Class ratings for Stakes races

[classhandicapper]

Originally Posted by cj

The more people deviate from the simple system Tom proposed, the less likely it will have any value in my opinion.

The problem is that class handicapping IS complicated. That's why almost no one is a class handicapper anymore.

No one wants to dig through a field and determine whether it was stronger or weaker than the norms for the class designation based on the accomplishments of the horses and their recent form. They'd rather look at whether the race was fast or slow according to the speed figures they use. It's easier and you can buy speed figures.

[Tom]

All I am doing it for is say I hsve the race down to three horse, all ran a 98 last time out.

6-5 ran 98 in 3 stakes.
3-1 ran a 98 in a 3 stakes
6-1 RAN A 98 in a 2 stakes

Who ya gonna call?

[cj]

Originally Posted by classhandicapper

The problem is that class handicapping IS complicated. That's why almost no one is a class handicapper anymore.

No one wants to dig through a field and determine whether it was stronger or weaker than the norms for the class designation based on the accomplishments of the horses and their recent form. They'd rather look at whether the race was fast or slow according to the speed figures they use. It's easier and you can buy speed figures.

Of course it is...my point was (and I see you agreed) once you start adding in speed, they basically turn into a modified speed figure. We've discussed this before, you know I'm against using any speed figures in class ratings.

I'd take it farther and say even pars and not a good idea. It should be based on money. That can get complicated with slots tracks and statebreds, but it can be done if one really wants to dive in.

[classhandicapper]

Originally Posted by thaskalos

IMO...the class factor loses "value" when it is closely associated with speed variables. I believe that the different handicapping factors should be contemplated SEPARATELY...not lumped together.

Here's a simple example of the way I do it.

Below are the top 5 finishers in a race and any other horse that ran very competitively before tiring

Finish Position - Points

1 - Grade 1 winner (111 points)
2 - Grade 1 placed (108 points)
3 - Grade 3 winner (102 points)
4 - Grade 2 placed (102 points)
5 - Grade 3 placed (99 points)
8 - Grade 2 winner (105 points)

I then use those point assignment to create a final rating. I give more weight in the rating to the winner than the 2nd place finisher, the 2nd place finisher than 3rd place finisher and so on down the line. Then I produce a weighted average.

Th tricky part is what I described previously.

What do you do if a horse was a Grade 1 winner 10 races ago but has been getting beaten up in Grade 3 races lately? Does he really deserve to be called a Grade 1 quality horse now?

What if he was a Grade 1 winner last year, but his first 2 races off the layoff were terrible.

What if he was Grade 1 winner on dirt, but today's race was on turf?

You need rules to cover all the situations where your own subjective thinking tends to take over if you want to automate it all.

[classhandicapper]

Originally Posted by cj

Of course it is...my point was (and I see you agreed) once you start adding in speed, they basically turn into a modified speed figure. We've discussed this before, you know I'm against using any speed figures in class ratings.

I'd take it farther and say even pars and not a good idea. It should be based on money. That can get complicated with slots tracks and statebreds, but it can be done if one really wants to dive in.

I've looked at money. If anything I came away with the conclusion that purses often have no relationship to the actual ability of the horses even at the same circuit, let alone others (even when controlling for slots and statebreds).

IMO you'll wind up with a lot of really bad ratings using just purses.

The most obvious example I found is that 2yo and 3yos will be slower than older horses (especially in claimers), but races limited to them tend to have similar purses to those for 3+ or 4+.

The other complication is that purses for the exact same quality of horse can be different at different tracks, but you can't make adjustments based on purses alone because costs are also different. I tried to locate cost information for various circuits, but the info is sketchy.

The only reasonable way to do it is with PARs and by monitoring how horses do when they cross circuits and various classes. With just stakes though, it's a walk in the park. It's very easy.

[cj]

Originally Posted by classhandicapper

I've looked at money. If anything I came away with the conclusion that purses often have no relationship to the actual ability of the horses even at the same circuit, let alone others (even when controlling for slots and statebreds).

IMO you'll wind up with a lot of really bad ratings using just purses.

The most obvious example I found is that 2yo and 3yos will be slower than older horses (especially in claimers), but races limited to them tend to have similar purses to those for 3+ or 4+.

The other complication is that purses for the exact same quality of horse can be different at different tracks, but you can't make adjustments based on purses alone because costs are also different. I tried to locate cost information for various circuits, but the info is sketchy.

The only reasonable way to do it is with PARs and by monitoring how horses do when they cross circuits and various classes. With just stakes though, it's a walk in the park. It's very easy.

I obviously wasn't talking only money. You consider sex, age, surface, distance, conditions, claiming levels, etc....just NO speed.

[classhandicapper]

Originally Posted by cj

I obviously wasn't talking only money. You consider sex, age, surface, distance, conditions, claiming levels, etc....just NO speed.

I've been looking at this for years, every approach is a minefield.

[Clocker]

Originally Posted by classhandicapper

I've been looking at this for years, every approach is a minefield.

Of all the classic handicapping factors (speed, pace, form, etc.), class is the most subjective. Perhaps the minefield is trying to objectively quantify a subjective, qualitative factor.

Tom's class rating here is more qualitative than quantitative. Any attempts in this thread to make it more quantitative have been ambiguous if not unsuccessful. Perhaps the answer is to use that and similar metrics as filters to identify contenders rather than as "figures" in the same sense as used for speed or pace.

Quinn's Condition Book is the standard on the subject. Neither there nor in his Figure book does he attempt to come up with an objective, computed class figure.

