Nascar Matchup Math Question for Justin7, Fishhead, Sharpcat or any other sharp.

[Geor99]

Does anyone have any information or a link that describes the process that is used to come-up with Nascar Head-to-Head matchup Numbers? Is there a way to convert "Odds to Win" to matchup numbers?

For Example if Driver A is +1200, B is +400, C is +350, etc to win the race; can these be used to get the numbers on matchups like Driver A versus Driver B or Driver B versus Driver C, etc?

I would imagine that this would be very similiar to converting odds to win to matchups in Horseracing.


I have a local who picks his own head-to-head numbers before the real Books release their numbers. He moves them to the market prices as soon as the real #'s come out. He is a Nascar fanatic and seems to enjoy releasing his own numbers. I am curious if I can use Odds to Win from larger books to see if there is any value in these early matchup numbers.


Any help in this arena would be greatly appreciated?

[TomG]

Is the probability of each individual driver winning basically independent? Or do some drivers team up or behave significantly differently depending upon who is in the lead?

I know nothing about NASCAR.

[Flight]

Start logging race odds from an existing sportsbook. Historical data (stats and market odds) is valuable.

[durito]

Start by amending your sharp list.

[sharpcat]

-convert all drivers odds to push %.
-get sum of all drivers push %'s.
-divide drivers push % by sum of all drivers to eliminate the vig.
-take the sum of the 2 drivers in your matchup and divide each driver by the sum of both drivers.

Books usually have a 30% over round in these markets and I would not consider "to win" the most efficient market but this should get you close enough to crush a local bookie who is putting out his own lines.

[brettd]

See whether this local is kind enough (and dumb enough) to give you historical data on the lines he's set for these match ups. Just say you 'love' NASCAR and want to see his historical lines out of 'curiosity'.

Then get the historical "odds to win" numbers.

What to do from there is then fairly easy.

[LT Profits]

-convert all drivers odds to push %.
-get sum of all drivers push %'s.
-divide drivers push % by sum of all drivers to eliminate the vig.
-take the sum of the 2 drivers in your matchup and divide each driver by the sum of both drivers.

Books usually have a 30% over round in these markets and I would not consider "to win" the most efficient market but this should get you close enough to crush a local bookie who is putting out his own lines.

"Field" complicates things though, but close enough for government work.

[Optional]

Does anyone have any information or a link that describes the process that is used to come-up with Nascar Head-to-Head matchup Numbers? Is there a way to convert "Odds to Win" to matchup numbers?

This thread might help you. Ganchrow gives me a bunch of formulas and scripts for my NASCAR capping calcs.

http://forum.sbrforum.com/handicapper-think-tank/526381-win-v-place-odds-value-math-question.html

It's a bit long, and doesn't start out asking about Head 2 Head matchups, but that is what I ended up making most use of the calcs for.

[Pancho sanza]

Fishface on the sharp list.

[RickySteve]

Start by amending your sharp list.

I lulz'd.

[RickySteve]

-convert all drivers odds to push %.
-get sum of all drivers push %'s.
-divide drivers push % by sum of all drivers to eliminate the vig.
-take the sum of the 2 drivers in your matchup and divide each driver by the sum of both drivers.

Books usually have a 30% over round in these markets and I would not consider "to win" the most efficient market but this should get you close enough to crush a local bookie who is putting out his own lines.

OMG, please do not do this if you don't hate money.

[RickySteve]

The Ganchrow method isn't going to work too well either. Clean Gaussian math is real fun to jerk off to but describes few things in the real world.

[LT Profits]

Clean Gaussian math is real fun to jerk off to

I prefer Katherine Heigl, but I digress

[RickySteve]

Do not want