Missing Class

[osophy_junkie]

What do you do when the class the horse is in either doesn't have pars or has too few races? The class latter doesn't appear to be linear by par time, so you can't really just pick the next claimg price up or down say. Do you look at other tracks that card similiar races and then find the classes with a close par times?

[classhandicapper]

You should look at the horses that are actually in the race instead of just the class label. If you are familiar with the general pecking order at your track, you should be able to figure out where they fit even if you don't have a PAR for that class.

The reason the class ladder PARs don't seem linear is mostly a function of sample size, but it's generally pretty efficient.

[sjk]

I have to add one or more pars nearly every day and do so assuming they are linear as to class levels. About 3 1/2 points for each increase in claiming level (an increase of 25% in claiming price).

[classhandicapper]

I have to add one or more pars nearly every day and do so assuming they are linear as to class levels. About 3 1/2 points for each increase in claiming level (an increase of 25% in claiming price).

How do you handle the odd ball multi conditional classes?

It's not that tough to figure out what's going on if you read the expanded conditions in the charts or at the top of a race, but the class description in the PP lines are often broad descriptors than don't tell you exactly which class you are dealing with.

For automated systems like you create, isn't that an issue or are you creating a table somewhere on your own where you read the conditions and say something like BEL, 10/1/2014, Race 5, PAR XXX.

[sjk]

I don't put much stock in any individual par and am sure that tracks that have very little interaction are out of sync with each other but since there is very little interaction that is not likely to be very costly.

I calculated a relationship between different classes at the same claiming level years ago and use it to help fill things in.

The only real use of the pars is to calculate par times and I don't need a lot of accuracy for that. Whether my pars are good enough to be a plus rather than a minus when shippers are involved is uncertain. I know my West Coast pars are too high but I don't bet those tracks and the horses rarely come East.

I am up to nearly 15,000 of them so I can't spend a lot of time concocting them or stewing over what doesn't look right. I could easily find something that looks out of line but it would not be value added work.

[classhandicapper]

I hear what you are saying on accuracy and use etc....

My question wasn't clear.

If a race says Clm25000B in a horse's PP lines, that could mean any number of things. Are you just accepting it as Clm25000B or do you have a way to determine what kind of Clm25000B race it actually was?

When I handicap manually, if for some reason I am interested, I just go to the chart and read the conditions. But for an automated process, there is no neat solution I can think of outside of setting up a table with track, date, race and a PAR. That's way more work that it's worth.

[sjk]

The data I use generally has a good representation of the condition. Here is a sample of some of the 4000 races at one track:

TRACK alw class a par type
BTP R4000NW1 9M 58.34407 C
BTP R4000NW13M 59.29534 C
BTP R4000NW13MX 61.01702 C
BTP R4000NW16M 56.89846 C
BTP R4000NW16MX 57.94524 C
BTP R4000NW19M 54.55709 C
BTP R4000NW19MX 58.0863 C
BTP R4000NW1L X 45.44835 R
BTP R4000NW1R Y 52.47072 C
BTP R4000NW1Y+ 46.48273 C
BTP R4000NW1Y+X 52.71347 C
BTP R4000NW2 9M 58.26407 C
BTP R4000NW23M 60.63165 C
BTP R4000NW23MX 58.99817 C
BTP R4000NW26M 60.73106 C
BTP R4000NW26MX 60.89799 C
BTP R4000NW29M 60.11368 C
BTP R4000NW29MX 61.79342 C
BTP R4000NW2L X 47.5841 R
BTP R4000NW2Y 56.95185 C
BTP R4000NW2Y+ 55.82472 C
BTP R4000NW2Y+X 55.35939 C
BTP R4000NW36M 62.61739 C
BTP R4000NW36MX 61.30359 C
BTP R4000NW39M 62.64925 C
BTP R4000NW39MX 61.86767 C
BTP R4000NW3Y+ 58.19024 C
BTP R4000NW3Y+X 61.42403 C
BTP R4000NW6M 56.58781 C

The open C4000 is at 61.4.

My latest pain is the optional claimers. I know they are more like an allowance race but I can parse the condition more easily using the claiming price so that is the par I am using even though I know it is not optimal.

[classhandicapper]

That looks like the raw Equibase class data (or similar). They generally don't put that kind of stuff in any PPs because people would freak out trying to understand it. :lol: They convert it to a more user friendly form that sometimes doesn't tell you everything you might want to know.

That's actually the kind of thing I need.

I could create a table like that, but I have no way to tie it to a specific race in the PP lines using track, date, and race unless I do it myself in another table (which would be monumental for many tracks).

[sjk]

I ask myself how in the world I ended up with 15000 class pars but if you do 2 a day for 20 years that is what you get. It would be a real piece of work to recreate it from scratch or to try to optimize it.

[classhandicapper]

I ask myself how in the world I ended up with 15000 class pars but if you do 2 a day for 20 years that is what you get. It would be a real piece of work to recreate it from scratch or to try to optimize it.

Do you have the ability to tie your table to the horse's PPs in an automated way?

That's the issue. Even if I had that table (or slowly created it for certain circuits) I can't tie it back to specific PP lines because what they give you in the PPs lines is not specific enough information.

[sjk]

I get my information from charts so tying data together is not an issue at all. I don't have the PPs you are looking at but make something equivalent in the course of the handicapping of a race in the program.