Charting Duke v. Kansas State

[pfrduke]

The Smith-Curry-Dawkins-Singler-Big lineups were dominant against K-State, accounting for 21 Duke points and giving up just 6 (all other lineup combinations were a collective -1).

Individuals
Kyle Singler 74-54 (+20)
Andre Dawkins 43-25 (+18)
Nolan Smith 74-57 (+17)
Seth Curry 48-33 (+15)
Miles Plumlee 29-24 (+5)
Ryan Kelly 22-17 (+5)
Kyrie Irving 61-58 (+3)
Mason Plumlee 57-56 (+1)
Casey Peters 0-3 (-3)
Josh Hairston 2-7 (-5)
Tyler Thornton 0-6 (-6)

Per 40 Minutes
Andre Dawkins +37.9
Seth Curry +30.0
Kyle Singler +23.5
Ryan Kelly +20.0
Nolan Smith +19.4
Miles Plumlee +16.7
Kyrie Irving +3.9
Mason Plumlee +1.2
Josh Hairston -66.7
Tyler Thornton -120.0
Casey Peters -120.0

Lineups (Score, times used, margin)
Smith-Curry-Dawkins-Singler-Kelly (9-2, x1, +7)
Smith-Curry-Dawkins-Singler-Miles (6-0, x1, +6)
Irving-Smith-Dawkins-Singler-Mason (14-11, x3, +3)
Smith-Curry-Dawkins-Singler-Mason (6-4, x2 +2)
Irving-Curry-Dawkins-Singler-Miles (4-2, x1, +2)
Irving-Smith-Curry-Mason-Kelly (3-1, x1, +2)
Irving-Smith-Curry-Singler-Mason (7-6, x4, +1)
Irving-Smith-Curry-Singler-Miles (4-3, x1, +1)
Irving-Smith-Curry-Hairston-Mason (2-1, x2, +1)
Irving-Curry-Dawkins-Singler-Mason (2-2, x1, 0)
Irving-Curry-Dawkins-Singler-Kelly (2-2, x1, 0)
Irving-Thornton-Dawkins-Singler-Mason (0-0, x1, 0)
Thornton-Smith-Curry-Hairston-Mason (0-1, x1, -1)
Irving-Smith-Singler-Mason-Kelly (8-9, x3, -1)
Irving-Smith-Singler-Mason-Miles (12-13, x3, -1)
Irving-Thornton-Dawkins-Hairston-Mason (0-2, x1, -2)
Thornton-Curry-Peters-Hairston-Kelly (0-3, x1, -3)
Irving-Smith-Curry-Mason-Miles (3-6, x1, -3)

[ice-9]

Interesting stuff. Shows that even though Singler and Nolan didn't have the bet statistical games, they made a huge impact on the game. Dawkins, Curry and Kelly made very good use of the limited minutes they had. Kyrie and Mason looked great to the eyes, but their +/- was nothing special.

[Saratoga2]

Interesting stuff. Shows that even though Singler and Nolan didn't have the bet statistical games, they made a huge impact on the game. Dawkins, Curry and Kelly made very good use of the limited minutes they had. Kyrie and Mason looked great to the eyes, but their +/- was nothing special.

Maybe we shouldn't have played Kyrie as much? What drivel!

[flyingdutchdevil]

Maybe we shouldn't have played Kyrie as much? What drivel!

Mason was also useless. A +1 difference? He needs to shape up! God! :rolleyes:

[ice-9]

Maybe we shouldn't have played Kyrie as much? What drivel!

Uh, what part is drivel?

[JohnGalt]

Maybe we shouldn't have played Kyrie as much? What drivel!

Come on, Saratoga...that's not what he's saying.

I know the +/- doesn't take into account individual performance against other individuals, but rather the entire opposing team as a whole...regardless, I find it odd Kyrie's score isn't higher seeing as how poorly Pullen shot the ball. I suppose that found its way into helping Seth and Andre's stats moreso than Kyrie's.

[Bob Green]

Come on, Saratoga...that's not what he's saying.

I know the +/- doesn't take into account individual performance against other individuals, but rather the entire opposing team as a whole...regardless, I find it odd Kyrie's score isn't higher seeing as how poorly Pullen shot the ball. I suppose that found its way into helping Seth and Andre's stats moreso than Kyrie's.

