Post position bias???

[Storm Cadet]

Here's one to chew on:

I use equisim...now we tune the sim based on past stats including post position winning places right? And we all make our handicaping choices partly based on our choices post position...

So lets take DG's race at AQU yesterday for example.

Many people (including myself) wrote him off because of his wide post 11. But as the day went on 2 inside posts scratched out moving him in towards the rail. Did that increase his odds or views?

Now computer programs that use past stats cannot take this move into consideration..at least not mine....here's NYRA's post position stats as of today as posted on their site:

LEADERS


(For Races Less Than One Mile on the Main Track)
Post 1 2 3 4 5 6 7 8 9 10 11 12
Starts 191 191 191 191 190 180 153 124 90 60 41 21
Wins 35 26 29 19 20 17 17 7 13 2 2 5
Win% 18.32 13.61 15.18 9.95 10.53 9.44 11.11 5.65 14.44 3.33 4.88 23.81
(For Races One Mile or Longer on the Main Track)
Post 1 2 3 4 5 6 7 8 9 10 11 12
Starts 165 165 165 165 165 158 132 108 66 41 24 15
Wins 30 22 27 20 18 11 16 8 11 1 0 2
Win% 18.18 13.33 16.36 12.12 10.91 6.96 12.12 7.41 16.67 2.44 0.00 13.33
The difference between the 11 post and 9 post is huge (0% versus 16.67%)

and these are the stats for the main track...or thats what NYRA says: are they lumping in their stats inner and outer tracks?

And here's the kicker...after watching the race replay, I was watching the 2 horse throwing MIgs for a loop. and I see where NYRA did not use the 1 post so the 1 actually was the 2 post and so on so DG was really the 10 hole not the 11 or the 9...confused yet? :confused:

[PaceAdvantage]

SC,

I edited your post to make the stats easier to read. If ever you are trying to post formatted text, make sure you surround the formatted text with the CODE function. To get the code function, simply click on the # button on the post composition screen. It's located on the second row of editing icons four in from the left....

[Storm Cadet]

nt

[sjk]

Storm Cadet,

The number of races cited would indicate that these are stats from this winter's inner meet.

Software should let you mark the scratched horses and then adjust posts, likely pace scenario etc. If yours doesn't that could be a serious problem when there are scratches.

I'm sure you realize that with only 24 starts, there is not enough data to conclude that it is impossible to win from the 11 post.

[alysheba88]

Any post position analysis must take into the number of horses running so you can get a true picture of what their expected odds of winning are.

Clearly the one hole should win more % wise in a 5 horse field then the 10 hole in a 10 horse field

[Turfday]

My software does the following in the post position category:

NOTE: As an Equibase strategic partner, all of my data comes directly from them.

1) Takes into account every varied distance (dirt or turf) at a particular track and does NOT "clump" all sprints or all routes together.

2) Has three available search features: (1) all races at that particular distance since Jan. 1, 1995; (2) all races at that particular distance in the last 1,095 days (3) all races at that particular distance in the last 365 days.

3) It then breaks down every single FIELD SIZE run at that particular distance and surface...NO CLUMPING. This is because the dynamics of a race figure to change depending on the size of any given field.

4) It shows you how many races were run at that distance at that particular field size. It shows the EXPECTED VALUE, AVERAGE ODDS PER STARTER and AVERAGE ODDS PER WINNER for comparative purposes.

5) It then RATES each post position based on a proprietary formula that includes the field size, closing odds, finish position and recency of when that race took place (to help identify trends).

5) We also have an "in-the-money" post position page for each surface and distance that includes all of same above, highlighted by red bars = below average post position; yellow bars = average post position; green bars = above average post position.

* * * * * * * * * * * * **

Even though this wasn't in the discussion, I'll throw this out. Nobody wants to believe me, but as an example, at most, not all of the big tracks, post positions 5 & 6 generally UNDERPERFORM at 6 furlongs in field sizes of 11 or 12. Not the inside, not the outside. The MIDDLE.

[hurrikane]

That's very nice Bob,

does your software also shamelessly sell itself on paceadvantage.com? :D

[alysheba88]

Great info Bob. Dont consider that self promoting at all. Thanks for sharing

[hurrikane]

it's a joke...sheesh