Successfully throw out the favorite and what do you have;

[formula_2002]

Losing favorites.
Successfully throw out the favorite and what do you have;

Based on a dutch play;

1751 actual winners compared to 1464 expected winners, yielding a 17% dollar profit.

That’s sounds fair since a dutch play on all favorites would yield the following;
920 winners compared to 1104 expected winners, yielding a 17% dollar loss

Cj’s top rank figure would return 264 act wins vs. 216 exp wins for 18% profit
2nd ranked fig would return 328 act wins vs. 236 exp wins for 28% profit
3rd ranked fig would return 281 act wins vs. 232 ex wins for 18% profit
4th rank would return 257 act wins vs. 205 ex wins for 20% profit.
5th ranked and higher would return 621 act wins vs. 574 ex wins for an 8% profit.

Expected wins = 1/(odds+1)

[RXB]

But that would be based on the favourite having zero chance of winning the race-- unfortunately, those lousy sob's running the parimutuels make me wager before the race.

Realistically, I'm happy if I can find a typical 9/5 favourite that I feel only has about a 10% chance of winning.

[kenwoodallpromos]

I would rather successfully throw out all the losers every time!

[formula_2002]

excellent point..

[kenwoodallpromos]

I took your advice and thnrew out Rockport Harbor and bet $3 on each of the others! You did good! Thanks!
Now I will have to go back to flatstats.com and study how the English "lay" horses!
Off topic- do you know where I can get trainer stats by track condition?

[mholbert]

i had to register just to respond to this. you make the same error that dick mitchell makes in calculating your expected win percent. let's say you only have two horses in the race and they both have the exact same money. both have 50% of the pool.

what are their posted odds? they are both 3/5 with a 17% takeout and breakage.

if i use your formula for expected win percent [Expected wins = 1/(odds+1)], i end up with an expected win percent of 62.5% instead of 50%.

finally, you may want to do a little reading on curve-fitting. you'll then find out why system 1,2,3,4,5, 6, 7, 8, 9 and 10 don't seem to work on a going forward basis.

i expect your usual response that i've seen when someone questions you, so i doubt i come back to read this thread.

[formula_2002]

Mholbert
You hurt my soul…
First you compared me to Dich Mitchell..I’m incomparable.

Second, it’s ok to calculate expected return the way I mentioned as long as you understand what is being done.

As I’m sure you know, summing 1/(odds+1) for all horses in a race allows the determination of the track take in that singular race.

But I would appreciate any advice you have to curve fitting.


Yes systems 1 to 10 did not hold up.. Neither did the 1000’s before. :bang:


ps
PA, I must get credit for his registration

[formula_2002]

Lets go on a bit with this.
Suppose that you have handicapped a 9/5 horse to have but a 10% chance to win a race.
9/5 is 35.7 % of an un-normalized pool. Say the track take is 17%.
So the 9/5 gets normalized to 29.6%.
If it has but a 10% chance to win then 19.6% in thrown back into the pool.
Say the track take it 15%. Then you could dutch all the other horses and make a 4.67% profit.

What an easy game :rolleyes:

[formula_2002]

Lets go on a bit with this.
Suppose that you have handicapped a 9/5 horse to have but a 10% chance to win a race.
9/5 is 35.7 % of an un-normalized pool. Say the track take is 17%.
So the 9/5 gets normalized to 29.6%.
If it has but a 10% chance to win then 19.6% in thrown back into the pool.
Say the track take it 15%. Then you could dutch all the other horses and make a 4.67% profit.

What an easy game :rolleyes:


The above is not correct. I'll post a "more accurate" post asap.
the short version is this

if the track take is 19%

and under certain conditions the favorites lost 28%

a dutch on all the others horses produced a 15% loss.

Thats what I get in the real world!

[formula_2002]

Lets go on a bit with this.
Suppose that you have handicapped a 9/5 horse to have but a 10% chance to win a race.
9/5 is 35.7 % of an un-normalized pool. Say the track take is 17%.
So the 9/5 gets normalized to 29.6%.
If it has but a 10% chance to win then 19.6% in thrown back into the pool.
Say the track take it 15%. Then you could dutch all the other horses and make a 4.67% profit.

What an easy game :rolleyes:


It looks like my previous post was correct. I'll explains it a bit differently.

At a 17% take out track, for each race,the sum 1/(odds+1)= about 1.21.
A 9/5 short represents 35% of the 121%.
If you threw out the favorite and dutched the rest of the field, it would cost 121units -35units.= 86.
If one of the horses you bet on, won ,you would get 100 units back for your 86 unit bet.
A net profit of 14 units.


In a series of 100 races;
If the 9/5 shot won 10% of the time, you would lose 100X86x10 units= 860 units
You would net 14 units 90 times, which = 1260 units
The net profit = 1260-860= 400 units

Total bet =100x86 = 8600
% profit= 400/8600 = 4.67%%