What are some good formulas to use?

OE = Pts/Possessions

where # of possessions can be easily and pretty accurately determined from the stats in the boxscore.

It's: 0.475* FTA + FGA - OReb + TO's

You can take the average of the number you get for each team or just use a single calculation from either team to save time (it's pretty close usually anyhow).

Top 20 teams usually have an Offensive Efficiency (Pts Scored/Poss.) above 1.05-1.10, and a good Defensive Efficiency (Pts Allowed/Poss.) below 0.95

For the college game, between missed 1-1 front ends and fouls on 3s, does that erase the "and-one"s and make the ratio more like .5*FTA? .475 seems reasonable. I think that they use .44 in the pros. Did you get that number from somewhere, or does it just seem like a good estimate?

Let me rephrase... what are some good formulas to use for individual players?

I'm no math whiz but I would think for individual players you can just adjust for the position. I'd guess OReb would be good for bigs, while you'd substitute assists for guards. Or you could put orebs and assists for both, and just get higher numbers, and adjust according for what's good and bad.

After putting a little more thought into it (as soon as I hit post reply) using a formula that consists of points per shot, offensive rebounds per game and assist:to ratio would be the best way to determine a players efficiency. Maybe somebody better with these type of forumlas than me can put someting together.

Quote: Originally posted by cornfused on 3/14/2009Let me rephrase... what are some good formulas to use for individual players?

I have a spreadsheet for use with individual players, offensive efficiency is one of the stats generated. It's complicated to generate the stat because you have to use estimated number of possessions for the individual player and points produced (which includes credit for assists, which is complicated to determine how much to count an assist for, again requires another equation).

Read Dean Oliver's Basketball on Paper if you want to see formulas, you can find most of it at Google Books.

And I found the 4.75 * FTA not to be the most accurate for HD. .5 comes out more accurate.

Has anyone used Win Score to determine efficiency?

Quote: Originally Posted By kmasonbx on 3/14/2009

After putting a little more thought into it (as soon as I hit post reply) using a formula that consists of points per shot, offensive rebounds per game and assist:to ratio would be the best way to determine a players efficiency. Maybe somebody better with these type of forumlas than me can put someting together.

kmason, asst:to ratio is totally screwy in HD, I would not use it as part of any formula. And the assist part (as jdtapp pointed out) is very difficult to quantify, I don't even include it when I'm looking at my players' offensive efficiency. Pretty basic efficiency models work well for HD, you don't have to delve too deep.

I use a fairly simple formula that adds points per minute, rebs per min, assists/min, steals/min and 0.5(blocks/min), then subtracts TOs/min and 0.75(fouls/min).

I understand what you're trying to do with the per/min stuff, but I'm not sure I like it ... points per minute is so hugely determined by distro. So the same exact two players, one with 15 distro and one with 5, would score wildly differently on your scale.

(Presumably the one with 15 distro would turn the ball over a little more, which would even it out just a bit, but not enough to offset 3x the distro).

Also, offensive efficiency is really how efficient a scorer you are ... yours would be interesting from an "overall" perspective, but I don't think it's a terribly good measure of pure offensive efficiency. (Not to mention that the values you assign are very arbitrary and we don't even know if they're accurate/representative in what to look for in determining a player's performance.)

Points per shot is probably a better way to determine efficiency. That way your grading every player essentially on the same scale regardless of distro.

Well, I did qualify it by saying it was "fairly simple." :)

And what I've found is that the guy with 15 distro, almost without regard to ratings, turns it over a LOT more than the guy with 5 distro. But that's a subject for a different thread.

In any case, you're correct--what I'm measuring is not just offensive efficiency, but what I call "contribution" to the team's entire effort. Probably not as applicable here.

On another note, why don't you take a short break from posting and accept my challenge in Allen? :)