Factor importance

[Ray2000]

I'll post this separate from the DR-TR thread. Sorry for the Hi-Jack Coach

Here are those Factor 'Importance' numbers by Track Size.
I don't have a large enough sample for individual tracks.

These are sorted on ROI which is more helpful than Win%
If you're after high Strike rates, bet the Crowd Fave or Morning Line Pick
Sorting on Win% does show Driver importance on the half mile track

If you want Impact values divide the Win% by 13 or the ROI by 0.71
That's the numbers for the Monkey Score (Random picks) in this data set.

I have starred the top 5 factors in each track size
My Speed is similar to Trackmaster and Morning Line varies wildly from track to track so both are excluded.


Races 3657 2 turns Mile or 7/8's
Win% ROI
17% 0.89 *Post
22% 0.86 My Speed Ave
28% 0.85 *TM Speed ave
16% 0.81 *Trainer
22% 0.81 *Earn$
34% 0.77 MoLine Pick
20% 0.75 *Class
19% 0.74 2YrWin%
17% 0.72 Class Drop
14% 0.71 Early Speed
16% 0.70 Driver
19% 0.69 Final Time
14% 0.64 4thQ Time

Races 6924 3 turn 5/8's
Win% ROI
36% 0.87 MoLine Pick
17% 0.85 *Post
23% 0.85 My speed Ave
22% 0.84 *Class
28% 0.83 *TM Speed ave
23% 0.82 *Earn$
16% 0.79 *4thQ Time
20% 0.76 Final Time
16% 0.73 Early Speed
17% 0.72 Trainer
18% 0.72 Class Drop
19% 0.71 2YrWin%
18% 0.67 Driver

Races 9196 4 turn 1/2 mile
Win% ROI
25% 0.86 My speed Ave
40% 0.82 MoLine Pick
30% 0.78 *TM Speed ave
22% 0.75 *2YrWin%
23% 0.75 *Earn$
23% 0.75 *Class
21% 0.74 *Final Time
21% 0.72 Driver
21% 0.72 Class Drop
19% 0.72 Early Speed
20% 0.72 Trainer
18% 0.72 4thQ Time
19% 0.69 Post

[sharkey11]

fantastic chart ray now thats some info that i can sink my teeth into :headbanger: ever hear of a guy up north in canada a professor of some sort he wrote a book harness gold or something like that in it there was a point count method in wich he used impact values to get plays but in todays game i think its dated it used mostly claming races of wich are few to be seen today with all the alowence races they have . thanks for the chart happy capping .

[Bogo]

I'll post this separate from the DR-TR thread. Sorry for the Hi-Jack Coach

Here are those Factor 'Importance' numbers by Track Size.
I don't have a large enough sample for individual tracks.

These are sorted on ROI which is more helpful than Win%
If you're after high Strike rates, bet the Crowd Fave or Morning Line Pick
Sorting on Win% does show Driver importance on the half mile track

If you want Impact values divide the Win% by 13 or the ROI by 0.71
That's the numbers for the Monkey Score (Random picks) in this data set.

I have starred the top 5 factors in each track size
My Speed is similar to Trackmaster and Morning Line varies wildly from track to track so both are excluded.


Races 3657 2 turns Mile or 7/8's
Win% ROI
17% 0.89 *Post
22% 0.86 My Speed Ave
28% 0.85 *TM Speed ave
16% 0.81 *Trainer
22% 0.81 *Earn$
34% 0.77 MoLine Pick
20% 0.75 *Class
19% 0.74 2YrWin%
17% 0.72 Class Drop
14% 0.71 Early Speed
16% 0.70 Driver
19% 0.69 Final Time
14% 0.64 4thQ Time

Races 6924 3 turn 5/8's
Win% ROI
36% 0.87 MoLine Pick
17% 0.85 *Post
23% 0.85 My speed Ave
22% 0.84 *Class
28% 0.83 *TM Speed ave
23% 0.82 *Earn$
16% 0.79 *4thQ Time
20% 0.76 Final Time
16% 0.73 Early Speed
17% 0.72 Trainer
18% 0.72 Class Drop
19% 0.71 2YrWin%
18% 0.67 Driver

Races 9196 4 turn 1/2 mile
Win% ROI
25% 0.86 My speed Ave
40% 0.82 MoLine Pick
30% 0.78 *TM Speed ave
22% 0.75 *2YrWin%
23% 0.75 *Earn$
23% 0.75 *Class
21% 0.74 *Final Time
21% 0.72 Driver
21% 0.72 Class Drop
19% 0.72 Early Speed
20% 0.72 Trainer
18% 0.72 4thQ Time
19% 0.69 Post

Thanks for posting this Ray. I couple of questions. How is Earn$ defined?
Why do you think there is such a difference between your speed and TM speed on a half mile?

