Post position 1 spot play

[Bill Cullen]

Here's a system I developed some years ago inspired by some findings from Mike Nunamaker's "Modern Impact Values."

For races at 9 furlongs or longer, both maiden and non-maiden races, as well as dirt and turf races are playable. For three years olds and up:

1) The horse must be starting from the number 1 post position.
2) There must be at least 11 horses in the field.
3) The horse must be going off at 6/1 odds or higher.

The rationale for the system: Four out of five atomic samples for routes in "Modern Impact Values" show higher-than-expected ROIs with relatively low impact values for horses starting from the number one post. The sprint atomic samples tend not to show this characteristic. I speculated that the higher-than-expected ROI might be a function of the horses breaking from the number one post having to cover, on average, less ground than horses breaking from the outside posts and that this effect would be magnified as the distance of the race increased. Thus I have a filter for only considering playable races at 9 furlongs or further.

I remember reading in Quirin's "Winning at the Races" many years ago that horses breaking from posts number 9 and higher are particualrly vulnerable in routes. Knowing that large fields have these extra horses breaking from these outside vulnerable post positions, and also knowing that these extra horses will be building the odds for the other horses and competing with the inside horse in the one post, this was my thinking for the 11-horse field or higher.

The odds rule of 6/1 or higher was determined by the fact that the winning percentage for the horses in the route samples in "Modern Impact Values" was 13.5% or about 1 out of every seven plays.

That's it.

[pmd62ndst]

Yikes. Out of nearly 70,000 races, I only have 287 that met that criteria (3 Yr & Up, 9 furlongs or more, 11 or more starters, Post position 1 is at odds of 6 or greater).

$ -74.70 (-13.0%) 65/287 (22.6%)

Here are the same results for the other horses who were not in Post position 1:

$ -649.00 (-13.3%) 530/2448 (21.7%)

Yes, there is a microscopic sliver of advantage for Post position 1, but definitely not enough to make this a spot play.

PMD

[Bill Cullen]

Originally posted by pmd62ndst
Yikes. Out of nearly 70,000 races, I only have 287 that met that criteria (3 Yr & Up, 9 furlongs or more, 11 or more starters, Post position 1 is at odds of 6 or greater).

$ -74.70 (-13.0%) 65/287 (22.6%)

Here are the same results for the other horses who were not in Post position 1:

$ -649.00 (-13.3%) 530/2448 (21.7%)

Yes, there is a microscopic sliver of advantage for Post position 1, but definitely not enough to make this a spot play.

PMD

I think you got your math wrong for the post postion one horses:

65 wins multiplied by $14.00( the minimum winning payoff for 6/1 odds or higher) = $910.00.

$910.00 minus $574.00 (287 plays times 2 dollars) = $336.00.

$336 divided by $574 = .58

That's a 58% positive ROI by my reckoning.

Let me know what you think.

Bill Cullen

[RXB]

The % wins are wacko-- way too high. Obviously, something went haywire. I wonder if he mistakenly used a < 6-1 odds filter instead of 6-1+.

[Buckeye]

Total Bets 118
WIN PLACE SHOW
Total Amount Bet 236.00 236.00 236.00
Wins 1 12 23
Pct. 0.85% 10.17% 19.49%
Amount Won 15.60 147.40 178.00
Profit/Loss -220.40 -88.60 -58.00
Pct Profit Loss -93.39% -37.54% -24.58%
Avg. Payout 15.60 12.28 7.74

[Bill Cullen]

Originally posted by RXB
The % wins are wacko-- way too high. Obviously, something went haywire. I wonder if he mistakenly used a < 6-1 odds filter instead of 6-1+.

Yes, the win percentage seems way too high.

[Bill Cullen]

Originally posted by Buckeye
Total Bets 118
WIN PLACE SHOW
Total Amount Bet 236.00 236.00 236.00
Wins 1 12 23
Pct. 0.85% 10.17% 19.49%
Amount Won 15.60 147.40 178.00
Profit/Loss -220.40 -88.60 -58.00
Pct Profit Loss -93.39% -37.54% -24.58%
Avg. Payout 15.60 12.28 7.74

PMD found only 287 plays out of 70,000 races and you find a 118 plays since August of 2003.

Something doesn't jibe here.

Also your win percentage seems incredibly low compared to anything you'd expect from the samples in Modern Impact Values.

Sorry, my friend, but for me the jury's still out.

Thanks for doing the research, though, and being kind enough to share it.

Bill Cullen

[Buckeye]

Post #1 appears to be a disadvantage because

Post Position NOT one
Field Size 11 or more
Odds of 6-1 or higher
Age 3up or 4up
Distance 9 furlongs or more

Total Bets 1407
WIN PLACE SHOW
Total Amount Bet 2814.00 2814.00 2814.00
Wins 63 141 243
Pct. 4.48% 10.02% 17.27%
Amount Won 1968.40 1889.20 2038.00
Profit/Loss -845.60 -924.80 -776.00
Pct Profit Loss -30.05% -32.86% -27.58%
Avg. Payout 31.24 13.40 8.39

[RXB]

The win % in Buckeye's first sample is aberrant, but the place and show totals aren't, considering the 6-1+ odds filter.

And considering that it is able to throw a 1/118 streak at you, do you really think that this might be a worthwhile angle?

[Buckeye]

Originally posted by RXB
And considering that it is able to throw a 1/118 streak at you, do you really think that this might be a worthwhile angle?

In a word NO.

Unless you're laying the one.

