clock menu more-arrow no yes mobile

Filed under:

Advanced Basketball Box Score Primer

NCAA Basketball: NCAA Tournament-Midwest Regional Jay Biggerstaff-USA TODAY Sports

If you follow me on Twitter, you’ve probably seen me tweeting out advanced box scores after each basketball game this year. A friend over at our Missouri sister-site, Sam Snelling at Rock M Nation, created this box score a few years back, and shared it with me to use for our purposes at College and Mag. They’ve been fun to share, but I thought it might be prudent to actually explain what all is in those box scores. I’ll try and keep things concise, but this article will mainly serve as a reference for those wanting to learn more about what all is in the box scores.

Here’s an example of the box score for the Colgate game. The first tab you see in the workbook is for the game on 11/18/19. After that, there are similar worksheets that show season stats and previous games’ box scores


Here’s the basics. For each player, you’ll see what you see in any box score - minutes, points, rebounds, assists, turnovers, steals, blocks, and fouls committed. What you’ll also see is columns for %MIN, OFFV%, and ORtg.

%MIN - This isn’t too complicated. This is just a calculation of time on the court vs on the bench. If a player played 40 minutes, in a regulation game, they’d have a 100% %MIN. If they played 45 minutes of an overtime game, same thing. Twenty minutes in a regulation game would be 50%. Easy enough?

OFFV% - This one’s a little more in depth. OFFV%, simply put, is a determination of how much the player contributed to his team’s offensive output compared to the rest of the team. The is calculated by dividing each player’s Adjusted Game Score (AdjGS, we’ll get to that later) by the team’s total points.

Generally speaking, this stats would be a way to determine a team’s “MVP” of any given game. The sum of each player’s OFFV% should equal 100%.

ORtg - Offensive Rating, by definition, is used to determine the number of points produced by a player per 100 individual possessions. It’s a pretty long and complicated formula that you can find here, but for simply scanning a box score, just know that a rating of 100 is about average. On offense, the higher the offensive rating the better, and the opposite will be true on defense. For even more reading, you can check this out.

Below the players stats, you’ll see team stats, which I don’t think need any explanation. Below that, you’ll see a few game stats. They are Poss, PPP, and PPM for both Auburn and the opponent.

Poss - This is an estimate on the total number of possessions in the game utilizing only box score stats. This number may be slightly off the true number of possessions, but without play-by-play data, this number will do well enough. Both teams should have the same number of possessions.

PPP - This is points per possession. The calculation is pretty straightforward, it’s just total points scored divided by the number of possessions. The winning team will always have a higher PPP than their opponent.

PPM - PPM is points per minute. Just like PPP, this is just points scored divided by minutes. If you’re feeling froggy with fractions you can divide PPM by PPP and pet possessions per minute.


There’s a ton of summary data down the spreadsheet from here, but I’m going to go ahead and work across. We’ll hit the summary data at the end.


PPS - You guessed it, this one is points per shot. The only trick here is that points scored from free throws are included, but FTA are not included in the shot total. So for example, if Austin Wiley scored 12 points on 4-6 FG and 4-5 FT, his PPS would be (8+4)/6 = 2.00. Therefore, guys who can get to the line get a little boost here.

ShPM - Shots per minute. Again, easy enough. Someone with a higher ShPM means they are the feature of the offense. While it’s a good indicator of who is the feature on a given team, it isn’t pace adjusted, so it shouldn’t really be used to compare players from different teams.

VersInd - This one’s actually pretty interesting. The versatility index measures a player’s ability to contribute in more than one way on the stat sheet, taking into account points, rebounds, and assists. If any one of these totals is zero, the VersInd will equal zero, so the stat is more valuable in a multi-game data set. states an average NBA player scores around a 5 in VersInd, while the best players score around a 10. Since the stat isn’t pace or minute adjusted, I would expect the scale would be lower for college players.

VI/Min - Here’s a time adjustment for you! Assuming pace is the same in the NBA and NCAA (it’s not), a player scoring a 10 in VersInd in the NBA would post an 8.3 in the NCAA, and a player scoring a 5 in the NBA would score a 4.2 in the NCAA. With VI/Min, though, the results should be standardized a bit.


TS% - True Shooting Percentage. This takes into account all FG and FT and attempts to assign an efficiency based on that. There’s a little extra math to the equation if you’re interested. This is intended to be an improvement on FG%.

FG% - Field goal percentage. Nothing complicated here. Shots made out of shots attempted, free throws excluded.

*eFG% - Effective Field Goal Percentage. I like this stat even better than TS%, because eFG% weighs two point shots and three point shots differently. Consider a scenario with two players, where Player 1 has a 40% FG% and Player 2 has a 50% FG%. However, Player 1 takes all three point attempts, while Player 2 takes only two point attempts. If each player takes 10 shots, Player 1 will have scored 12 points, while Player 2 will have scored 10 points. eFG% fixes that by assigning Player 1 a 60% eFG% and Player 2 a 50% eFG%. The formula can be found here.

2PT%/3PT%/FT% - Two point percentage, 3 point percentage, and field goal percentage. This is just a breakdown of the rate at which each type of shot is made.

*FTA/FGA - Free throw attempts per field goal attempt. This isn’t necessarily a stat that reveals a ton about player value, but it does reveal a player’s style. For most players, this will be a fairly low number (sub 0.5), but for bigs it’s usually quite a bit higher. More fouls occur down in the paint than in other parts of the floor. The benefit is that while most bigs don’t shoot threes, if they have a FTA/FGA around 1.0, then they are average three possible points per attempt. If they can hit free throws at a decent rate, that’s a huge plus to their offensive value.

