Calculating exacta fair values

[fmolf]

Originally Posted by duckhunter3

If you are using your own betting line, that line should add up to 100% and represents what you think is fair pay AFTER the track takeout. That is the formula we have been discussing, with the same result reached with several different mathematical approaches.

BUT, if you are using the tote board odds to represent your opinion of each horse's chances, those odds total far more than 100% to take into account track takeout. And the formula takes into account the effect of track takeout by becoming this: amount bet x odds on Horse A x odds +1 on Horse B DIVIDED BY the track PAYOUT squared after takeout. So, if a track pays out 76% on exacta and takes out 24%, then the track payout squared would be .76 x .76 or .5776 being the divisor.

That formula, witho that divisor accounts for the track takeout. You now have what is fair pay for the quoted TOTE BOARD odds on this exacta combination, and are free to add any premium or cushion you are comfortable with to make it an overlay. If no overlay, no bet.

duck

my odds line adds up to 80% as i assign 20% to the field....then i use my odds for the horses i like and look at the chart and then the track payout matrix....some people like robert are very good and quick with their math and can handle doing it that way..i cannot

[duckhunter3]

Originally Posted by fmolf

my odds line adds up to 80% as i assign 20% to the field....then i use my odds for the horses i like and look at the chart and then the track payout matrix....some people like robert are very good and quick with their math and can handle doing it that way..i cannot

Well your odds line is really 100% as it should be, 80% to the contenders and 20% to the rest of the field. Nothing wrong with that.

I usually take a look at the exacta combinations I am interested in well in advance of the race and quickly calculate fair pay + 50% (very easy with the formula), then prepare a matrix with horses on top across the top of the table and horses second on the left side of the table. I can easily pencil in the required payoff for a play and compare to the tote board payoffs about 5 minutes before post.
duck

[dav4463]

Originally Posted by Robert Fischer

the main thing is that the formulas you mentioned suck at determining exacta fair value. They are just fun shortcuts. Field size and pool distribution run amok with general shortcuts.

If you want an accurate formula, instead of being lazy and using the win odds of the two horses, you have to estimate what hit percentage that exacta has of occurring.

Once you have done the work, there are two things you can do.
You can divide 1 by the estimated hit percentage to find a fair payout for $1 exacta. EXAMPLE = 1/ 3%estimated hit percentage = $33.34

The other thing you can do is multiply the estimated hit percentage by a low estimate of the $1 exacta probable. This will tell you how much of an overlay you have. Example = 3%estimated hit percentage x $32 = 0.96 or a 4% underlay.

So is this right?

I have an exacta combination that hits at a 3% clip. If I divide 2/3%, I get $66.67. So I should add $33.33 for a 50% overlay and play this combination that hits 3% when it pays $100 or more.

If I have an 11% combo. I divide 2/11%, and get $18.18.....with a 50% overlay I should play this combo when it pays $27 or more?

Is this correct?

[CBedo]

Originally Posted by dav4463

So is this right?

I have an exacta combination that hits at a 3% clip. If I divide 2/3%, I get $66.67. So I should add $33.33 for a 50% overlay and play this combination that hits 3% when it pays $100 or more.

If I have an 11% combo. I divide 2/11%, and get $18.18.....with a 50% overlay I should play this combo when it pays $27 or more?

Is this correct?

Your math is right; just don't forget to adjust your return calculations if you are playing more than one combination.

[dav4463]

Originally Posted by CBedo

Your math is right; just don't forget to adjust your return calculations if you are playing more than one combination.

What do you do? Add $2 for each extra combination?

[InFront]

The bottomline line does playing exacta combos at overlay prices show longterm profit?. Regardless what formula you use if there is a combo of two horses and based on their win odds should pay a $40 exacta but the board shows $60 should you play it?

As we learned on win bets using Odds or Probability Lines just cause something has higher odds above the norm doesn't mean it works. As with any probability formula as the odds go up that formulated probability goes out the window.

In an average 8 horse race there are 56 exacta combos. If you quickly used the win odds of every horse and the calculated the fair value of all exacta payoffs and only played those at a 50%+ payoff what would happen in the real world if you did this based on thousands of races. Can anyone state for a fact would you profit or lose your shirt?

[fmolf]

Originally Posted by InFront

The bottomline line does playing exacta combos at overlay prices show longterm profit?. Regardless what formula you use if there is a combo of two horses and based on their win odds should pay a $40 exacta but the board shows $60 should you play it?

As we learned on win bets using Odds or Probability Lines just cause something has higher odds above the norm doesn't mean it works. As with any probability formula as the odds go up that formulated probability goes out the window.