[raybo]

Originally Posted by cj

Of course it is...my point was (and I see you agreed) once you start adding in speed, they basically turn into a modified speed figure. We've discussed this before, you know I'm against using any speed figures in class ratings.

I'd take it farther and say even pars and not a good idea. It should be based on money. That can get complicated with slots tracks and statebreds, but it can be done if one really wants to dive in.

I agree, a class "assessment" or "classification" should include money earned, and pace/speed etc, should be separate.

That being said, if one is going to use a class "figure" to rank, grade, or compare horses, then one almost has to include factors other than money earned. I personally calculate a class figure which includes money earned, an overall performance figure, and a form cycle component. I also have another ranking that combines class with a distance component, so I get money, performance, form, and distance, all put together in a single ranking method. Of course, all my class and distance ratings are further defined by surface, so surface is a big part of it all, too.

Many times when I just can't figure out how a horse did well in a race, the only thing that explains it is my class stuff. "Class" can, and often does, take pace and speed to school.

[classhandicapper]

IMO the idea behind class handicapping is to have an alternative approach to measuring performance because the accuracy and complexity issues of pace/speed figures can occasionally lead you astray. Over and above that, IMO there are more subtle things going in races that aren't measured well with time.

The flip side is that race and performance quality measurement can get tricky without an objective measurement like "time", especially when we are talking about lightly raced or inconsistent horses that haven't sorted themselves out yet. Maiden and limited ALW race quality is all over the map and difficult to get at without time.

IMO class and time analysis should be done separately, but the ultimate answer to how strong a field was or how well a horse ran can and IMO should often be a blend of both. They are independent pieces of evidence to draw upon to reach a conclusion.

I think what Tom is trying to accomplish (and what I've been trying to add to) is how to go about getting at field quality in the non speed/pace figure way. To that I have been adding some of the problems that come up much the same way figure makers have to deal with wind, run up, rails, track maintenance etc.. There is a different set of issues for qualitative analysis of field strength that have to be addressed to improve accuracy.

Purses often don't reflect quality. There are issues with statebreds, 2yo and 3yo development, costs structures are different at different circuits, there are constantly evolving claiming and other conditions etc....

In the end though, I look at both.

[cj]

Originally Posted by classhandicapper

In the end though, I look at both.

Agree, but this is also why I think it is important the two not be related to each other.

[raybo]

Originally Posted by classhandicapper

IMO the idea behind class handicapping is to have an alternative approach to measuring performance because the accuracy and complexity issues of pace/speed figures can occasionally lead you astray. Over and above that, IMO there are more subtle things going in races that aren't measured well with time.

The flip side is that race and performance quality measurement can get tricky without an objective measurement like "time", especially when we are talking about lightly raced or inconsistent horses that haven't sorted themselves out yet. Maiden and limited ALW race quality is all over the map and difficult to get at without time.

IMO class and time analysis should be done separately, but the ultimate answer to how strong a field was or how well a horse ran can and IMO should often be a blend of both. They are independent pieces of evidence to draw upon to reach a conclusion.

I think what Tom is trying to accomplish (and what I've been trying to add to) is how to go about getting at field quality in the non speed/pace figure way. To that I have been adding some of the problems that come up much the same way figure makers have to deal with wind, run up, rails, track maintenance etc.. There is a different set of issues for qualitative analysis of field strength that have to be addressed to improve accuracy.

Purses often don't reflect quality. There are issues with statebreds, 2yo and 3yo development, costs structures are different at different circuits, there are constantly evolving claiming and other conditions etc....

In the end though, I look at both.

Good post! I agree that purse alone is a poor strategy, but when combined with a good performance rating, a current form rating, a distance capability rating, and a surface rating, then the inconsistency in purse values is leveled out pretty well. But, if you're just trying to get at logical contender classifications, regarding the field, Tom's class method is probably pretty good. Then the tough part starts, you have to separate those contenders, and that is where other factors have to be included in the overall method, IMO.

[classhandicapper]

Originally Posted by cj

Agree, but this is also why I think it is important the two not be related to each other.

Agree.

I don't have a problem using PARs for this because even though they are derived from figures, they are an average that's just being used to come up with the general pecking order of the class strcuture, not to rate a specific race on class.

They give you a numeric scale to work with that lets you say for example a Grade 1 race for older males = X on average and so on down the line.

If the class rating is on the same scale as the speed figures, it makes it easy to say whether the race was faster or slower than average for the class in addition to saying it was of higher/lower quality than average in non figures terms.

It just always made the process easy for me to think about.

The Test looked like a high quality race on paper, but it came up slower than average. So my thinking might be it was probably not as good as it looked on paper, but perhaps of higher quality than the time indicates. That's not saying that the speed figure is wrong, just that perhaps something in the development made it come up a little slow while perhaps those 3yo sprint fillies are not a great crop. My guess would be that reality is somewhere in between.

[Billnewman]

If I were going to incorporate Tom's rating system into my sheets I would use 3 numbers.

#1 the raw number as it stands

#2 the raw number multiplied by a competitiveness number 1,2,3,4,or 5. 5 being the best and than divide by 4. So if the class number is 4 and the competitiveness number is 3 than 4x3=12 divided by 4= 3

#3 multiply the second number by average pace in feet/second and than divide by 30 So 3x55.6=166.8 divided by 30= 5.5

So my finial numbers are 4 - 3 and 5.5

[PressThePace]

There's way too much "good" information in this thread.....I'm hoping two of you can piss each other off and start throwing around insults, then PA can boot the thread. One can only hope...