Friday, December 05, 2008

More Tempo-Free Stats

I like to throw these things in a post every once in a while. At first, it can have a sort of death-by-numbers effect, especially if this is your first taste.

Over at NDNation, poster Final_Flanner has an exceptional spreadsheet breaking down individual statistics from Notre Dame players under Mike Brey's tenure (post here).

For a more in-depth explanation than I can give you on these numbers, check here. I also highly recommend
Basketball on Paper by Dean Oliver, the best book available on the topic and one that I am currently rereading.

"What should I care?" You must be asking this right now. You know Chris Thomas was a ballhog who couldn't hit the side of the barn with his terrible shot selection. On the other hand, Kyle McAlarney and Luke Harangody are offensive stalwarts who can do no wrong. Fair enough, but an added knowledge about tempo-free stats and the reality they represent will give you an increased appreciation for the game. Bear with me,
I'll bring this home.

Four stats tell you everything about the game of basketball (taken directly from Big Ten Wonk):

1. Shooting. Effective FG pct. (eFG pct.) = (FGM + (0.5 x 3PM))/FGA. (Read more here.)
2. Offensive Rebounding. Oreb pct. = team orebs/(team orebs + opponent drebs). (Read more here.)
3. Turnovers. TO pct. = TOs/possessions.
4. Free Throws. Free throw proficiency = FTM/FGA.

Don't worry too much about "eFG." It's just common sense, weighting three point field goals a little bit more because of the added point value. Offensive rebounding is clear, a team with more chances per possession does better than a team with one shot per trip down the floor. Turnovers: you can't score if you can't shoot. And foul shots, free chances to add points. There, now you know how to break down a box score more effectively than 95% of analysts who get paid to do that stuff on televsion. If a team loses in all four categories (or even three most of the time), they lose the game.

Let's take a look at what Final_Flanner has to offer. You already saw eFG% so now you just need to know PPWS (Oliver has somewhat different calculations, but they can be a pain to figure out. Points Per Weighted Shot is an effective tool to see how efficiently a player scores per field goal and free throw attempt.

PPWS = PTS/(FGA + (0.475 x FTA))

There you go. Plug the formula into a grpahing calculator if you'd like to figure the numbers out yourself (the only good use for those things aside from playing Super Mario knock-offs) or just let others do the dirty work for you. Simply put, a player with a high PPWS is more effective on offense and deserves to take more shots than a player with a low PPWS. Now take another look at the spreadsheet of last season.

Instantly, you will have a couple questions:

Wait, you're saying Jonathan Peoples and Tim Abromaitis are better players than Big East Player of the Year Luke Harangody?

Certainly not. While these stats are a fantastic way to compare players with similar abilities and playing time, they sometimes require a reality check. Peoples (and to a greater extent, Abro) received fewer minutes than Harangody and, as a result, fewer shots. When he did have his chances, he converted very efficiently, making him an excellent player off the bench. With more playing time, his PPWS would undoubtedly go down. Some story with Ryan Ayers. Though Ayers played a lot, he took very few shots (and made a big percentage due to his conservative shot selection).

Is eFG% or PPWS better at describing player performance?

Another good question. You're an excellent pupil. eFG% tends to favor three points shooters (as evidenced by Mac, Ayers, and Peoples leading by a wide margin). PPWS is better at bridging the gap between positions, though I would use eFG% when comparing two big time shooters.

What do the other numbers mean?

% Shots on the right side describes the percentage of possessions (with the player in the game) in which a player shot the ball. The higher the percentage, the more likely the other stats will dip a little (due to fatigue and the decrease of hot streaks). Though Harangody was only the third most effective scorer among the starting lineup of Mac, Tory, Kurz, and Ayers, he took shots on 37% of his possessions. Therefore, he was extremely valuable as an offensive weapon. As for the others, rebounding, steal, and block percentage (reb%, steal%, block%) is pretty self-explanatory. Same with assists and turnovers per 100 possessions. Free throw rate calculate how often a player gets to the line.

I'll refer to these stats a bit as the season goes on. Ken Pomeroy usually does a pretty good job posting similar stats on his site as the year goes on as well. Of course, I won't abandon the good old points and rebounds per game for you old timers (in part because I am too lazy to calculate all of this stuff every time).

So there you go. Didn't think you would be getting a math lesson today? Too bad. Back to the normal stuff later with a big game tomorrow. Have a good weekend!

EDIT: It's 16 degrees in South Bend. Just awful.


Anonymous said...

Nice analysis...the combination of tempo-free stats and a little common sense can really paint a nice picture of a player's contribution.

Anonymous said...

That first formula should probably be FGM + (1.5 x 3PM))/FGA, right?

Anonymous said...

Never mind. I forgot the 3PM are also included in the FGM. The 3-pointers are already getting a 1.5x factor this way.

BlackandGreen said...

Correct. Nice catch.