PDA

View Full Version : Charting Duke vs Miami



loran16
01-03-2011, 05:05 PM
The Numbers for this game are interesting, but should be taken with some caution, as Nolan's personal run came during a single lineup that didn't play together again (Smith, Curry, Singler, Kelly, Mason). Still, the numbers show the good game by Kelly.

NOTE: Smith and Singler played all 40 minutes.

Ryan Kelly +12
Nolan Smith +11
Kyle Singler +11
Mason Plumlee +11
Miles Plumlee +6
Seth Curry +2
Andre Dawkins +2

Per 40 Minutes
Kelly +19.2
Mason Plumlee +17.6
Singler +11
Smith +11
Miles Plumlee +10.43
Curry +4.44
Dawkins+2.75


Smith-Curry-Singler-Kelly-Mason +8 (11-3) (1x)
Smith-Dawkins-Singler-Miles-Mason +6 (18-12) (4x)
Smith-Dawkins-Singler-Kelly-Mason +2 (12-10) (3x)
Smith-Curry-Singler-Kelly-Miles +1 (12-11) (3x, Starting Lineup)
Smith-Dawkins-Singler-Kelly-Miles +1 (10-9) (2x)
Smith-Curry-Singler-Miles-Mason +0 (0-0) (1x)
Smith-Curry-Dawkins-Singler-Miles -2 (2-4) (2x)
Smith-Curry-Dawkins-Singler-Mason -5 (9-14) (3x, Ended the Game)

Kedsy
01-03-2011, 08:02 PM
Funny, when everyone was ragging on Miles, he was routinely putting up great plus/minus numbers, and now people are snarking on Mason and his plus/minus line is 2nd on the team. Just goes to show that the eyes don't always tell the whole story.

loran16
01-03-2011, 10:53 PM
Funny, when everyone was ragging on Miles, he was routinely putting up great plus/minus numbers, and now people are snarking on Mason and his plus/minus line is 2nd on the team. Just goes to show that the eyes don't always tell the whole story.

I'm not wholly disagreeing, but I'd hesitate to conclude such a thing from this statement. 8 points of Mason's +/- come from the line on the floor while Nolan went on his run, and I don't recall Mason's play being especially big in that. Of course, I'd need to rewatch the game to check.

Remember, there are four other guys on the court, and if their play results in a run, the fifth player, who may have done nothing, gets credit as well.

1 game is a small sample size for conclusions using +/-

Kedsy
01-03-2011, 11:18 PM
1 game is a small sample size for conclusions using +/-

Oh, I agree. I just thought it was interesting, that's all, especially after Miles did so well in plus/minus early in the season while being roasted on the board.

Mason, by the way, was one of our top plus/minus performers against UNCG also, so it's two games rather than one (although still way too small a sample size for conclusions).

nocilla
01-04-2011, 09:05 AM
I'm not wholly disagreeing, but I'd hesitate to conclude such a thing from this statement. 8 points of Mason's +/- come from the line on the floor while Nolan went on his run, and I don't recall Mason's play being especially big in that. Of course, I'd need to rewatch the game to check.

Remember, there are four other guys on the court, and if their play results in a run, the fifth player, who may have done nothing, gets credit as well.

1 game is a small sample size for conclusions using +/-

Well Mason actually was a big part in Nolan's run. He set the screen for the drive/dunk and I believe set a screen for one or two of the 3pters. Not to say that Miles or Kelly wouldn't have made the same plays had Mason not been in the game because it was mostly Nolan's efforts, but maybe not.

I do agree with it being a small sample size though.

MChambers
01-04-2011, 09:48 AM
Oh, I agree. I just thought it was interesting, that's all, especially after Miles did so well in plus/minus early in the season while being roasted on the board.

Mason, by the way, was one of our top plus/minus performers against UNCG also, so it's two games rather than one (although still way too small a sample size for conclusions).
I'm glad you pointed this out. Folks seemed to think that Miles did very well and Mason not so well in the last two games, using the eye test, but the +/- numbers suggest the opposite.
Don't which is right, but it is interesting.

CDu
01-04-2011, 12:13 PM
I'm glad you pointed this out. Folks seemed to think that Miles did very well and Mason not so well in the last two games, using the eye test, but the +/- numbers suggest the opposite.
Don't which is right, but it is interesting.

And this is the challenge with +/-. It can be useful to identify contributions that fall outside of the standard individual metrics (points, rebounds, assists, etc). But it can also be heavily influenced by a ton of factors unrelated to the quality of the individual's play (performance of teammates, quality of the opposing lineup, luck of the bounce, etc).

