When I first started blogging, I tried to use numbers, graphs and charts to explain the 2012 Auburn-Clemson game. Fortunately (in a silver lining sort of way), Scot Loeffler killed off my enthusiasm pretty quickly and I basically reset the blog before the 2013 season. (I don't know how the original CaM writers made it through. In fact, former editor Chris Fuhrmeister's first post in the archives is about Loeffler's hiring).
And, though I dabble in stats here and there, I think I contribute much better with screen shots and GIFs with colorful circles and arrows. But that only works for football. I don't know much about basketball. I couldn't draw up a play on the grease board like the coaches do. Plus, there's a thing called burnout and off-seasons can be very therapeutic. But, as I told our own UcB,
@peggyrossmanith Thanks. I'm just getting in on the ground floor of this whole "Auburn basketball in the 21st century" thing.— WarRoomEagle (@WarRoomEagle) November 27, 2014
So how can I contribute to what should be an exciting time in Auburn basketball history in a knowledgable way while not over doing it? Automatic Charts! [wild applause]
Enough about me. Let's see what we can do with standard box scores. First, if you follow me on Twitter or read Walt's basketball review posts, you've likely seen the Game Flow Chart. This is nothing earth-shattering, as you can find one anywhere you can find a game tracker online. But, hey, now CaM has its own. This one also shows the maximum lead held by each team. Here's the one from Auburn's last game. Hover over it for the score at any point in the game.
Now to the reason I started this post in the first place. Dean Oliver, a statistician and basketball fan, did a lot to bring statistical analysis to the game he loved. Much of it I don't understand (or at least don't care to understand right now), but one of his contributions is relatively clean and simple, plus, as Oliver said himself, it "really starts allowing a strategic understanding of the game." That contribution was measuring four distinct aspects of the game itself. Shooting, turnovers, rebounding, and free throws. He called these four elements of the game the Four Factors.
Shooting: Effective Field Goal Percentage
A standard basketball box score will include a field goal percentage and a three point field goal percentage separately. Oliver's Effective Field Goal Percentage (eFG%) combines the two and gives a 50% bonus for made three-pointers (since three points is 50% more than two points). So eFG% is just a way to use one number to describe how well a team shoots the ball.
Turnovers: Turnovers per Possession
One key Oliver used in his analysis was to use "per possession" stats rather than "per game" stats. That way, hurry-up teams don't get a statistical benefit by simply running more plays. You've probably noticed something similar in football recently, too. As hurry-up teams can accumulate ridiculous yardage and point totals only because they run so many plays, "per play" or "per drive" stats begin to make more sense.
Basketball possessions are much more fluid than football possessions, so Oliver found a way to estimate the number of possessions with a combination of field goal attempts, rebounds, turnovers and free throws. Using that number, turnovers per possession measures how careful a team is with the ball, regardless of their pace of play.
Rebounding: Offensive Rebounding Percentage
Offensive rebounds give a team additional chances to score while defensive rebounds prevent those additional chances, so rebounding is the third factor. Simply put, offensive rebounding percentage just measures how often a rebounded ball was rebounded by the offense.
Free Throws: Free Throws Attempted per Field Goal Attempted
Finally, free throws have to be accounted for. Oliver determined that just getting to the free throw line was more important than actually making them. Think about it, if a team shoots 50% from the line but shoots 20 of them, they have as many points as a team that shoots 66% of 15 shots or 100% of 10 shots. (That's actually not that convincing now that i reread it. Oh well, I'm not the statistician.) In a particular game, missing free throws might matter, but more often than not, just getting those opportunities is more important. So Free Throw Rate is the fourth factor.
Combining the Four Factors
We don't have open access to great advanced basketball stats like we do for football (Bill Connelly's S&P+ or Brian Fremeau's FEI, for example). Pomeroy's site, kenpom.com, probably has lots of information behind its paywall, but we can at least measure these Four Factors with just the box score for each game. Furthermore, Oliver showed that each factor was not equal and actually gave them a weight. Using those weights, we can combine the factors into one rating for each game. I don't know how useful it is by itself, but I'm including it.
Here is the Four Factors Chart from Auburn's last game.
So that's it. I can have those two charts ready about 10 minutes after each game (assuming the box score is available from either auburntigers.com or espn.com). I looked around other SBNation college sites to see how their basketball coverage goes and I saw some neat things, but nothing I'm going to tackle until next season, unless I get an unexpected swath of free time.
For example, Connelly, writing for rockmnation.com, knocks it out of the park with charts, tables, and actual words. I'm excited about Auburn basketball and its future, but I'm not that excited.