Background on Basketball Statistics
The Four Factors come from the work done by Dean Oliver, whose book "Basketball on Paper" is like Moneyball for basketball. A lot of the work done by Ken Pomeroy is based on the fundamentals of Dean Oliver. All of the analysis that we do is based on his book as well.
The basic premise is that there are Four Factors that dictate how well a team does. The principles are Field Goal Percentage, Turnovers, Offensive Rebounds, and Free Throws. These are all fundamentals of basketball. However, the key difference for the Four Factors is that they've been modified to reflect possession-based statistics. In other words, a team that plays an extremely slow pace (Georgetown - 62 possessions / game) can now be compared to a team that plays at an extremely fast pace (Tennessee - 72 possessions / game). The team's points per game are not comparable, but their points per possession may be comparable.
What's a possession anyways?
A possession is anything that ends the action on the offensive end of the court. Any field goal attempt, any turnover, or even a Free Throw Attempt is a possession. Offensive rebounds negate possessions (because the player extends action on the offensive end of the court). The simplified equation to calculate possessions is FGA-OR+TO+0.475xFTA. Pretty basic, right?
Four Factor Review (Example - Florida Gulf Coast Game 3/4/08)
Offensive and Defensive Efficiency
The first thing to start with is the Efficiency review. Whichever team wins the Efficiency battle wins the game 100% of the time. Efficiency is as simple as how many points a team scores per 100 possessions. Another format of this is just points / possessions.
Against Florida Gulf Coast, we scored 67 points in just around 66 possessions. That works out to 1.01 points / possession, or an Offensive Efficiency of 100.9. Typically, an Offensive Efficiency of 100 is average, and an Offensive Efficiency above 110 is considered good. Therefore, our offensive output was around average.
Florida Gulf Coast scored only 37 points in 64 possessions, which is 0.58 points per possession, or an MU Defensive Efficiency of 58. Obviously, the lower a team's Defensive Efficiency, the better it did. I usually consider 90 to be the threshold for a poor efficiency outing, so FGC did not do very well at all, thanks to Marquette's defense.
Rating Teams and Tracking Trends
Pomeroy rates all of the teams on the basis of Efficiency Margin. This is simply the difference between Offensive Efficiency and Defensive Efficiency. In this game, our Efficiency Margin was 42.9 because the game was a blowout. In games that are fairly close, the Efficiency Margin is obviously close as well.
At Cracked Sidewalks, we track a moving average of Marquette's efficiencies to provide a view of how well the team is playing throughout the season. One always wants to see trends of Marquette's offensive efficiency sloping up, and of Marquette's defensive efficiency sloping down. In addition, by looking at the Efficiency breakdown following a game, one can tell if it was a game decided on the offensive end of the court (both teams with efficiencies above 100) or the defensive end (both teams with efficiencies less than 100).
The Four Factors
effective Field Goal Percentage (eFG%) - This is the same as regular Field Goal Percentage, but it takes into consideration the extra value of a three point basket. The example I always use is that a team gets the same amount of points (12) if they shoot 4/12 from three as if they shot 6/12 from inside the arc. eFG% adds in the extra value.
This is the most important of the four factors and it plays the greatest role in determining which team wins. I usually look above 55% as a good outing, and below 45% as a poor outing. Against, Florida Gulf Coast, Marquette won the effective Field Goal battle at 52.1%. Our game FG% was 45.8%. The formula for this calculation is (FGM + 0.5*3PM) / FGA.
Turnover Rate (TO Rate) - Turnover Rate is turnovers / possessions. A turnover rate of 20% is considered average. Less than 20% is considered good, and higher than 20% is considered bad. Against Marquette, Florida Gulf Coast turned the ball over 39% of the time, or one out of every two-three possessions. Marquette also did a poor job protecting the ball in this game.
Offensive Rebounding Percentage (OR%) - This is a little more sophisticated than just looking at total rebounds or a comparison of offensive rebounds between teams. The basic idea is that every defensive rebound is a potential offensive rebound (and vice versa). Consider for example:
- Team A misses forty shots, and grabs 15 offensive rebounds. This means that the opponent secured 25 defensive rebounds. OR% for Team A is 37.5% = 15 / (15 +25)
- Team B misses thirty shots, and grabs 15 offensive rebounds. This means that the opponent secures 15 defensive rebounds. OR% for Team B is 50% = 15 / (15 + 15)
Against FGC, we had 10 offensive rebounds and they had 17 defensive rebounds. Our OR% was 37.0% = 10 / (10 + 17). Florida Gulf Coast had 13 offensive rebounds and we had 23 defensive rebounds. FGC's OR% was 36.1% = 13 / (13 + 23). Even though Florida Gulf Coast had more offensive rebounds, we did better at offensive rebounding percentage (marginally).
Free Throw Rate (FTR) - There are a couple different ways of looking at Free Throw Rate. For consistency, the equation is Free Throw Attempts / Field Goal Attempts. In other words, what percent of the time did a field goal result in a free throw attempt? Against FGC, we took 24 free throw attempts against 48 field goal attempts. FGC took 10 free throw attempts against 47 field goal attempts.
In our next view, we'll take a look at how to read the individual player ratings through the FGC review.
I was told there would be no math. :)
ReplyDeleteI have read your stat reviews on this blog since you started writing them here and have always stuck to looking at the charts and seeing if we won in each column and that was it. Thank you for taking the time to explain the method and the thought process behind the stats you use so I can read them more intelligently. I have always enjoyed watching college basketball but never took the time to delve into all if the statistics involved and used in the different analyses and rankings.
ReplyDeleteI do have one question regarding the stats you use. What is the importance of the Free Throw Rate (FTR)? Why is knowing how many free throws attempted divided by the number of field goals attempted important? It seems to me that as long as your making your field goals, a lower FTR would be fine even if the other team is making a lot of free throws but missing their field goals. Also this does not seem to take into account if the team is making the free throws given. I would think the calculation given would tell how the refs are calling the game. Remember, I am a stat novice so any insight would be greatly appreciated.
mueng93, Free Throw Rate is important because it generally correlates to Defensive Efficiency.
ReplyDeleteYou'll have to copy and paste the following link. However, if you do so and scroll down, you will notice that our Opponent's FTR negatively correlates to our defensive efficiency. In other words, if our opponent is shooting a lot of free throws, it has a negative effect on our ability to do well defensively.
http://kenpom.com/expsked.php?team=Marquette&y=2008
Intuitively, this makes sense. If your opponents are getting to the line a lot, then your team has players in foul trouble/etc.
For considering "made" free throws as opposed to FTA, this is what I was referring to with the different ways of calculating free throw rate. For consistency, I just use the format that Pomeroy uses (FTA / FGA). The impact of a team's FT% shows up in their defensive efficiency.
Bottom line - If our opponents get to the free throw line at a high rate, Marquette is usually in trouble.
Thank you for the explaination and taking the time to respond.
ReplyDelete