3PTA/FGA - Three point attempts per field goal attempt. This is a fractional measure of the amount of 3 point attempts a player attempts out of their total shot attempts.

FTPM - Free throws per minute. This is just a minute-adjusted rate for free throw attempts.


Touches and Tch/Min - This is an estimate of the number of touches he player had during the game. You would expect to see a lot of these for the point guards, and less for other players, since the offense is generally run through the point guard. If you want the actual formula, you can find it here. There’s also a minute-adjusted rate, as well.

Usage% - Usage percentage is an estimate of the number of plays the offense runs that end in a player either shooting, getting fouled, or turning the ball over. In essence, this determines how often the player is the one “responsible” for how the possession ends while they are on the court. Ideally you want your best scorers to have the highest Usage%, otherwise the offense is running inefficiently. You can find the formula for it here.

Floor% - Floor percentage is, put simply, the ratio of possessions that end in points to total possessions each player is responsible for. This is pretty easy to look at from a team-level perspective, but a little more difficult to determine for individual players. Players who provide assists also receive a portion of the credit for creating scoring possessions. For more reading, let me take you deep into the early internet.

Tch/Pos - This is the average number of touches a player has per possession that they are on the court.

%Pass/%Shoot/%Foul/*%TO - These four stats break down what a player does with the ball whenever they get it. Optimally, your guards are going to pass more, your scorers are going to shoot more, and your bigs are going to get fouled. And hopefully nobody turns it over much. The four percentages should add up to 100% for each player. These number don’t do much to show you how effective a player is, but they can reveal if there is poor optimization of the offense. If you have a player with, for example, a low eFG% but a high %Shoot, you might need to adjust your offense to have him shoot less.


Columns AI-AZ in the box score break down the traditional box score stats on a per minute and per possession basis. I think you can handle interpreting those. The one thing I do feel the need to point out is OR% and DR%.

*OR% - Offensive Rebounding Percentage is the rate at which an offense rebounds it’s own missed shot attempts, free throws included. Offenses generally do this at a 32%-37% clip (meaning the defense gets the rebound around 63%-68% of the time), but it depends on the types of shots the offense is taking. Missed three point attempts are usually rebounded less often by the offense due to simply having more players away from the basket.

DR% - Defensive Rebounding Percentage is the rate at which the defense rebounds the opponents’ missed shot attempts, free throws included.

Generally, individual rebound rates aren’t super important, but instead the overall team rebound rates. However, players with extraordinarily high rebound rates are valuable. When you consider that for any given shot, ten players can rebound that ball, a player with a higher than 10% rebound rate on either side of the court is doing “better than average”. For example, a player with a 30% DR% is grabbing 30% of all possible rebounds when he’s playing defense, and the other 9 players are splitting the remaining 70%.


GmSc, GmSc/G, and GmSc/M - Game Score is an attempt to assign players a value for their overall performance, and summing up the Game Scores for each player yields a “team score” that can be compared against the opponent. Game Scores can be read similarly to point totals, but they are intended to include other contributions, such as scoring efficiency, rebounds, assists, steals, blocks, turnovers, and fouls. The formula can be found here. The next two columns are rate-based on a per-game and per-minute basis.

AdjGmSc, AdjSC/G, and AdjSC/M - This is an adjustment of a player’s game score so that when all of the player’s scores are summed, the total for the team matches the teams actual score. Mathematically, Adjusted Game Score takes each player’s game score as a fraction of the team’s game score, then multiplies it by the total points scored.


Below columns G-P on the left side of the worksheet, you can find a few different printouts that summarize the box score in a much cleaner fashion that scrolling across the worksheet horizontally. You’ve already been introduced to all of the statistics, but let’s talk about what each printout contains.


In the breakdown above, I put an asterisk (*) by four stats - eFG%, FTA/FGA, %TO, and OR%. These four stats, or Four Factors according to Dean Oliver, are the key stats that are most telling about a game. Performance in these stats can be used to model an expected win probability to a reasonable degree of accuracy, although on a game to game basis statistical anomalies are bound to occur. If you’re familiar with Bill Connelly’s advanced football box scores, the concept is very similar. This first printout allows a 1:1 comparison to be made to the opponent in the Four Factors, plus other game-defining numbers.

A few new stats are introduced as well.

BCI - Ball Control Index is, as far as I can tell, unique to this box score thanks to Mr. Snelling. You can see the formula above, but it’s essentially a way to see how well a team is controlling the ball. Sitting in the mid 1’s is fine, but if a team has a BCI above 2.0, they’re probably going to win pretty easily.

Expected Off. Rebounds - This is an attempt to estimate how many offensive rebounds a team would have based off of league average offensive rebounding rate for a given shot type (32% 3PA, 37% 2PA, 18% FTA). This expected number of offensive rebounds can be compared to the actual number of offensive rebounds to see if each team is under- or over- performing in terms of grabbing offensive boards.


Nothing new to see here, but this is probably the cleanest place to see each players stats. GS/Min helps provide context for who the most valuable players were on a per/minute basis, as well.


All of these stats were discussed above, but again, this is organized a little more cleanly. Remember, Usage% is a measure of the percentage of possessions each player is “responsible” for on the offensive end, and Floor% is the percentage of possessions that end in points out of the possessions each player is responsible for.

Alright everyone. 2300+ words later, you are now EXPERTS in how to read these box scores. Going forward, I’m going to try to share these on the site for each game and hit on a few key observations, both for the game and for the season totals.