower paced races.

ICX

[Light]

Often, higher class horses produce mediocre times because the "herd pace" decided it. These are not inferior entries, just naturally slower paced races.

ICX

That's a very good point. Puts a dent in someone saying "I don't like that horse because he came out of a slow race". I have heard Matt C on TVG say that several times when I had my $ on just such an animal.Of course Matt C. changes his tune after the horse wins. The flaw is that if you go another race or two back on said horse, you can see that the horse has competed well in faster races and for some reason,the last race showing was run slow which also fools the speed figure makers.

[fmolf]

the pace of race ratings can tell you how fast the race was......the rr is calculated solely (according to bris website)on the cr of horses beaten and horses that beat this horse and not on pace or final times.So if a horse finished second at a higher class race than running in today,theoretically at least this should be reflected in its rr.They also say that its best use is in determining which horses faced the strongest fields in the classtheir racing in today(class within the class).Confusing to say the least and i am sure different ratings work better at some tracks than others.

[pondman]

In my computer analysis the single most win producing factor is speed. The second major factor is class. (Pretty obvious). When Beyer came out with his book in the 70's about making speed figs it revolutionized the way we handicap.

Just as you'll not be able to compute speed ratings intra- track, or compare speed ratings between S.A. and G.G., you'll have a difficult time comparing intra-track class. There are a number reasons for this, including local shenanigans.

When I bet a shipper, it's because I have the background data to say, this track has superior horses over another track, and an allowance horse at Sarataga is superior to an allowance horse at a casino track such as Zia. If the connections wants to spend the money shipping, they'll toy with and dust the field at a casino track. And it often doesn't matter how bad the horse's past looks, you've got to blindly bet the horse.

I would recommend not trying to normalize data within lower claiming horses. The industry has moved towards B style racing with older tired horses running against young horses. And unless you really can pick these, I would look elsewhere for entertainment and profit.

[classhandicapper]

To clarify my little chart of the horses with their RR's: That's how I handicapped the RR's before the race occurred.What you present in your blog is your numeric assessment (RR) of the horses after they run.

Yes, but not as relevant as you think.

It's kind of rare that I have a completely different opinion of the horses after the race than I had before the race. IMO, it's usually trips, pace, bias etc... that separate the otherwise reasonably equally matched horses. The only thing that will change my rating is if it's clear that several very good horses didn't fire at all or several improved sharply.

In addition, it's these kind of ratings that are the input to the subsequent race when I handicap.

So you could for example look at the rating I gave to Tizway's last two races and compare them to another Grade 1 winner and see who beat the better field or which field he beat was better.

Privately, I am thinking about all horses just like this, I simply don't have time to publish trips and ratings for every stakes race, let alone every horse in every race. Numbers like these WILL solve your problem and improve on both RR and CR.

[classhandicapper]

That's a very good point. Puts a dent in someone saying "I don't like that horse because he came out of a slow race". I have heard Matt C on TVG say that several times when I had my $ on just such an animal.Of course Matt C. changes his tune after the horse wins. The flaw is that if you go another race or two back on said horse, you can see that the horse has competed well in faster races and for some reason,the last race showing was run slow which also fools the speed figure makers.

The entire purpose of classifying horses is to try to avoid the accuracy issues and other pitfalls of speed figures when measuring the ability and performances of horses.

It's also easy to shoot holes in class (especially the way most people understand and measure it), but IMO it's a very useful way of looking at horses that these days may have more value than speed figures.

It doesn't have to be an either/or situation the way it has become in some circles. It's possible for one to reinforce the other, one may be more accurate and predictive in some circumstances than the other, one may clarify the other etc....

[Light]

classhandicapper

Your blog says this:

Explanation of Ratings and Methodology
Ratings are based on fractional times, final times, quality of competition the horses have been successful against in the past,

I'm sure Bris's RR's uses mathematical times as well. What I dislike about this method for evaluating class is that it is not a pure class number. It is at best a hybrid of speed ratings and its flaw can be seen with some upper class horses who produces crappy pace and final times,maybe checking in last. Then they takes a significant class drop and air a lower class field.