[Ray2000]

Bogo

Earnings is total for current year and is not used in my current 'Bot' programs.
Jan-Apr are too unreliable, I pulled the numbers just to see how the factor would rank.

The difference between My speed calcs and TM's has always puzzled me. I know I'm using a different base track speed rating and adjust for pace and driver ability, but the difference is always there, fortunately in the right ROI direction.:)

[Bogo]

Is class figured out as an average like the speed ratings are?

[Ray2000]

Is class figured out as an average like the speed ratings are?



Yes, it is the average of the class listed in the past 3 or 4 races as found in the pplines.
I don't include lines older than 35 days, qualifiers, breaks etc.
All ties are broken by using the inside horse as Best in category.

On Class drop
I normally use a difference of 4 or more to be a true class up or down.
In this case I changed that to using the horse with the most drop points from the previous race to be called 'Best Class Drop' (same tie break) It's possible but rare that a race has all horses with the same number or all but one are class ups, so the "Best Drop" title goes to a horse that isn't dropping.:rolleyes:

[MutuelClerk]

I really like your chart Ray. Excellent work. Thanks for sharing. If i believed in Rep Power I'd give you like five points.

[Bogo]

Ray,

Have you taken a look at lowest odds last time or lowest odds at today's class as a factor?

[Ray2000]

Thanks MutuelClerk, keeps me busy:)

Bogo
My data base for 'Best on Tote' isn't as large as the other 'Best in x' factors.
Roughly 12,000 vs 20,000 races.

For race time Favorites
Track All Mile 5/8's 1/2
Races 12153 2523 4042 5588
Win% 41% 38% 38% 45%
ROI 0.84 0.85 0.81 0.85

I don't have Favorite in Last race data.

[BaconLover]

Ray, have you ever tried analyzing payouts? For example, when might it be better to play exa vs. tri vs. sup? How does this compare from track to track? Short fields vs large fields? Fields with 1 or 2 heavy chalks...or no solid favs? Etc.

Although I come across a lot of good cappers, I often find that many of them really don’t know how maximize their returns when they have say a solid 5-1 top choice or a live 20-1 bomber or a 4-5 fav that they think won’t hit the board.

I know that there are many good handicapping books out there, but I haven’t found a good book about maximizing returns when you have a winner picked. I don’t mean the basic overlay/underlay stuff, but something a bit more “Freakonomics”.

For example, for years I enjoyed playing the M1 p4 because the payout was normally significantly higher than parlaying 4 win bets. The reason for this was not necessarily because of the reduced p4 takeout. Instead, I strongly believed that the average player wagered 81 combinations or less when they played. All things being equal, these players statistically hovered around the top 3 favorites in each leg. As soon as you get an outlier that is 4th fav or worse, that kills a lot of tickets. If you have two such winners that are 4th fav or worse in a p4 sequence, the payout increases exponentially. If any heavy favs lose, that kills a lot of players who had singled those favs, further multiplying the payout, even if a 2nd fav won. I think the value declined when M1 switched from $1 to 50-cent mins since the average person could afford more combinations.

Anyway, just curious if you or anyone else has looked at the payout side of things.

[Ray2000]

BaconLover

I've done a lot of analysis on my own wagers but not on overall industry patterns.
The biggest complication is rebates.

This game we play has been getting tougher and tougher to turn a profit because :
1. the lack of new 'dumb' money and
2. the widespread use of computer board watching and data analysis.

Profit margins are so thin that without good rebates you're cooked.
Pools with high take-outs can be in play with rebates and tend to attract less 'sophisticated' players.

My own betting is on Pick 3's - Exactas - 20¢ Supers, in that order.
Superfecta returns depend on a few big hits so they're more of a throw-away type bet.

Your pick 4 comments are spot-on and apply to the pic3s as well.
Of all the ways I've looked at Pic3's the best pattern I've found is to single the leg with the horse that has your lowest line < 5/2 and use no more than 4 picks in the other 2 legs (Max 16 combo's)

“Freakonomics” might be there but impossible to uncover. :)

Best of Luck with your plays...