Class Shifters

[Ray2000]

More useless information...

TrackMaster class shift of 5 or more from previous start.

9,501 class droppers
2,132 won (22.4%)... roi 0.83

11,601 class risers
809 won (7.0%)... roi 0.63

[tyler522]

Love it, keep it coming.Thanks

[LottaKash]

[
TrackMaster class shift of 5 or more from previous start.



Ray, 5 or more classes ???...Or too much Amaretto ?..

[Ray2000]

John

Trackmaster Class ratings, like a race last time rated 83 down to 78 this week.

Corresponds to about a 1 (sometimes 2) real class drop.

[LottaKash]

John

Trackmaster Class ratings, like a race last time rated 83 down to 78 this week.

Corresponds to about a 1 (sometimes 2) real class drop.

Ok Ray, thx for the cleanup there...

Also, perhaps you may want to take a further look at a horse rising up 1-class, when he is moving up into the top 1 or 2 classes at any given track...Especially so, when a horse, in has last 10 running lines or so, has been shown competing in that upper class somewhere, if not more than occasionally, in that new Class-Up... I think that you will get a nice strike/ROI number(s)... One of my faves..

[Ray2000]

Looks like a good angle, LK, I'll keep it in mind.

This is one area where humans beat robots hands down.
You and I can look at the race classification and judge if this is a class move or not.
But it takes a lot better programmer than me to get a robot to recognize if these 2 races are the same class

Example

BigM purse $6500 (Trackmaster 82)
Fillies & Mares N/W $3,500 in Last 5 Starts (N/W Last 3 Allowed $500)
Starters in a Class Higher Than N/W $6,000 Last Start or W/O $55,000 2016/17 Inel. Opt.Claim$10,000

Northfield purse $6800 (Trackmaster has one at 75, other at 78)
FILLIES & MARES NON WINNERS $4000 IN LAST 5 STARTS THAT ARE N/W $20,000 IN 2017.


At Yonkers, you might use purses as a guage
but sometimes Race Secretaries write a novel describing the conditions and robots are lost.

[traynor]

More useless information...

TrackMaster class shift of 5 or more from previous start.

9,501 class droppers
2,132 won (22.4%)... roi 0.83

11,601 class risers
809 won (7.0%)... roi 0.63

The essence of research is that the answers you get are contingent on the questions you ask. In this case (based on a LOT of similar research) the best question may be "what happens in the SECOND RACE after the class move?"

[pandy]

More useless information...

TrackMaster class shift of 5 or more from previous start.

9,501 class droppers
2,132 won (22.4%)... roi 0.83

11,601 class risers
809 won (7.0%)... roi 0.63


These class dropper numbers are not bad at all. Harness racing has such an extreme post position bias, that if you took out all of the horses from the 9,501 class droppers that had a bad post, the win % will definitely go up, and the ROI will most likely go up as well. I'm not sure what the best criteria is for bad posts, perhaps any post that wins less than 8%.

If you look at Buffalo, post 7 is winning at 2%, post 8, 1.5%. Post 6 is 8.8%. If you eliminate 7 and 8, all the numbers improve sharply.

[Jess Hawsen Arown]

Post 6 is the most winning post at the Meadowlands this meet. Other than half-mile tracks, best winning post position changes regularly each meet.

I find class change based on purse or condition incredibly irrelevant. The nw3,500 class at the Meadowlands is very wide and horses move up and down in field strength without changing class.

In addition, time of year is critical. From October through April (i.e., away from when most trainers keep their horses the sharpest), I find sharp horses moving up destroy class horses dropping down. However, that reverses during the hot weather when drop downs are more often caused by bad racing luck or temporary illness.

October begins FORM CHANGE season in harness racing.

[Postime]

Pandy makes an objective point here, to many variables within thousands of data points, track size, post bias,drivers,weather conditions, class is good angle when use as a winning level point, most horse's level off at a certain price range, the only thing TM is doing is trying to establish that point level on an individual race.

[pandy]

Pandy makes an objective point here, to many variables within thousands of data points, track size, post bias,drivers,weather conditions, class is good angle when use as a winning level point, most horse's level off at a certain price range, the only thing TM is doing is trying to establish that point level on an individual race.


Another point on this, Trackmaster's class ratings are not based on class and purse... they are based on speed ratings. Trackmaster basically averages out the speed ratings of all of the horses in the race for the last six months to create the class ratings.

Personally, I think that's a pretty efficient way to get a general idea of class levels. I prefer it over some quasi-artificial intelligence that attempts to create class ratings off of purses and class levels, which are hard to gauge.

Naturally, a handicapper who follows a circuit closely can spot hidden class moves within a class, which are the best...such as when two former class horses drop into a field and finish 1-2 several lengths ahead of the others...horses that finished behind them that are coming back in the same class are likely to be facing weaker.

[Stick]

Pandy,

Weren't you contemplating coming out with some type of system based on droppers at certain prices? Thought I remember you mentioning that a little while back.

[pandy]

Pandy,

Weren't you contemplating coming out with some type of system based on droppers at certain prices? Thought I remember you mentioning that a little while back.



Yes, I may still do that but I keep getting sidetracked. I have been doing well overall with droppers in both flats and trots.