Charting Duke vs. Cornell

[Jumbo]

A few interesting stats from this game. Taylor King's stellar plus/minus took a big hit, while the small lineup that we expect to see finish games (Paulus, Scheyer, Nelson, Henderson, Singler) was far and away Duke's most effective group.

Individuals
Jon Scheyer 66-48 (+18)
Gerald Henderson 71-54 (+17)
Kyle Singler 60-44 (+16)
DeMarcus Nelson 60-48 (+12)
Greg Paulus 49-42 (+7)
Nolan Smith 32-27 (+5)
Brian Zoubek 20-16 (+4)
Lance Thomas 18-17 (+1)
Jordan Davidson 0-0 (0)
David McClure 20-23 (-3)
Taylor King 9-16 (-7)

Per 40 Minutes
Jon Scheyer +24
Gerald Henderson +22.7
Kyle Singler +22.1
Brian Zoubek +16
DeMarcus Nelson +15.5
Greg Paulus +12.2
Nolan Smith +11.1
Lance Thomas +4.4
Jordan Davidson 0
David McClure -8.6
Taylor King -46.7

Lineups
Paulus-Scheyer-Nelson-Henderson-Singler (x3) 20-7 (+13)
Paulus-Scheyer-Nelson-Henderson-Thomas 5-0 (+5)
Smith-Scheyer-Henderson-Singler-Zoubek 9-5 (+4)
Smith-Scheyer-Nelson-McClure-Singler (x2) 6-3 (+3)
Smith-Nelson-Henderson-Singler-Zoubek (x2) 11-9 (+2)
Paulus-Scheyer-Nelson-Henderson-McClure 5-3 (+2)
Paulus-Scheyer-Henderson-King-Thomas 4-2 (+2)
Smith-Nelson-Henderson-McClure-Singler 1-0 (+1)
Paulus-Scheyer-Nelson-McClure-Thomas 2-2 (0)
Paulus-Davidson-Scheyer-Henderson-Singler 0-0 (0)
Scheyer-Nelson-Henderson-Thomas-Zoubek 0-0 (0)
Smith-Scheyer-Nelson-Henderson-Singler 5-6 (-1)
Paulus-Scheyer-Henderson-Singler-Thomas 5-7 (-2)
Paulus-Scheyer-Henderson-McClure-Singler 3-5 (-2)
Paulus-Scheyer-Nelson-King-Thomas 2-4 (-2)
Smith-Scheyer-Nelson-Singler-Zoubek 0-2 (-2)
Paulus-Smith-Scheyer-Nelson-Thomas 0-2 (-2)
Paulus-Nelson-Henderson-McClure-King 3-10 (-7)

[Jumbo]

Should I stop posting these? Doesn't seem like they're drawing much interest anymore.

[JasonEvans]

Should I stop posting these? Doesn't seem like they're drawing much interest anymore.

I disagree!! I find them fascinating. There is just not much to comment upon after reading the numbers... at least not for me after last night's snoozer.

I really wish the Smith plus-minus had been better as I like the dimension he adds at PG because of his insane quickness, though I am not convinced he should be the starter. I also like what Zoubek brings much of the time, though I don't think he is nearly fast enough to be more than about a 15 minute per game player for Duke. Duke succeeds by using its speed and athleticism on D and Zoub hinders our ability to do that a bit.

Taylor King was godawful last night.

--Jason "look at that, I commented a bit ;) " Evans

[Johnboy]

Should I stop posting these? Doesn't seem like they're drawing much interest anymore.

I never (rarely?) comment, but I always check them out with great interest. thank you for providing this service to our little community.

[Jeffrey]

Taylor King was godawful last night.

--Jason "look at that, I commented a bit ;) " Evans

Hi,

As were the announcers for not being with it enough to figure out why he wasn't playing in the 2nd half. I was amazed by some of their lame comments concerning Taylor towards the end of the game.

Best regards,
Jeffrey

[bird]

I too really look forward to the quantifications.

I was telling my daughter when Smith and Zoubek took the floor that "it doesn't matter who starts as much as who finishes," a line stolen from this board, and then, guess what, the money group was the small lineup with Scheyer and Paulus in there -- making me look smart for my daughter.

[Troublemaker]

Yeah, I love looking at plus-minus, and it's been a welcome addition to NBA box scores this season. However, there usually isn't much room to comment about them on a game-to-game basis, so no commentary does not indicate a lack of interest.

But really, you should only continue charting if you still find it interesting, and not for our benefit. If it ever becomes a burden...

[Jumbo]

Not a burden at all. But if I'm just carrying on a conversation with myself, it's pointless. Seems that people still read the stuff, though, so I'll keep posting. The season totals are updated and I've added some net numbers. We're far enough into the season that some trends have emerged there.

[gofurman]

Not a burden at all. But if I'm just carrying on a conversation with myself, it's pointless. Seems that people still read the stuff, though, so I'll keep posting. The season totals are updated and I've added some net numbers. We're far enough into the season that some trends have emerged there.

What is this plus-minus? Can someone explain? I am sure it is some sort of "net" numeric but of what?