The point is where is the intrinsically pure class rating on that class dropper? You cannot use numbers on such an animal because this kind of horse's numbers are subject to class not the other way around. His numbers don't dictate what class he should be in. His class dictates what numbers he will produce.

So to me,this is the problem with making class numbers. Using a horses speed or fractional ratings takes out the purity of a class number and hybridizes it. That's why I am still not totally sold on the Bris RR's. I would say eliminate all pace and final times. That is the speed figures department. Positional and final calls,lengths gained or lost,final lengths behind,and level of competition in designated races might be the place to start. But a pure class number should be just that and keep the speed numbers out.

[thaskalos]

So to me,this is the problem with making class numbers. Using a horses speed or fractional ratings takes out the purity of a class number and hybridizes it. That's why I am still not totally sold on the Bris RR's. I would say eliminate all pace and final times. That is the speed figures department. Positional and final calls,lengths gained or lost,final lengths behind,and level of competition in designated races might be the place to start. But a pure class number should be just that and keep the speed numbers out.
I am not sure if pace and speed numbers should be completely left out when constructing a class rating...but there is definitely an element of class which cannot be measured by pace and speed.

I have, for many years, tried to augment the horse's actual performance figures by incorporating the figures of the RACE that the horse competed in...trying to create a COMPOSITE rating, giving the horse added (or less) credit according to the strength of the race itself -- and I have failed at every turn.

A $10,000 horse with impressive speed and pace numbers, coming out a highly rated (number-wise) $10,000 claiming race, still figures to lose to a sharp $12,500 dropper...even if the $12,500 horse has LESSER speed and pace figures...and is coming out of less impressive (number-wise) races.

But, I suppose, that Sammy "The Sage" can easily provide me with numerous examples of the exact opposite...:rolleyes:

[Elliott Sidewater]

I'm not taking anyone's side here, but here are a few thoughts:

1. not 100% sure whether BRIS RR's use time or speed or pace figures, but it appears to me that they do not. I am 100% sure that BRIS Prime Power numbers do use time/speed/pace. Not only that, it is also possible to convert differences in Prime Power ratings to projected beaten lengths at the finish in today's race!

2. The RR's appear to me to be damn accurate, when all horses are coming from the same track. I have found problems with overrating and underrating of shippers, not consistently, but enough to bug me.

3. Horses coming out of the race with the highest fastest total fractions E1+E2+final should never be eliminated without at least some consideration in today's race.

I am a longtime BRIS user - about 15 years.

Elliott

[thaskalos]

I'm not taking anyone's side here, but here are a few thoughts:

1. not 100% sure whether BRIS RR's use time or speed or pace figures, but it appears to me that they do not. I am 100% sure that BRIS Prime Power numbers do use time/speed/pace. Not only that, it is also possible to convert differences in Prime Power ratings to projected beaten lengths at the finish in today's race!

2. The RR's appear to me to be damn accurate, when all horses are coming from the same track. I have found problems with overrating and underrating of shippers, not consistently, but enough to bug me.

3. Horses coming out of the race with the highest fastest total fractions E1+E2+final should never be eliminated without at least some consideration in today's race.

I am a longtime BRIS user - about 15 years.

Elliott
Thanks for your input Elliott...as I haven't used the Bris numbers much myself...

Have you found their PACE figures (both early and late) to be of much use?

[classhandicapper]

classhandicapper

Your blog says this:

Explanation of Ratings and Methodology
Ratings are based on fractional times, final times, quality of competition the horses have been successful against in the past,

I'm sure Bris's RR's uses mathematical times as well. What I dislike about this method for evaluating class is that it is not a pure class number. It is at best a hybrid of speed ratings and its flaw can be seen with some upper class horses who produces crappy pace and final times,maybe checking in last. Then they takes a significant class drop and air a lower class field.

The point is where is the intrinsically pure class rating on that class dropper? You cannot use numbers on such an animal because this kind of horse's numbers are subject to class not the other way around. His numbers don't dictate what class he should be in. His class dictates what numbers he will produce.

