quantum beats the favorite(?)

[formula_2002]

If testing means ANYTHING, I have got to believe, that Quantum can pick those races where the short price horse (less than 2-1) wil return an ROI of less than .75.

I tested 385 such plays ,and the roi was less than .75.

Then I tested it against over 700 out of sample races.
The ROI was even less, about .70.

Average odds are about 1.2 -1 or 46 % of the win pool.
The book in a 15% take-out track is about 1.18
That means if I dutch all the remaining horses, the total bet would be (1.18-.46) =.72 and I would get back 1.0 for a profit of .28.

If you bet into a 18% take track the profit goes down to .24 (too bad right ;) ).

There is a bit of reality to overcome..and that is of course late changing odds, but thats not my problem (unless I bet )
I'll post the plays in the "selection" forum.

[chickenhead]

Average odds are about 1.2 -1 or 46 % of the win pool.
The book in a 15% take-out track is about 1.18
That means if I dutch all the remaining horses, the total bet would be (1.18-.46) =.72 and I would get back 1.0 for a profit of .28.
.

Something is out of whack with those numbers. You're getting back 1.0 only when you win, which is something like 69% of the time, so your profit should be -0.03 over all of the plays?

What if you excluded any horses over 10-1, (or 12 or 15-1), in your dutch.

[ryesteve]

Average odds are about 1.2 -1 or 46 % of the win pool.
6/5 shots don't have 46% of the win pool.

[formula_2002]

6/5 shots don't have 46% of the win pool.
It's their book % 1/(odds+1)

% of pool in a 15% take would be .46/1.18 = .38 % pool..
I do get sloppy..

[formula_2002]

Something is out of whack with those numbers. You're getting back 1.0 only when you win, which is something like 69% of the time, so your profit should be -0.03 over all of the plays?

What if you excluded any horses over 10-1, (or 12 or 15-1), in your dutch.

I think you are closer the truth than I.

in 378 out of sample plays
ave odds were 1.25-1
the favorites lost 73 times in 100 plays

I would have played (1.18- 1/(2.25)) =74 units

I win 73 times x (100-74)= 1898

I lose 27 times x 74 = 1998

I lose 100 units having bet (378*74)= 27972 units
or less than 1%
well that sucks

In that sample the favorites had an roi of but .62.

[chickenhead]

I don't think that's bad....just not enough to dutch everyone. There is no reason to assume that the overlay is spread evenly amongst all the other horses (on avg), I would bet the majority of it is concentrated on the next few lowest odds horses, with the long shots still losing close to the track take.

[formula_2002]

I think you are closer the truth than I.

in 378 out of sample plays
ave odds were 1.25-1
the favorites lost 73 times in 100 plays

I would have played (1.18- 1/(2.25)) =74 units

I win 73 times x (100-74)= 1898

I lose 27 times x 74 = 1998

I lose 100 units having bet (378*74)= 27972 units
or less than 1%
well that sucks

In that sample the favorites had an roi of but .62.

I think I need a rest.

Thats not right..
My total play would have been 100*74=7400
100/7400= 1.35% loss

[formula_2002]

I don't think that's bad....just not enough to dutch everyone. There is no reason to assume that the overlay is spread evenly amongst all the other horses (on avg), I would bet the majority of it is concentrated on the next few lowest odds horses, with the long shots still losing close to the track take.

I'll have to work on it some more..
I'll have to see who's winning..

[formula_2002]

The proof now in the live selections..
It looks like the magic number is >=4-1.
Dutch every thing >=4-1 where the posted horse is <2-1. Ofcourse if you have handicapped a horse in this and its not the posted horse..jump on it.

It will be like playing at a track with no take, and hopefully we will do the taking!!

I even salted my one line acount with a few bucks so I can play along..

Good luck to us.

Joe M