What do you value in a box score (when evaluating players)?

[ceiling raiser]

The traditional box score, since 77-78 (when individual turnovers were added), has consisted of the following:

MP-FG-FGA-FG%-3P-3PA-3P%-FT-FTA-FT%-ORB-DRB-TRB-AST-STL-BLK-TOV-PF-PTS

Removing FG%, 3P%, FT%, TRB, PTS (since they can be derived from other numbers), we get a few pieces of information describing events that happened over the course of games. Here are my general thoughts on each:

MP - I do have some concerns about minutes being recorded in the pre-PbP era, but I trust it for the most part. Straightforward enough.

FG - Not too comfortable judging a player by made FG, since it depends a ton on role, coverage faced, and momentum in the course of games.

FGA - I think this is generally accurate, but there are some issues. In particular, in some older games, there was some issue of official attempts in the box score matching up with video (Jerry West vs the Celtics was one such guy, but it was mentioned that there's a possibility that blocked FGA were not recorded as misses, but as turnovers). General note though, I do like FG-FGA zone diagrams, if it's made clear that the goal is to ascertain a guy's scoring ability relative to other zones (taking into consideration context, such as coverage faced based on how much a guy is respected in a given zone).

3P - Same as FG.

3PA - Same as FGA.

FT - Not a real issue, but rule changes can impact this greatly.

FTA - Same as FTA.

ORB - Scorekeepers are better now than they've ever been, but I'm pretty uncomfortable with this stat. Star players, big-name rebounders, and guys playing at home in general seem to get the benefit of the doubt in traffic (again, this may not be as big of an issue as it once was, but it still happens, including when rebounds are tapped by one player and secured by another). In addition, individual rebounding totals can belie a player's rebounding ability and impact (depending on team rebounding schemes, guys may box out, and not grab as many boards; teams also may eschew offensive rebounding to get back on transition defense).

DRB - Same as ORB.

AST - Two issues with this. It's not a perfect indicator of passing ability (and could more so reflect ball dominance), and there's inconsistency in how assists are credited. Home scorekeepers may be more liberal in handing out assists to their own guys, and the definition of assists seems to have changed across eras.

STL - Similar to rebounds, in traffic they might be assigned to higher profile players involved. I don't mind this as much, but I don't think it's appropriate when used as a proxy for defensive ability.

BLK - Same as rebounds and steals, things get dicey in traffic, and there's an issue when blocked shot totals are conflated with

TOV - Home/away scorekeeping bias (particularly on bad passes, stars may/may not get the benefit of the doubt). Also linked to ball-dominance.

PF - No issue whatsoever. Refs don't always get foul calls right, but they form the framework of the game.

Apologies for ranting, and feel free to ignore some or all of the above. I just wanted to start a conversation about box scores to get an idea of how the PC board feels about some of the constituent stats. I'm open-minded, so if good evidence is presented/arguments are made as to why some columns should be highly regarded, I'm perfectly willing to adjust my stance. As it stands though, I look to the box score to get a vague idea of a player's role on a team (as opposed to performance), when I haven't watched enough tape to make this determination on my own.

[Quotatious]

I think you have to look at ALL of these boxscore categories to get still a rather vague idea of how good a player is, or what his role on a team might be (I think you can derive roles from seeing boxscores of a few consecutive games of a team, assuming you didn't watch them play during that particular stretch, or had no idea about a player or a team, at all, never watched them - I think that looking at a boxscore of just one game wouldn't be more than 65% accurate - there are so many variables based on the flow of the game - for example a role player taking and/or making a few more shots than he usually would've taken, because he was left wide open more often than usual, etc.).

In general, I'd say that knowing minutes played, points, rebounds, assists and field goal percentage would give you the most basic, somewhat reliable knowledge. I didn't mention steals or blocks, because I'm generally reluctant of evaluating a player's defense based on those numbers - it would indicate whether a player is a good help defender, more often than not, but you would never know that, for example, Bruce Bowen was a great defender, if you just looked at his boxscore defensive stats.