In an average 8 horse race there are 56 exacta combos. If you quickly used the win odds of every horse and the calculated the fair value of all exacta payoffs and only played those at a 50%+ payoff what would happen in the real world if you did this based on thousands of races. Can anyone state for a fact would you profit or lose your shirt?

first i select my win horse/horses then i select my second place horses, then i check the board to see if i have overlaid fair value combos...some races i only bet one or two combos because the last one is underlaid...i would say yes the answer is yes bet only the overlaid combos that you like....for long term profits...not just random exacta combos that are overlaid.

[InFront]

Originally Posted by fmolf

first i select my win horse/horses then i select my second place horses, then i check the board to see if i have overlaid fair value combos...some races i only bet one or two combos because the last one is underlaid...i would say yes the answer is yes bet only the overlaid combos that you like....for long term profits...not just random exacta combos that are overlaid.

What you are saying is you are FIRST handicapping every horse to find who you think will win and who will place and then ONLY use the win odds and exacta payoffs to see if you have value or not. That would make sense.

But what I was asking is if you want to call it "random" that playing ALL exacta combos based on their win odds and exacta payoffs can possibly show long term profits? For example those that are at a 20% overlay, 30%, 50%, etc. I don't think it would but was curious has anyone done much study on this?

[CBedo]

Originally Posted by InFront

What you are saying is you are FIRST handicapping every horse to find who you think will win and who will place and then ONLY use the win odds and exacta payoffs to see if you have value or not. That would make sense.

But what I was asking is if you want to call it "random" that playing ALL exacta combos based on their win odds and exacta payoffs can possibly show long term profits? For example those that are at a 20% overlay, 30%, 50%, etc. I don't think it would but was curious has anyone done much study on this?

I did a little with something similar to this a while ago (really small sample--maybe a week at 5 or 6 tracks), and what I discovered quickly was that the late odds changes in both the win and the exacta pools made it tough to implement effectively. What looked like value would disappear, and what looked like value after the fact, you might not have seen before hand. People arbitrage pools in some crude way most of the time. If we had real time data of what was going on in the pools and didn't have to wait until after the fact to get the odds, then I think there would be even more "equality" across pools (where there is enough size to guarantee liquidity).

[rrbauer]

Originally Posted by CBedo

I did a little with something similar to this a while ago (really small sample--maybe a week at 5 or 6 tracks), and what I discovered quickly was that the late odds changes in both the win and the exacta pools made it tough to implement effectively. What looked like value would disappear, and what looked like value after the fact, you might not have seen before hand. People arbitrage pools in some crude way most of the time. If we had real time data of what was going on in the pools and didn't have to wait until after the fact to get the odds, then I think there would be even more "equality" across pools (where there is enough size to guarantee liquidity).

Very good! Unless you are dealing with huge pools, such as what we see on the TC cards or BC cards, then your best guess is simply that: a guess. And, for that matter relating to the thread, "fair value" relative to what?

[InFront]

This is why I never liked methods that relied on Post Time odds or payoffs. While using even Win Odds as a rule within querying a database may show this or that to use it in real life day to day betting is tough. Yes much easier these days with Conditional Wagering interfaces but still tough.

The only way post time odds may work if all tracks first closed down all betting at assigned post time, all pools are then recalculated and locked in and then they waited about 2 minutes more to start loading the horses. But of course they would never do that, they want betting to continue right until the gate pops open to make even more money which I can't blame them.

[rrbauer]

They can load and lock ten minutes before the race and it won't make any difference from the perspective of being able to make betting decisions based on what's being displayed on the board.

[formula_2002]

I'll add my CAA


CALCULATING A FAIR EXACTA PAYOUT

THE PROBABILITY OF A HORSE WINNING IS (1/(ODDS+1))/ THE RACES TOTAL BOOK PERCENTAGE.
THE TOTAL BOOK PERCENTAGE FOR A TRACK WITH A 15% TAKE OUT SUMS TO ABOUT 1.18
THE PROBABILITY OF ANY HORSE WINNING IS (1/(ODDS+1))/1.18
TAG THE WIN HORSES WIN PROBABILITY WITH "A"
TAG THE PLACE HORSE WIN PROBABILITY WITH "B"
EXACTA PROBABILITY PROBABILITY FOR A/B IS:
A X (B/(1-A))
EXAMPLE: A HORSE ODDS =2-1, B HORSE ODDS = 3-1, TRACK TAKE OUT IS 15%
A HORSE WIN PROBABILITY IS (1/(2+1))/1.18 = .282
B HORSE WIN PROBABILITY IS (1/(3+1)/1.18 = .212

PROBABILITY FOR THE A/B COMBINATION IS,
.282 x(.212/(1-.282)) = .083
THAT EQUATES TO A $1 RETURN OF (1/.083)= 12.04
THE FAIR $1 PAYOUT IS $12.04

[elhelmete]

Topping a zombie thread from 13 years ago...a new narcissistic low.

[formula_2002]

Originally Posted by formula_2002

I'll add my CAA


CALCULATING A FAIR EXACTA PAYOUT

Sorry, there is an error in the excel charts. I will fix