The final margin of victory was 14 points. Duke went on a 9-0 run in the 1st half with Curry and Dawkins on the court, while Irving took a rest. This run partially explains Irving's lower than anticipated +/- number. For all of us who watched the game, Irving clearly was the MOTM so this once again illustrates the limitations of the +/- statistic when looking at a single game. However, I certainly agree with JohnGalt in that the +/- numbers are important and certainly not drivel. Perhaps Saratoga2's tongue was firmly in his cheek. At least that is the way I interpreted his comments.

[ice-9]

The final margin of victory was 14 points. Duke went on a 9-0 run in the 1st half with Curry and Dawkins on the court, while Irving took a rest. This run partially explains Irving's lower than anticipated +/- number. For all of us who watched the game, Irving clearly was the MOTM so this once again illustrates the limitations of the +/- statistic when looking at a single game. However, I certainly agree with JohnGalt in that the +/- numbers are important and certainly not drivel. Perhaps Saratoga2's tongue was firmly in his cheek. At least that is the way I interpreted his comments.

Setting Saratoga2's rude tone aside, I too had voted Kyrie MOTM, but while our eyes are subjective these numbers are not. Of course numbers are still contextual, so you can't just blindly follow them but Kyrie's +3 was surprisingly low enough to make me do a double take.

Maybe it can be explained away in that whenever Kyrie was in the game we were matched up against K-State's best five. Perhaps those layups Kyrie couldn't finish led to easy transition buckets for K-State. Or could it be that Dawkins and Curry were just that much more effective when Kyrie wasn't in the game.

We track +/- for a reason, and to dismiss the results because they don't support what our eyes see is...well, that's drivel. The whole point of these numbers is to provide a different perspective. E.g., like the one last season where the numbers showed Zoubek to be a valuable player while many DBR pundits called for his benching.

[COYS]

Setting Saratoga2's rude tone aside, I too had voted Kyrie MOTM, but while our eyes are subjective these numbers are not. Of course numbers are still contextual, so you can't just blindly follow them but Kyrie's +3 was surprisingly low enough to make me do a double take.

Maybe it can be explained away in that whenever Kyrie was in the game we were matched up against K-State's best five. Perhaps those layups Kyrie couldn't finish led to easy transition buckets for K-State. Or could it be that Dawkins and Curry were just that much more effective when Kyrie wasn't in the game.

We track +/- for a reason, and to dismiss the results because they don't support what our eyes see is...well, that's drivel. The whole point of these numbers is to provide a different perspective. E.g., like the one last season where the numbers showed Zoubek to be a valuable player while many DBR pundits called for his benching.

I actually think that what really killed Kyrie's +/- stats was a period in the second half when we were trying to slow the game down and three possessions in a row resulted in a turnover, two by Kyrie and one by Kyle. All three turnovers resulted in scores for the other team. This knocked his +/- back. He sat out for a bit after that and Duke settled down. When he came back in, we continued to slow the game down but at that point we were not interested in increasing the margin as much as continuing to slow the game down. I really think those three rapid-fire turnovers hurt his overall stats.

[CDu]

The final margin of victory was 14 points. Duke went on a 9-0 run in the 1st half with Curry and Dawkins on the court, while Irving took a rest. This run partially explains Irving's lower than anticipated +/- number. For all of us who watched the game, Irving clearly was the MOTM so this once again illustrates the limitations of the +/- statistic when looking at a single game. However, I certainly agree with JohnGalt in that the +/- numbers are important and certainly not drivel. Perhaps Saratoga2's tongue was firmly in his cheek. At least that is the way I interpreted his comments.

Bob's point here is important. Because points can be scored so easily in basketball, there is a tremendous amount of variability inherent in any single game +/-. Therefore, it's very possible that the +/- for any single game can be very disparate from the true quality of the player's performance.

There are a number of factors that can come into play. The quality of the opposing players on the floor matters. If you're in against the starters more than another player, then it stands to reason that your +/- might be worse (this may explain part of why the Dawkins/Curry line is so often impressive - we're subbing in starter-quality players while the opposition subs in weaker guards. But beyond just the matchups, it could be that a player happens to be in the game when a player on the other team hits a hot (or cold) stretch. Even just a few aberrant possessions can skew a single game +/-.

