Vi and the Fav

[hurrikane]

I've been toying around with using the Vi number and converting it directly to an odds number to give me an idea of where the favorite's odds should be. Over or under creates opportunity.
For instance,

a Vi of 40 would mean the fav should have odds of 3/2

a Vi of 20 would set the odds on fav at 4-1.

over or underlays create opportunity.

If you are using the 4Qs can automate the starting process of setting the odds line based on the Vi.

I haven't run a lot of test and it may actually be a factor of Vi and not a direct conversion.

Anyone else done any work with Vi or have any ideas or input.

Thanks!

[MtKen]

Kane:

I did a quik look at % of payoffs over $15:


nVi.....#races.....>=$15.....%
>=35.....844..........118......14%
30-34....703...........152......21.6%
25-29....653...........180......27.6%
20-24....342...........110......32.2%
<=19.....66..............23.......34.8%


Seems to work as advertised & I'm hoping it will save me a lot of wasted bets.

[fred4now]

HK, you're ahead of me on this. Been thinking the same thing since the seminar. I'm going to run about a years worth of data through the HX4 exporter and then I will do some tests.

[hurrikane]

That's great Rick. I'm upgrading my computer so I haven't had a chance to back run the HX4. Let me know what you find. Small samples have looked promising. I'm trying to figure and algorithm to set the second place horse in the odds line.

Any suggestions are welcome.

[George]

Kane...am probably off base here but thought would inject
my two cents on making an odds line. Following is a quick
approach I have used for years. Any number or factor can
be used for the basis as long as there is some seperation.

I use VEL fron either DOS or HTR2001. Adding all the VEL
numbers in a race usually sums to 101 or 102. To save a math
step assume they sum to 100.

Race 8 at SAR today horse #2 has VEL of 26. Divide 100 by 26.
Subtract 1 from result of 3.8 for fair odds of 2.8

result of above for all horses in race:
#2 vel 26 odds to $1 2.8 payoff 5.60
#3 vel 22 odds to $1 3.5 payoff 7.00
#5 vel 15 odds to $1 5.6 payoff 11.30
#6 vel 15 odds to $1 5.6 payoff 11.30
#7 vel 17 odds to $1 4.8 payoff 9.70
#8 vel 06 odds to $1 15.6 payoff 31.30

Interestingly, AT THE RACES listed these actual payoffs at
post time.
#2 projected 5.60 actual 5.60
#3 ......... 7.00 ...... 8.90
#5 .........11.30 ......19.90
#6 .........11.30 ...... 9.60
#7 ......... 9.70 ...... 9.30
#8 .........31.30 ......33.50

Most races follow the calculated line from VEL very closely.
I think that is a real testimony to the accuracy of KM numbers.
Not sure that can be adapted to 4Q templates since VEL usually
has top choice around 2-1. Have seen a few 1.5-1 but not many.

BTW, I use this in tote watching cheap tracks. When you see a
horse like #8 above with 15-1 projected and it gets beat down
to 4-1 it is time to take the rubber band off the bankroll.
Underlays are sometimes the best bet of the day:)

[fred4now]

Thanks George, the line I have been using over the past 1000 races based on nHTR hardly ever gives a fair odds lower than 2 also. Been profitable so far.

[hdcper]

George,

Believe I noted an error in your fair odds payoffs. The 2.8 to 1 should payoff at $7.60 and so on for the other fair odd payoffs or calculations(must add back in the $2 wagered).

Just to support this thought consider the win percentage of #2 to be 26%(26/100) or .26.

Thus, .26 times $7.60 = 1.976(approx breakeven based on ROI of 2.00)

Hope this makes sense,

Bill

[George]

Bill..........reason it didn't make sense is because of the subtraction of 1 from the result in the formula, 100/26=3.8 -1 = 2.8 odds to one dollar.

You are correct that my example did not add the $2(1*2) back in the payoff.
Sorry, I normally just use pure odds in totewatching and screwed up in translating it to compare to AT THE RACES calculated payoffs.

With the correct calculated payoffs the match of projected and actual is not as close but the relative relationship is the same.

I also failed to mention in the above thread what I guess thought was obvious, prior to calc of the odds line, any set of numbers should be summed and each components percent of total sum calculated. VEL is handy because the sum is always very near 100 and that step can be skipped.

[hurrikane]

thanks George. Appreciate the input. What you say about the underlay I agree with completely. that is part of the reason for looking at the Vi, to see if there is a point where you can automatically throw out the fav. As in the false K***