Comments on: Fantasy Basketball Draft Tip #1: Percentages are not Created Equal.

By: Nels

Nels — Sat, 01 Nov 2008 00:47:05 +0000

Since Patrick is the one who could answer your statistical query, I will simply direct you to Basketball Monster.

By: Zach

Zach — Fri, 31 Oct 2008 21:21:24 +0000

Here’s what I’m trying to do and have run into a slight problem. I’m trying to create a spreadsheet based on the 11 categories our fantasy basketball league uses to come up with a composite score for each player in each category that I can total at the end to give me a list of the best players given the scoring of our league. The categories we use are PTs, FG%, FT%, 3ptMade, 3PT%, OREB, DREB, Blocks, Steals, Assists, TO’s. Basically what I have started to do is assign a value between 0-1 for each player for each category based on last years stats. For points it was easy. I used Lebron’s 30ppg as the high water mark and took the remaining top 200 fantasy players in the league and divided their ppg by 100, Lebron being the only one that earned 1 full point. FG% is a little trickier, but I basically followed what you posted above and made minor tweaks to come up with a fairly normal distribution of scores between 0 and 1 for this category. When I come to 3pt% though, I run into a small problem…big men don’t shoot threes, so they do not penalize you for shooting a bad percentage necessarily, but they do not help you either. In the FG% category, I basically assigned the avg. to be worth .5 points, all the way to Stephen Jackson or maybe someone else gaining only .1 points out of a possible 1. The problem is, a guy like Stephen Jackson actually hurts your FG%, whereas Dwight Howard wouldn’t hurt your 3pt%, but the question is how should you rate players that don’t shoot 3′s in regards to 3pt%? Would it be alright to arbitrarily assign them the half a point since, sort of similar to FG%, the players I assigned .5 point to (the players whose FG% mirrored the average) don’t necessarily hurt or help your %’s, or would this throw off the overall composite scores. Just wondering how somebody else may have attacked this similar problem.

By: M-Dawg

M-Dawg — Sun, 26 Oct 2008 01:38:26 +0000

Ive wondered if using the median instead of the average would be better when tackling fantasy hoops. It seems in many categories there are a handful of guys that perform substantially higher than the crowd, and this may skewer the average. Ive checked and there is a significant difference between the median and average in most categories.

Id appreciate any comments

By: T-Time

T-Time — Wed, 06 Aug 2008 21:46:17 +0000

Ahhhhhhhhhhhhhhh

“I only use the top 100 to get rid of the noise from part time and other players who don’t shoot threes.”

That’s the key right there I think. Because I had been using everybody the league average figure was wayyyyy too low, distorting the results a bit. For the exact reasons you mentioned about players who don’t shoot 3′s.

Thanks for the tip Patrick, I’m off to hit up the spreadsheet again.

By: Patrick

Patrick — Wed, 06 Aug 2008 14:52:10 +0000

Hey T, thanks for the comment and complement.

So, if I understand you, you play in a league that uses three point percentage as a category? If so, then yes, you should be able to calculate an effective three point percentage for every player. And I think you’re right saying that it should vary even more than free throw percentage, because there are some players who take a lot of threes and some who take none.

To calculate it, first calculate the league average number of threes and league average 3pt percentage for the top 100 three point shooters in the league. I only use the top 100 to get rid of the noise from part time and other players who don’t shoot threes.

Then, take a player’s 3pt percentage and subtract the league average percentage. Multiply that value by the number of threes a player takes per game divided by the league average numbers of three taken a game. Add that number to the league average 3pt percentage to get a player’s E3pt%.

Say a guy shoots 3 threes a game at 40% and the league average is 1.5 threes a game at 35%. His E3pt% would be:

((40% – 35%) * (3 / 1.5)) + 35% = 45%

I hope that’s understandable. Good luck!

By: T-Time

T-Time — Wed, 06 Aug 2008 13:46:11 +0000

Hey guys,

I really love your Effective Field Goal and Free Throw Percentages, I cant wait to use them to get a leg up in my draft (in particular in order to overcome Lebron’s gross FT%)

Have you tried to develop a 3PT% version? I’ve made a few attempts using info and equations from some of the sites you’ve linked on the topic but I’m not yet satisfied as my results seem a little screwy.

In an old article you guys mentioned:

“Everything written about EFT% also applies to EFG%. However, the range of field goal percentages and the number shots attempted from the field (average of 11.5, std dev of 4.2) varies less than the number of free throws attempted (average of 4.2, std dev of 2.3), so EFG% stays fairly close to the standard FG%”

Could it be possible that the reverse is true of E3PT%? That the range of 3 point %’s and the number of 3′s attempted varies a great deal and so therefore E3PT% is wildly different from standard FG%??

Any advice would be great.

Thanks.