[CDu]

What is this plus-minus? Can someone explain? I am sure it is some sort of "net" numeric but of what?

It is the point differential observed when you are on the floor. For example, if you enter the game with the score tied at 10-10, and then subsequently leave the game with the score 15-10, you'd have a +/- of +5. A player's +/- stat is just the sum of all points scored by Duke when that player is on the floor minus the sum of all points scored by the opposition when that same player is on the floor. Similarly, the +/- per 40 minutes is an adjustment for the amount of minutes a player has played, which can be calculated as +/- divided by total minutes and then multiplied by 40.

[MChambers]

Not a burden at all. But if I'm just carrying on a conversation with myself, it's pointless. Seems that people still read the stuff, though, so I'll keep posting. The season totals are updated and I've added some net numbers. We're far enough into the season that some trends have emerged there.

I read your post on this after each game. What I found most interesting today was that, although Jon Scheyer had some ugly lines in the boxscore, he was tops in the plus/minus category.

[CDu]

I read your post on this after each game. What I found most interesting today was that, although Jon Scheyer had some ugly lines in the boxscore, he was tops in the plus/minus category.

Yeah, sometimes the plus/minus shows funny things from game to game as compared to the box score. That can be either because a player had a better game than the box score would indicate OR that the plus/minus for any particular game can be a misleading metric (i.e., happening to be on the floor at the right times). It's one of the difficulties in assessing much with a single game plus/minus. Jumbo: please don't read this as a vote in favor of you not posting these - I also enjoy reading the charting of each game.

In this case, I wouldn't have said Scheyer had a particularly outstanding game (not having even seen the box score). So I'd guess this was in some part a case of being on the floor at the right times. Though there have certainly been games where Scheyer's contributions haven't been fully captured by the box score, and he's certainly one of our best players, as evidenced by the fact that he's consistently been among our top 3-4 players in plus/minus per 40 minutes.

[Jumbo]

Yeah, sometimes the plus/minus shows funny things from game to game as compared to the box score. That can be either because a player had a better game than the box score would indicate OR that the plus/minus for any particular game can be a misleading metric (i.e., happening to be on the floor at the right times). It's one of the difficulties in assessing much with a single game plus/minus. Jumbo: please don't read this as a vote in favor of you not posting these - I also enjoy reading the charting of each game.

In this case, I wouldn't have said Scheyer had a particularly outstanding game (not having even seen the box score). So I'd guess this was in some part a case of being on the floor at the right times. Though there have certainly been games where Scheyer's contributions haven't been fully captured by the box score, and he's certainly one of our best players, as evidenced by the fact that he's consistently been among our top 3-4 players in plus/minus per 40 minutes.

That's the thing -- the team just tends to play better when Scheyer is in. He didn't shoot well, but he moved the ball well, disrupted some plays on defense and was often in the right place at the right time. At this point, I don't take his high plus/minus as an accident anymore. He's leading the team in that category over the season by a fairly healthy margin, so I don't think he got "lucky" by any means against Cornell.

[CDu]

That's the thing -- the team just tends to play better when Scheyer is in. He didn't shoot well, but he moved the ball well, disrupted some plays on defense and was often in the right place at the right time. At this point, I don't take his high plus/minus as an accident anymore. He's leading the team in that category over the season by a fairly healthy margin, so I don't think he got "lucky" by any means against Cornell.

Minor quibble: Scheyer is leading in TOTAL plus/minus, but trails King (substantially) and Henderson in plus/minus per 40 minutes among players who actually play in real minutes. Still very good, but not leading by a healthy margin, either.

Also, I wasn't saying Scheyer's high plus/minus last night was due to luck. I'm just saying that it is certainly possible that some of Scheyer's numbers last night were due to timing rather than performance. I simply don't think Scheyer was the most valuable player on the floor for the team last night, and thus I think his plus/minus is somewhat inflated relative to performance. That's all I was trying to convey - that sometimes the plus/minus (if read improperly) can lead to faulty conclusions.

Could Scheyer also have had a better game than his box score indicates? Sure, but the two aren't mutually exclusive.

[Jumbo]

Minor quibble: Scheyer is leading in TOTAL plus/minus, but trails King (substantially) and Henderson in plus/minus per 40 minutes among players who actually play in real minutes. Still very good, but not leading by a healthy margin, either.

Also, I wasn't saying Scheyer's high plus/minus last night was due to luck. I'm just saying that it is certainly possible that some of Scheyer's numbers last night were due to timing rather than performance. I simply don't think Scheyer was the most valuable player on the floor for the team last night, and thus I think his plus/minus is somewhat inflated relative to performance. That's all I was trying to convey - that sometimes the plus/minus (if read improperly) can lead to faulty conclusions.

Could Scheyer also have had a better game than his box score indicates? Sure, but the two aren't mutually exclusive.

I think plus/minus is the best window we have into intangible contributions. I also think King's plus/minus is extremely inflated because the bulk of his success came against Duke's worst opponents. When Duke has played good teams, his minutes have declined dramatically, and that's why his +/- per 40 minutes is so good. Guys who play a ton of minutes (DeMarcus is an obvious example) actually get penalized because they rarely leave the court against good teams. The fact that Scheyer's raw plus/minus AND plus/minus per 40 are so high, despite his high mpg, is a feather in his cap.