These are simply small sample size data points. The +/- numbers from any single game don't lie, but they also don't provide the full picture for that game. They are just one additional bit of information to use in evaluating a player's performance in any game. You shouldn't simply look at a good +/- and say "see, that player had a great game in spite of looking like he played poorly," and you also shouldn't look at a not-so-good +/- and say "well, I thought he'd played well, but the +/- proves I was wrong." And you also shouldn't say "well, the +/- disagrees with what I saw, so it's bunk."

In other words, single game +/- can be a useful evaluative tool, but it takes a great deal more care in interpretation/conclusion due to the high variability of scoring during any single game.

[Newton_14]

Bob's point here is important. Because points can be scored so easily in basketball, there is a tremendous amount of variability inherent in any single game +/-. Therefore, it's very possible that the +/- for any single game can be very disparate from the true quality of the player's performance.

There are a number of factors that can come into play. The quality of the opposing players on the floor matters. If you're in against the starters more than another player, then it stands to reason that your +/- might be worse (this may explain part of why the Dawkins/Curry line is so often impressive - we're subbing in starter-quality players while the opposition subs in weaker guards. But beyond just the matchups, it could be that a player happens to be in the game when a player on the other team hits a hot (or cold) stretch. Even just a few aberrant possessions can skew a single game +/-.

These are simply small sample size data points. The +/- numbers from any single game don't lie, but they also don't provide the full picture for that game. They are just one additional bit of information to use in evaluating a player's performance in any game. You shouldn't simply look at a good +/- and say "see, that player had a great game in spite of looking like he played poorly," and you also shouldn't look at a not-so-good +/- and say "well, I thought he'd played well, but the +/- proves I was wrong." And you also shouldn't say "well, the +/- disagrees with what I saw, so it's bunk."

In other words, single game +/- can be a useful evaluative tool, but it takes a great deal more care in interpretation/conclusion due to the high variability of scoring during any single game.

Great points CDu. Another killer is end of games with walk-ons playing. We only have 2 walkon's this year so 3 of our scholarship players have to be on the floor during this time. In almost every blowout game this year where our walkons got minutes, we were outscored quite a bit as the other team had their better players still on the floor. That has hurt guys like Curry, Thornton, Kelly, and Hairston.

So that has to be accounted for when considering the data.

[JG Nothing]

For all of us who watched the game, Irving clearly was the MOTM so this once again illustrates the limitations of the +/- statistic when looking at a single game.

On the other hand, maybe the +/- statistic illustrates the limitations of assessing performance simply from watching the game. ;)

Basketball is a team game. Irving played three-quarters of the game and, during that stretch, Duke was three points better than an excellent Kansas State team. In this case, a three point margin is pretty darn good. However, Smith, Curry, Dawkins, Singler, plus a big was easily the most effective lineup at plus 15. That lineup broke the game wide open and, on Tuesday night, was the best lineup by far.

Yes, the +/- statistic has its limitations, but here it may show two things: (1) people other than Irving had great games and (2) the whole is greater than the sum of its parts.

[Saratoga2]

On the other hand, maybe the +/- statistic illustrates the limitations of assessing performance simply from watching the game. ;)

Basketball is a team game. Irving played three-quarters of the game and, during that stretch, Duke was three points better than an excellent Kansas State team. In this case, a three point margin is pretty darn good. However, Smith, Curry, Dawkins, Singler, plus a big was easily the most effective lineup at plus 15. That lineup broke the game wide open and, on Tuesday night, was the best lineup by far.

Yes, the +/- statistic has its limitations, but here it may show two things: (1) people other than Irving had great games and (2) the whole is greater than the sum of its parts.

I was away for Thanksgiving yesterday so couldn't get involved in the discussion. It is my view that +/- numbers developed from a team game with 10 total players is a dynamic process and that it is a fallacy to try to draw up conclusions about individual player performances from those statistics.

Kyrie was recognized almost unanimously as MOTM because he was such a threat to penetrate, the Duke offense stayed fluid and effective with him at point and he played very effective defense. To imply from the statistics that he wasn't particularly important draws the use of the +/- statistics in that way into serious question.

Watching Nolan run point gives one a comparison. More dribbling, more bringing the ball to the sideline and getting poorer shots. It's not that Nolan played a poor game, he just does much better off the ball.