[ceiling raiser]

I think you have to look at ALL of these boxscore categories to get still a rather vague idea of how good a player is, or what his role on a team might be (I think you can derive roles from seeing boxscores of a few consecutive games of a team, assuming you didn't watch them play during that particular stretch, or had no idea about a player or a team, at all, never watched them - I think that looking at a boxscore of just one game wouldn't be more than 65% accurate - there are so many variables based on the flow of the game - for example a role player taking and/or making a few more shots than he usually would've taken, because he was left wide open more often than usual, etc.).

In general, I'd say that knowing minutes played, points, rebounds, assists and field goal percentage would give you the most basic, somewhat reliable knowledge. I didn't mention steals or blocks, because I'm generally reluctant of evaluating a player's defense based on those numbers - it would indicate whether a player is a good help defender, more often than not, but you would never know that, for example, Bruce Bowen was a great defender, if you just looked at his boxscore defensive stats.

Thanks for the response.

I think my main issue is with assists/blocks/steals/turnovers, but I'm not sure if I should have one. I know they have some value, but I'm not quite sure what it is.

Regarding rebounding, the on/off splits here:

http://www.basketball-reference.com/pla ... -off/2001/

for DRB%/ORB% are very useful IMO, since both (to me at least) are really team efforts. I need to try and play around on stats.nba.com and media.nba.com to see if I can find those pages for 96-97 through 99-00 for the season and playoffs.

As I said in the OP, I love shot charts (and synergy-type breakdowns) for scoring. I really appreciate the efforts of guys like Dipper 13 and others to provide that data for seasons predating the PbP era.

[john248]

From the traditional box score, I generally look at points, FGA, rebounds, and turnovers. There was a time when seeing a center block a shot near the rim was almost as exciting as a dunk, but the game isn't as compact as it used to be.

[ceiling raiser]

Just thought I'd bump this thread since a lot of box score-based player evaluation is picking up steam in the project. Would love to hear from more people on the topic.

[ThaRegul8r]

Speaking for myself, there's nothing in the box score I value in particular when evaluating players. I'm interested in how much a player helps his team win or lose. That's it. So the pertinent question for me is: What does the team need from a player in order to win at that particular point in time, and did that player provide it? There is no specific box score for that, as box score production will vary, as what a team needs from a player will vary from game to game and opponent to opponent.

For example, Magic Johnson could put up 42/15/7 to help his team win a title one year, then take only three shots another game and still help his team win a title another year. The box score production for those two games are different, yet he still did what the team needed in order to win in each instance.

[Clyde Frazier]

For example, Magic Johnson could put up 42/15/7 to help his team win a title one year, then take only three shots another game and still help his team win a title another year. The box score production for those two games are different, yet he still did what the team needed in order to win in each instance.

I'm reminded of anthony davis controlling the NCAA championship game in 2012 without scoring the entire first half. It was pretty incredible. Also reminded of lebron controlling an entire finals game passing out of the post in 2012.

Pretty sure it was the game when mike miller went off, but sorta drawing a blank here.

http://www.basketball-reference.com/box ... 10MIA.html

[Winsome Gerbil]

Basically all the stuff EXCEPT the scoring columns. How is a guy impacting the game when he's not looking for his own? Assists, rebounds, steals, blocks. Show me somebody who is filling up the whole boxscore and I'll show you a real basketball player.

I also like to see guards who rebound/steal/get tot he line, show me they play physical. And amongst star level players, or guys accused of being stars, I look at that assist column a lot. Can they create? do they make teammates better? Or are they just a finisher their teammates have to spend effort making great?

[ronnymac2]

For offense, I look at points, assists, offensive rebounds, turnovers, and makes and attempts from the field and free throw line. Then I check out the minutes played and, if available, individual ORTG along with USG%. Basically I can see how efficient a player was using his own possessions and get a grasp of the offensive load he had to carry.

Even with assists, this type of analysis doesn't take into account opportunities creates, spacing, screens, passing in general, offensive rebounding activity (to draw fouls or help create chaos so another player can hit the offensive glass), HOW a player is scoring (post up, iso, flow of offense, long 2's), or the effect a player is having on his teammate's production. So it is limited.

But as far as showing the efficiency of the offensive possessions used, I really like looking at all the boxscore stats and then looking at minutes, individual ORTG, and USG%. I feel like pairing those last 2 gives us a better sense of how turnovers affect a player's offensive efficiency.