So to me,this is the problem with making class numbers. Using a horses speed or fractional ratings takes out the purity of a class number and hybridizes it. That's why I am still not totally sold on the Bris RR's. I would say eliminate all pace and final times. That is the speed figures department. Positional and final calls,lengths gained or lost,final lengths behind,and level of competition in designated races might be the place to start. But a pure class number should be just that and keep the speed numbers out.

I need to be clearer in my explanation. Thanks.

In my case, the Race Class Rating is entirely based on the horses' performances relative to the norms for the class designation without looking at the final time. Only in very rare cases where I'm not sure how to interpret something subjectively might I look at the final time and fractions for clarification.

The Time Quality Rating is a combination of my subjective class analysis above plus the speed figure. It's an effort to find large discrepancies between the two where one or the other could be wrong or where pace and other considerations impacted the time.

So I am attempting to do what you want and to combine them as a separate exercise.

The real problem is the beaten horses. It's not so easy to assign class ratings for individual beaten horses because that gets into trips, bias, pace, race development etc... So rather than provide ratings, I provide trips, bias, and other notes that I then in turn use for comparison purposes.

In actual practice, I rarely actually assign numbers to all the horses in a race. I make comparisons between horses. It's easier for me to say horse "A" is clearly better, marginally better, similar etc... to horse "B" than it is for me to assign each of them a number.

[Light]

I'm not taking anyone's side here, but here are a few thoughts:

1. not 100% sure whether BRIS RR's use time or speed or pace figures, but it appears to me that they do not.

I stand corrected . I assumed they did. Bris says:


Final Time is NOT a factor used in calculating the ratings.

[classhandicapper]

Here's an example.

The Jim Dandy came up WAY faster than the Haskell, but IMHO those races are a lot closer in quality/class than the final times indicate. This edition of the Haskell was not a strong one, but the speed figure makes it look like a mid level ALW race. It was stronger than that.

[Tom]

BRIS agrees with you, Class - only a point apart in the speed ratings.
If you get the Travers day file, let's talk about the class ratings for the Travers and King's Bishop.... a couple of interesting things there, I thought.

[sammy the sage]

But, I suppose, that Sammy "The Sage" can easily provide me with numerous examples of the exact opposite...:rolleyes:

Happened MORE than ONCE today at MORE than ONE track TODAY...or EVERY day for that matter...certainly not gonna waste my time trying to validate a point that IS ONLY accurate say...70% of the time...from YOUR point of view...

I'm just VERY glad that there are MANY like you BETTING into the pools...MANY, many thanks... :p

[thaskalos]

Happened MORE than ONCE today at MORE than ONE track TODAY...or EVERY day for that matter...certainly not gonna waste my time trying to validate a point that IS ONLY accurate say...70% of the time...from YOUR point of view...

I'm just VERY glad that there are MANY like you BETTING into the pools...MANY, many thanks... :p
Your welcome...

I am always glad to see people winning big at the track. I use big winners like you as an example, which motivates me to try harder...so I can, one day, join your ranks.

[classhandicapper]

BRIS agrees with you, Class - only a point apart in the speed ratings.
If you get the Travers day file, let's talk about the class ratings for the Travers and King's Bishop.... a couple of interesting things there, I thought.

I'll get a chance to review those some time tomorrow.

I think the biggest issue in assigning a class rating to the Travers is going to be the same one I've been facing all season in the Triple Crown and other 3YO races. The 3YO crop does not look especially good relative to average and some of the better ones have already fallen by the wayside. But until a few start taking on older horses it's hard to know exactly where they fit. Some are still developing also.

This is one of those times where I will look at the speed figures in a more general sense. Any "one" or "two" figures can be way off for a variety of reasons. But when all the races are coming up slower than PAR, you more or less know they are a weak group. So while I might not look at the final time of the Travers to classify the race, the fact that so many of the spring and summer 3YO races have been slow will weigh in my decision.