[CDu]

I think plus/minus is the best window we have into intangible contributions. I also think King's plus/minus is extremely inflated because the bulk of his success came against Duke's worst opponents. When Duke has played good teams, his minutes have declined dramatically, and that's why his +/- per 40 minutes is so good. Guys who play a ton of minutes (DeMarcus is an obvious example) actually get penalized because they rarely leave the court against good teams. The fact that Scheyer's raw plus/minus AND plus/minus per 40 are so high, despite his high mpg, is a feather in his cap.

I agree with you on all of the things you just said. But this post doesn't make my previous post false (or make your previous post correct, as Henderson still leads Scheyer in plus/minus per 40). Hence it was only a minor quibble. One additional point I'd make is that, while I agree that plus/minus is the best we have, it is also still very flawed (especially in a single game context). That's kind of what I was saying in my post. Can it be more valuable than the box score? Sure. Can it also be more misleading than the box score? Absolutely.

Also, your points in this post are not really related to what I was saying with regard to Scheyer's plus/minus last night's game. Again, what you said here and what I said about the plus/minus from last night are not mutually exclusive.

[bird]

Here are the cumulative stats grabbed from the other thread:

Indviduals
Jon Scheyer +198 (733-535)
DeMarcus Nelson +187 (740-553)
Gerald Henderson +183 (627-444)
Greg Paulus +181 (655-474)
Kyle Singler +177 (672-495)
Taylor King +147 (404-257)
Nolan Smith +104 (409-305)
Lance Thomas +89 (413-324)
Brian Zoubek +76 (291-215)
Martynas Pocius +19 (74-55)
Jordan Davidson +12 (34-22)
David McClure +7 (128-121)

Let's say these are the "objective" stats.

If I were to rate players on overall contributions while on the floor, I might have a similar, but not identical ranking:

Nelson
Scheyer
Henderson
Paulus
Singler
Smith
Lance
Taylor
McClure
Zoubek

Nelson's minutes might skew his objective numbers down, as mentioned. Lance's recent performances may have skewed subjective judgment down. I would put a larger gap between Paulus and Henderson subjectively.

[socaldukie]

I personally like to look the +/- numbers. I look forward to these after each game. I appreciate the effort that goes into gathering this information. Box scores, for me, only tell a small story of a games' outcome. Hockey has been keeping these stats for years. They show some insights into team effectiveness from an individual perspective. Sure they are not perfect. But, you can see trends established over time. It is the best means to capture (objectively) intangible contributions to team success and failure.

Keep 'em coming Jumbo.

[CDu]

I personally like to look the +/- numbers. I look forward to these after each game. I appreciate the effort that goes into gathering this information. Box scores, for me, only tell a small story of a games' outcome. Hockey has been keeping these stats for years. They show some insights into team effectiveness from an individual perspective. Sure they are not perfect. But, you can see trends established over time. It is the best means to capture (objectively) intangible contributions to team success and failure.

Keep 'em coming Jumbo.

I pretty much agree (especially with the last sentence - definitely keep'em coming!). However, I think one slight caveat should be mentioned with regard to bringing up hockey. I think plus/minus is much more meaningful in hockey. For one thing, there are a LOT more games in a hockey season as compared to basketball (81 versus ~30-35), which helps minimize the impact of an outlier performance. For another, because hockey has much less scoring, there is less risk of outlier performances. And finally, there are just so few stats to measure in hockey (especially for defenders) that plus/minus is much more useful in rating players who aren't prone to get assists or goals, as compared to basketball where players can be rated for various contributions (points, rebounds, assists, steals, blocks, etc).

Again, that's not to say plus/minus doesn't have value. I think it's a good step in the progression of stat collection for basketball, and can aid in understanding intangible contributions that box scores don't capture. And hopefully, somebody will eventually figure out an even better measure of these intangible qualities in the future.

[Jumbo]

I agree with you on all of the things you just said. But this post doesn't make my previous post false (or make your previous post correct, as Henderson still leads Scheyer in plus/minus per 40). Hence it was only a minor quibble. One additional point I'd make is that, while I agree that plus/minus is the best we have, it is also still very flawed (especially in a single game context). That's kind of what I was saying in my post. Can it be more valuable than the box score? Sure. Can it also be more misleading than the box score? Absolutely.

Also, your points in this post are not really related to what I was saying with regard to Scheyer's plus/minus last night's game. Again, what you said here and what I said about the plus/minus from last night are not mutually exclusive.

Yeah, I'm working under the impression that we agree on pretty much everything.

[mgtr]

Assuming that Coach K is rational (which I do assume), another way of looking at the plus/minus is by simply looking at minutes played. If Coach K is a good coach, then the best players will probably get the most minutes. In the Cornell game, we have Nekson, Scheyer, Henderson and Singler all in the 29-31 minute range. Who would challenge that kind of decison making? Each of those players has been a Man of the Match this season.