For defense...I suppose defensive rebounds, minutes, and defensive rebound rate if available. None of the boxscore stats are very reliable as far as measuring defensive prowess though. I mean, steals and blocks are indicators, but you can't really get much from them.

So Minutes is probably the thing I use the most that is most reliable.

[BmanInBigD]

Points, man, points!!

82! 62 in one quarter! 50-point games all the time!

It's all about the points, mang! Kobe's the best ever!!!

[trex_8063]

Points, man, points!!

82! 62 in one quarter! 50-point games all the time!

It's all about the points, mang! Kobe's the best ever!!!

Not sure if you're serious or not. If you are indeed championing for Kobe, though, I'd have you consider that within the realm of the (mostly) civil discourse that occurs on this and other basketball forums, this type of post probably does more harm than help to Kobe's legacy.

It's so hyperbolic and narrow-focused that the people who think highly of Kobe can't take it seriously. And on the flip-side it's exactly the kind of zealous rhetoric that his most out-spoken critics latch on to as a means of trying to discredit any and all of Kobe's supporters.

[Quotatious]

Not sure if you're serious or not. If you are indeed championing for Kobe, though, I'd have you consider that within the realm of the (mostly) civil discourse that occurs on this and other basketball forums, this type of post probably does more harm than help to Kobe's legacy.

It's so hyperbolic and narrow-focused that the people who think highly of Kobe can't take it seriously. And on the flip-side it's exactly the kind of zealous rhetoric that his most out-spoken critics latch on to as a means of trying to discredit any and all of Kobe's supporters.

He's clearly joking. Seems like he's making fun of the Kobe stans who like to mention impressive, small sample size scoring averages of his, or even just single games, to make a point about his greatness, ignoring the fact that over a big sample size (entire season) Kobe generally didn't seem to be more impressive than about 15 or 20 other players in NBA history (perhaps even more, if we include single seasons, as then you have terrific, but outlier campaigns like 2003 McGrady, which certainly rivals prime Kobe).

[john248]

Not sure if you're serious or not. If you are indeed championing for Kobe, though, I'd have you consider that within the realm of the (mostly) civil discourse that occurs on this and other basketball forums, this type of post probably does more harm than help to Kobe's legacy.

It's so hyperbolic and narrow-focused that the people who think highly of Kobe can't take it seriously. And on the flip-side it's exactly the kind of zealous rhetoric that his most out-spoken critics latch on to as a means of trying to discredit any and all of Kobe's supporters.

He's clearly joking. Seems like he's making fun of the Kobe stans who like to mention impressive, small sample size scoring averages of his, or even just single games, to make a point about his greatness, ignoring the fact that over a big sample size (entire season) Kobe generally didn't seem to be more impressive than about 15 or 20 other players in NBA history (perhaps even more, if we include single seasons, as then you have terrific, but outlier campaigns like 2003 McGrady, which certainly rivals prime Kobe).

You make it sound like being the 21st rated player of all time is a bad thing, even if it is Kobe. Then you say if "if include single seasons" but who cares. This thread isn't about that.

[Quotatious]

You make it sound like being the 21st rated player of all time is a bad thing, even if it is Kobe. Then you say if "if include single seasons" but who cares. This thread isn't about that.

I admit it was probably unnecessary to add that thing about Kobe, and no, it wasn't my intention to make it sound like being the 21st best player of all-time is a bad thing (and I was talking only about peaks - career-wise, he's clearly higher) - I just wanted to make a point that quite a few players ever reached the level of prime Kobe, at some point in their careers, but oh well, it seems that you always have to be prepared for someone who will find it a bit offensive...What makes Kobe so great is that he sustained his peak level of play for a long time - I mean, it's not even clear which year was his peak - he was at his peak basically for 4 straight seasons (2006-09), and had a few other amazing seasons, too (2001, 2003, 2010).

Okay, that's all I had to say. Just wanted to make it clear that I'm not a Kobe hater, even if I may perhaps sometimes come off as one, to some people who perceive anyone who ranks him lower than 11 as a hater.