TOGA Trends

[Tom]

After 3 weeks, some trends using BRIS data.

Dirt sprints
Looking at top 3 prime power, top 2 SR last race, 45% winners [33 races]
Bet $30 Ret $97.10 +67.10

Dirt routes
20 OUT OF 27 [74%] won by top 2 prime power
Bet $108 Ret $131.80 +23.80

Turf sprints
9 0f 18 races [50%] won by top 2 ACL
Bet $72.00 Ret $93.40 +21.40

Turf routes
CC and ACL the most predictive, but no profits

[BarchCapper]

Originally Posted by Tom

After 3 weeks, some trends using BRIS data.

Dirt sprints
Looking at top 3 prime power, top 2 SR last race, 45% winners [33 races]
Bet $30 Ret $97.10 +67.10

The others were clear to me - this one is a little fuzzier. I'm reading it as taking the top 3 prime power, and out of those top 3 adding the filters that they also needed to be in the top 2 speed ratings last race, and have a 45% winning percentage?

If that's right - is that winning percentage filter exclusively dirt sprints?

Thanks!

[Tom]

The winner ranked 1-2-3 PP AND 1-2 SR.
If the top 2 SR horse are not in the top 3 PP, no bet.

I calculated everything exclusively to dist/surf, but not all races were used. Races with a few FTS were not used, off turf with massive scratches, etc, but most "playable" races.

[BarchCapper]

Originally Posted by Tom

The winner ranked 1-2-3 PP AND 1-2 SR.
If the top 2 SR horse are not in the top 3 PP, no bet.

I calculated everything exclusively to dist/surf, but not all races were used. Races with a few FTS were not used, off turf with massive scratches, etc, but most "playable" races.

That's more clear - thank you!

[ranchwest]

Probably a dumb question, but what is ACL? And CC?

[CheckMark]

Originally Posted by ranchwest

Probably a dumb question, but what is ACL? And CC?

ACL-Average Competitive Level based on the Race Ratings when finishing in the money

CC-Current Class is based on horses' most recent starts with more importance on the most recent starts at or near today's distance

[ranchwest]

Originally Posted by CheckMark

ACL-Average Competitive Level based on the Race Ratings when finishing in the money

CC-Current Class is based on horses' most recent starts with more importance on the most recent starts at or near today's distance

Thanks!

[ranchwest]

Is the reference here to the actual BRIS Race Rating and Class Rating? Wouldn't it be better to use the Speed Par for both measures? I find that the listed Par Race Rating and Par Class Rating are often much different from the ratings found among the horses in the field. Any thoughts?

[Jeff P]

Tom, I ran a similar query for SAR dirt routes using top 2 HDW PSR.

UDM Definition:   SAR-DirtRouteTest
Divisor:          # UDM Def  Divisor: 999
Surface Req:      D*
Distance Req:     R

HDW PSR:          MinRank= 1  MaxRank= 2   MinVal= -999  MaxVal= 999   MinGap= -999  MaxGap= 999
Running Style:    ALL
Track:            SAR

Data Window Settings:
Divisor = 999      Odds Cap: None
Show on Text Report:  False
Filters Applied: -DATE: 7/14/2022 to 7/31/2022

Dirt (All*)  ROUTES  (From Index File: C:\2022\Q3_2022\pL_profile.txt)
     Track: SAR 

Data Summary          Win         Place          Show
-----------------------------------------------------
Mutuel Totals      130.80        111.70         89.50
Bet               -112.00       -112.00       -112.00
-----------------------------------------------------
P/L                 18.80         -0.30        -22.50

Wins                   20            32            33
Plays                  56            56            56
PCT                 .3571         .5714         .5893

ROI                1.1679        0.9973        0.7991
Avg Mut              6.54          3.49          2.71

From there I broke the above HDW PSR rank <= 2 data out by rank for a handful of other factors and pasted the breakout data into a text file: here.

By: CompoundAP Rank

Rank       P/L        Bet        Roi    Wins   Plays     Pct     Impact     AvgMut
----------------------------------------------------------------------------------
 1       28.00      34.00     1.8235      10      17   .5882     1.6471       6.20  
 2       19.30      20.00     1.9650       5      10   .5000     1.4000       7.86  

 3      -24.00      24.00     0.0000       0      12   .0000     0.0000       0.00  
 4        0.50      16.00     1.0313       3       8   .3750     1.0500       5.50  
 5       -6.40      12.00     0.4667       1       6   .1667     0.4667       5.60  
 6        0.00       0.00     0.0000       0       0   .0000     0.0000       0.00  
 7        1.40       6.00     1.2333       1       3   .3333     0.9333       7.40  
 8        0.00       0.00     0.0000       0       0   .0000     0.0000       0.00  
 9        0.00       0.00     0.0000       0       0   .0000     0.0000       0.00  
10        0.00       0.00     0.0000       0       0   .0000     0.0000       0.00  
11        0.00       0.00     0.0000       0       0   .0000     0.0000       0.00  
12        0.00       0.00     0.0000       0       0   .0000     0.0000       0.00

By: CompoundE2 Rank

Rank       P/L        Bet        Roi    Wins   Plays     Pct     Impact     AvgMut
----------------------------------------------------------------------------------
 1       28.70      32.00     1.8969       9      16   .5625     1.5750       6.74  
 2       10.20      20.00     1.5100       5      10   .5000     1.4000       6.04  

 3       -8.50      18.00     0.5278       2       9   .2222     0.6222       4.75  
 4       -2.80      20.00     0.8600       2      10   .2000     0.5600       8.60  
 5       -6.20      12.00     0.4833       1       6   .1667     0.4667       5.80  
 6       -0.60       8.00     0.9250       1       4   .2500     0.7000       7.40  
 7       -2.00       2.00     0.0000       0       1   .0000     0.0000       0.00  
 8        0.00       0.00     0.0000       0       0   .0000     0.0000       0.00  
 9        0.00       0.00     0.0000       0       0   .0000     0.0000       0.00  
10        0.00       0.00     0.0000       0       0   .0000     0.0000       0.00  
11        0.00       0.00     0.0000       0       0   .0000     0.0000       0.00  
12        0.00       0.00     0.0000       0       0   .0000     0.0000       0.00

Speaking from my own personal experience here:

It's not terribly difficult to create rules based spot plays that point out +ROI areas in the past.

I'm often amazed that seemingly obvious factors are not just overlooked, but to the extent they are sometimes overlooked.

That said, most of the time, when I subject rules based spot plays to forward testing:

The +ROI is seldom repeated any time soon.

But every once in a while, the shelf life is longer than you might think.

And I ask myself:

What confounding factors are out there making that possible?

Haven't subjected this to any kind of forward testing yet.

But either way --

Tom, interesting thread.


-jp

.

[Tom]

Dirt routes now 23 of 30 top 2 PP.

[ranchwest]

Jeff, have you analyzed the power number in combination with either jockey or trainer numbers? If so, do those findings also tend to regress to the mean? I ask because it seems to me that at Saratoga an unusually high percentage of races are won with top jocks and trainers. Admittedly, that is based on perception and not actual statistical studies.

[stok$$]

Hi, I have a dumb question... what does TOGA stand for?

[geroge.burns99]

Originally Posted by stok$$

Hi, I have a dumb question... what does TOGA stand for?

Saratoga

[ranchwest]

Or a party where people wear sheets and drink beer.

[stok$$]

Oh, now it makes sense. Thanks