[Bill Cullen]

Originally posted by RXB
The win % in Buckeye's first sample is aberrant, but the place and show totals aren't, considering the 6-1+ odds filter.

And considering that it is able to throw a 1/118 streak at you, do you really think that this might be a worthwhile angle?

I stand by what i said before:

One winner out of 118 races is not just different in degree but is totally different in kind and absurdly low from anything that I would normally expect starting from the initial conditions.

I'd like to see what PMD's figures show when he re-does his math and/or fixes his typos.

Thanks,

Bill Cullen

[RXB]

I had a 41-race streak two years ago where I had 13 seconds, 8 thirds, but not a single winner.

I was just about ready for the bridge. Now imagine a 1-for-77 skein tacked on top of that. {Splash}

Realistically, I'd expect the win % long-haul to be about 5%.

[Bill Cullen]

Originally posted by RXB
I had a 41-race streak two years ago where I had 13 seconds, 8 thirds, but not a single winner.

I was just about ready for the bridge. Now imagine a 1-for-77 skein tacked on top of that. {Splash}

Realistically, I'd expect the win % long-haul to be about 5%.

You might well be right. I hope PMD and/or someone will crunch the numbers for additional data. An unavoidable reality with these micro-spot plays is that it takes almost forever to get a good baseline when there are so few plays over the course of a year.

Bill Cullen

[Buckeye]

Post Position #2

Total Bets 115
WIN PLACE SHOW
Total Amount Bet 230.00 230.00 230.00
Wins 7 11 22
Pct. 6.09% 9.57% 19.13%
Amount Won 220.30 141.20 153.10
Profit/Loss -9.70 -88.80 -76.90
Pct Profit Loss -4.22% -38.61% -33.43%
Avg. Payout 31.47 12.84 6.96

Seriously, these win percentages are just too low, notwithstanding, tDAT tTRK nRAC tPRG tHOR nWIN nFIN nDIS nAGE nFLD nODDS nPPO
4/7/04 KEE 7 2 Literacy $77.60 1 9 1 11 37.8 2

[RXB]

There might be individual distances at certain tracks where it might do okay, but across the board I can't see it. Besides, ultra-low win percentages don't work-- eventually you will hit a vicious losing streak that savages your bankroll and your confidence, a la Buckeye's sample. Gotta get the wins around 20%. (That's about where I'm at, and I've still had that 41-race skid.)

And really, single factors can only be expected to get you closer to break even, not produce anything like a profit. It's being able to apply knowledge to the specifics of an individual race that seems to matter.

[BillW]

Bill,

I'm not set up for filtering for starters odds, but I come up with some pretty sour nos. too. This meets all of your criteria except is for all odds ranges for the starters, which should increase the win pct. greatly, but drop the ROI. This data possibly does not have any significance relative to your angle, but I post it for your perusal.


Strt Wins Pct. IV ROI PIV
Post Position
Rail 390 32 8.21 0.98 -0.48 0.69


Bill

[Bill Cullen]

Originally posted by BillW
Bill,

I'm not set up for filtering for starters odds, but I come up with some pretty sour nos. too. This meets all of your criteria except is for all odds ranges for the starters, which should increase the win pct. greatly, but drop the ROI. This data possibly does not have any significance relative to your angle, but I post it for your perusal.


Strt Wins Pct. IV ROI PIV
Post Position
Rail 390 32 8.21 0.98 -0.48 0.69


Bill

Yes, I agree. It looks like more of a confirmation of what Buckeye and PMD found than what I would expect on extraploating from Nunamaker's route samples done back in the early 1990's.

Many thanks to you Bill, Buckeye, and PMD.

Bill Cullen

[hurrikane]

I believe you will find when tested that many of Nunamakers stats do not hold up....at least that has been my finding.

In fact I have found none that made money.

[pmd62ndst]

I >DID< reverse the odds!! Those numbers I posted were for the Final Odds <= 6. Thanks so much for double-checking my work. How embarassing...

Here are the numbers for the Final Odds to be >= 6 and they are far worse for the rail:

Post Position 1:
$ -453.20 (-46.9%) 20/483 (4.1%) $512.80/20 (25.64)

Other Post Positions:
$ -2517.30 (-23.4%) 260/5371 (4.8%) $8224.70/260 (31.63/0.00)


PMD

[InsideThePylons-MW]

Originally posted by hurrikane@HTR
I believe you will find when tested that many of Nunamakers stats do not hold up....at least that has been my finding.

In fact I have found none that made money.

Seems like this one fits the bill too! :rolleyes:


I have always preached that the #1 horse is usually anti-value. I don't have stats or computer queries to back this up, just my logical theories.

[MichaelNunamaker]

Hi hurrikane@HTR.

You wrote "I believe you will find when tested that many of Nunamakers stats do not hold up"

Neither have Fred Davis's<G>.

Seriously, it depends on what you are looking at. Large sample size winning percentages and IVs have held up extremely well. NETs in many cases have changed quite dramatically. The book really should be updated, but it is very much not worth the grief to me.

There are so many people with large databases out there, and all of them could write an update to MIV that would be quite useful to many people in racing. I hope someone does because it really helps a good number of people.

Mike Nunamaker

[hurrikane]

Mike

I know you post here and hope you know I meant you no disrespect whatsoever...or Fred. :D

all of us with dbs find things, they work well for a while..then maybe work to the mean...or everyone up here gets the word. :D

Want to say I have the utmost repect for your work and what you did early on.....

h