Converting win percentage to a point spread, etc.?

[dirk93]

I have been using the pythagorean/log5 formulas for capping the NBA and I get a win percentage for each team, which I can convert to a moneyline. My main problem is that I am trying to incorporate injuries and rest into the formulas (which are used most easily through a plus/minus rating).

Does anyone know a way to either convert my win percentages into a spread or convert my win percentages into a predicted score so I can accurately incorporate injuries and rest? If I could convert it to a spread, then I would be able to make my formulas more accurate including injuries and rest.

I know there is the calculator on SBR, but I would like a formula so I can just enter it into excel instead of going through the website everytime.

Thanks

[splash]

I doubt there is a formula. The SBR tool uses the actual results of years of data. I suppose if you had that data, you could plot it, and extract an equation from the curve.

[Justin7]

You need to make a ML-Spread conversion chart. It will be more accurate if you use totals also.

[dirk93]

Thanks for the help. As an aside, has anyone experimented with incorporating adjusted plus/minus ratings into their formulas when trying to account for injuries? I initially thought that may be beneficial but now it's looking like there is too much variance for it to be significant. Any ideas?

[dirk93]

Sorry to keep adding on the questions but this applies to my model. Does anyone know where to find an NBA schedule by date that has the amount of rest for each team on each day such as B2B, 4 in 5, 1 day rest, 2 days rest, 3 in 4, etc.

[Dimwhit]

NBAstuffer has the schedule with days rest in an excel spreadsheet. It costs $5, though. Might be a free one out there somewhere.

[StraitShooter]

NBAstuffer has the schedule with days rest in an excel spreadsheet. It costs $5, though. Might be a free one out there somewhere.

thks for the heads up. Been looking for this info past couple days

worth the 5 bucks but will post if I find a free one

[ManBearPig]

I"m not sure if this is what you're looking for but I used this one this season and found it quite useful. HERE

The NBA schedule link is on the left side.

[uva3021]

http://sportsobjective.com/wordpress/?p=3554

The pythagorean formula - (RS^exp / (RS+RA)^exp) - can also be a calculation using the ratio of runs scored to runs allowed - (RS/RA)^exp / ((RS/RA)^exp +1)

Switch your variables, making the win probability a known variable, and the runs scored - runs allowed differential unknown, with the assumption that there is an expected total

using pencil and paper you can solve for x, this is the result

P = win probability
n = exponent