The eyeball test agrees with the +/- statistics for Andre, as he had an exceptional game on both sides of the ball.

I would vote that a close watching of the game provides a much surer way of determining how well a player has done in a game.

[Kedsy]

I was away for Thanksgiving yesterday so couldn't get involved in the discussion. It is my view that +/- numbers developed from a team game with 10 total players is a dynamic process and that it is a fallacy to try to draw up conclusions about individual player performances from those statistics.

Kyrie was recognized almost unanimously as MOTM because he was such a threat to penetrate, the Duke offense stayed fluid and effective with him at point and he played very effective defense. To imply from the statistics that he wasn't particularly important draws the use of the +/- statistics in that way into serious question.

Watching Nolan run point gives one a comparison. More dribbling, more bringing the ball to the sideline and getting poorer shots. It's not that Nolan played a poor game, he just does much better off the ball.

The eyeball test agrees with the +/- statistics for Andre, as he had an exceptional game on both sides of the ball.

I would vote that a close watching of the game provides a much surer way of determining how well a player has done in a game.

Well, I think there are too many independent factors that can affect +/- for the statistic to be especially meaningful in one game. So if that's what you're saying I agree.

However, I think as you get more games in, the cumulative data is very instructive, and can show you things the eyeball test cannot, and thus should not be ignored.

Looking at the cumulative +/-, then, Kyrie has our fourth best +/- total (fifth best average). That puts him behind Kyle and Nolan, which shouldn't surprise because they're seniors and perhaps know how to do the little things to win a bit better than Kyrie does. He's also behind Andre for total +/-, and behind both Andre and Miles in +/- per 40. What I think that shows is that Andre is playing at a very high level. Having the extra shooter in there makes us a better offensive team and his defense has been good enough that there's no noticeable defensive drop off to the "small" lineup (which most of us were expecting if Seth were playing the 3). But does it mean Andre is better than Kyrie? Of course not. Does it mean if Andre took Kyries minutes we'd be a better team? Of course not. It simply means Andre has been playing very well.

Miles's outstanding +/- is more of a surprise. To me it says the stats can show you things the eye test doesn't. He hasn't scored much or rebounded as much as we expected, and he makes noticeable miscues on offense. So the eye test says he's been a disappointment. But his defense (and perhaps other things that are difficult to see while watching the game) has been so good that through five games our team has performed at it's best level while he's been on the floor.

My point is, while I agree the +/- is flawed for single game comparisons, it's very valuable to see a player's contributions that don't end up on the stat sheet.

[JG Nothing]

I was away for Thanksgiving yesterday so couldn't get involved in the discussion. It is my view that +/- numbers developed from a team game with 10 total players is a dynamic process and that it is a fallacy to try to draw up conclusions about individual player performances from those statistics.

Kyrie was recognized almost unanimously as MOTM because he was such a threat to penetrate, the Duke offense stayed fluid and effective with him at point and he played very effective defense. To imply from the statistics that he wasn't particularly important draws the use of the +/- statistics in that way into serious question.

Watching Nolan run point gives one a comparison. More dribbling, more bringing the ball to the sideline and getting poorer shots. It's not that Nolan played a poor game, he just does much better off the ball.

The eyeball test agrees with the +/- statistics for Andre, as he had an exceptional game on both sides of the ball.

I would vote that a close watching of the game provides a much surer way of determining how well a player has done in a game.

I never said Irving "wasn't particularly important." I actually said the team was "pretty darn good" when he was on the floor. However, Smith, Curry, Dawkins, Singler, plus a big was easily the best lineup that night in terms of the scoreboard. I find it hard to believe that a plus 15 is not meaningful, particularly since it was accumulated during different stretches of the game in both the first and second halves. There may be a couple reasons for the greater effectiveness of that lineup. First, those players have been around each other for well over a year and are use to playing together. Second, we may be a much better defensive team when Singler plays the four, which always happens when Curry and Dawkins are on the court together.

Irving played a great game (although 5 for 12 is nothing special compared to the rest of the team, which shot 21 for 42) and he was particularly impressive at getting to the basket almost at will. However, as I said before, the whole is greater than the sum of its parts and other players also had great nights. Curry's defense was stellar and Dawkins was lights out from the field.