Quote:
Originally Posted by mountainman
DD and exacta pools aren't always predictive of final win-odds at Mnr. But, then again, my way of consulting them isn't very sophisticated.
I wish a Dave S. or a Jeff P. would post the quickest fairly accurate way to guesstimate final win odds based on exacta or DD pools.
|
Keep in mind that the math shown below is rough... It's something I do in my head (and every once in a while on a scrap of paper) when trying to predict where late money is likely to land.
If you really want it done quickly and accurately - I'm guessing it wouldn't be that hard for someone to code out as a VBA macro in Excel. (Imo, if you're going to do that any serious effort at calibration would probably get you something useful.)
And if anyone else has a more accurate method, feel free to jump in. (I'm not against learning new stuff.)
Ok. Let's jump in and do an example race --
These are the $2.00 will pays for the double spanning races 1 and 2 earlier today (Monday 08-03-2020) at Thistledown:
#1 $86.20
#2 $113.80
#3 $18.80
#4 $15.80
#5 $165.20
#6 $36.00
#7 $22.00
First, a description of the basic steps:
A. The first step involves calculating the cost to have singled the winner in (each of) the previous leg(s) to all of the live will pay runners in the final leg.
Let c be equal to the amount it costs to do that.
c = (7 runners) x (2.00 base wager amt)
or
c = 14
B. Next, calculate the adjusted will pay for each runner using the the following formula:
adjWillPay = (actual willpay) x (1/(1-takeout))
C. Finally, calculate estimated odds using the following formula:
est_odds = (adjWillPay/cost)
Next, let's perform each of the steps:
Step #1. cost = (7 runners in final leg) x ($2.00 base wager amt)
or cost = 14
Step #2. Calculate the adjusted will pays for each of the live will pay runners using 0.225 as the takeout
Let's do the right hand side of the adjWillPay formula first because it's the same for each horse.
Let r=(1/(1-takeout))
Substituting the actual takeout rate of 0.225 I get r = (1/0.775) calculated as follows:
r = (1/(1-0.225))
or r = (1/0.775)
Now that we have r, we can calculate the estimated will pay for each runner using the following formula:
adjWillPay = (actual willpay) x r
#1. adjWillPay = 86.20 x (1/0.775) or adjWillPay = 111.23
#2. adjWillPay = 113.80 x (1/0.775) or adjWillPay = 146.84
#3. adjWillPay = 18.80 x (1/0.775) or adjWillPay = 24.26
#4. adjWillPay = 15.80 x (1/0.775) or adjWillPay = 20.39
#5. adjWillPay = 165.20 x (1/0.775) or adjWillPay = 213.16
#6. adjWillPay = 36.00 x (1/0.775) or adjWillPay = 46.45
#7. adjWillPay = 22.00 x (1/0.775) or adjWillPay = 28.39
Step #3. Now that we have adjusted will pays, we can calculate estimated odds for each runner using the following formula:
estOdds = (adjWillPay/c)
#1. estOdds = (111.23/14) or estOdds = 7.95
#2. eestOdds = (146.84/14) or estOdds = 10.49
#3. estOdds = (24.26/14) or estOdds = 1.73
#4. estOdds = (20.38/14) or estOdds = 1.46
#5. estOdds = (213.16/14) or estOdds = 15.23
#6. estOdds = (46.45/14) or estOdds = 3.32
#7. estOdds = (28.39/14) or estOdds = 2.03
Finally, the following table shows the estimated odds using this method vs. the final odds for each runner in the second half of the double spanning races 1 and 2 for Mon 08-03-2020:
Code:
Est Final
# Odds Odds
-- ------ ------
#1 7.95 11-1
#2 10.49 23-1
#3 1.73 7-2
#4 1.46 1-1
#5 15.23 25-1
#6 3.32 4-1
#7 1.57 7-2
Fyi, the above method makes no attempt to account for the odds of the winner in previous legs which can and does impact the will pays.
When a short priced favorite wins the first leg of a double the will pays are often lower than normal.
Conversely, when a longshot wins the first leg of a double the will pays are often quite a bit higher than normal.
Both scenarios can cause the estimated odds using this method to vary from final odds.
The first leg of the above example race was won by a favorite that paid $4.40 to win. (Which might skew things a bit.)
However, if you look at odds rank, the rough math shown above came really close to nailing every horse in perfect order.
It missed when it predicted the #7 horse to have lower odds than the #3.
Other than that, it got the odds rank of the horses in perfect order.
-jp
.