Mason had some good moments (solid rebound numbers, some good screens), but he had some pretty bad moments as well (let Johnson blow by him for a baseline drive/dunk, some awkward-looking shots). It's unclear whether the recent +/- numbers are truly reflective of his performance or more reflective of those outside factors.

ACCBBallFan
01-04-2011, 12:20 PM
With Nolan playing all 40 minutes and Duke only alternating Seth/Dre at SG, I was expecting their +/- average out to +11 instread of +2 each?

Then I saw there was a three guard set that in this case did not work out too well, probably one of the times when Reggie Johnson took over while not on bench with foul trouble.

He is quite a load and good preparation for MD's Jordan Williams.

MChambers
01-04-2011, 12:25 PM
And this is the challenge with +/-. It can be useful to identify contributions that fall outside of the standard individual metrics (points, rebounds, assists, etc). But it can also be heavily influenced by a ton of factors unrelated to the quality of the individual's play (performance of teammates, quality of the opposing lineup, luck of the bounce, etc).

Mason had some good moments (solid rebound numbers, some good screens), but he had some pretty bad moments as well (let Johnson blow by him for a baseline drive/dunk, some awkward-looking shots). It's unclear whether the recent +/- numbers are truly reflective of his performance or more reflective of those outside factors.
Yes, I agree with that, but the eye test can be overly influenced by things too, such as spectacular plays around the ball (good or bad), and miss more subtle things, such as plays away from the ball (not to mention that TV limits what you can see). Some of the folks who post here are sufficiently knowledgeable that I trust their eye tests quite a bit, but others not so much (including me).

CDu
01-04-2011, 12:27 PM
With Nolan playing all 40 minutes and Duke only alternating Seth/Dre at SG, shouldn't their +/- average out to +11 instread of +2 each?

If Seth was in the game for that Nolan takover after the bogus charging call, I would guess it is his numbers that are too low, but have not done the math.

This would be true if Curry and Dawkins only played at the SG spot. The problem is that Dawkins also played some at the 3, and had a -7 at that spot. So while the SG +/- was +11, Dawkins "gave back" 7 of those points to the Plumlees and Kelly.

Edit: And it looks like you discovered that while I was typing this post!

loran16
01-04-2011, 12:34 PM
Yes, I agree with that, but the eye test can be overly influenced by things too, such as spectacular plays around the ball (good or bad), and miss more subtle things, such as plays away from the ball (not to mention that TV limits what you can see). Some of the folks who post here are sufficiently knowledgeable that I trust their eye tests quite a bit, but others not so much (including me).

Indeed. +/-'s best use really is proper capturing defense and other non-boxscore things a player brings to the table.

The problem is that it can get overswamped in small sample sizes, and that it doesn't on it's own separate a teammate's ability from the player who plays with him.

There's a way of getting partially around this problem long-term, called WOWY (with-or-without-you), where you look at how the team does with one player and without that player, with everything else being the same and compare. But to be truly accurate with it, you need minutes played of each lineup, and we haven't been charting that (it's a pain).


With Nolan playing all 40 minutes and Duke only alternating Seth/Dre at SG, I was expecting their +/- average out to +11 instread of +2 each?

Then I saw there was a three guard set that in this case did not work out too well, probably one of the times when Reggie Johnson took over while not on bench with foul trouble.

He is quite a load and good preparation for MD's Jordan Williams.

Errr, I don't think Johnson is a good simalcrum of Jordan Williams. When he's on the floor, Johnson may be better than Williams, in that he's a far stronger offensive rebounder (Williams is nothing special in that regard). Johnson is even a better scorer.

The difference is that Johnson commits 6.1 fouls per 40, and thus is on the bench a lot, meaning a team can manage to relax with him off the floor. Meanwhile, Williams only commits 2.8 fouls per 40, meaning he can basically stay in a whole game.

Also, Johnson can make free throws (71% shooter), while Williams....has issues (50.5%). You can afford to hack Williams. Johnson, less so.

They're both big men who are very good. Other than that they're different players who pose different challenges, and have different weaknesses.

ACCBBallFan
01-04-2011, 01:19 PM
Indeed. +/-'s best use really is proper capturing defense and other non-boxscore things a player brings to the table.

The problem is that it can get overswamped in small sample sizes, and that it doesn't on it's own separate a teammate's ability from the player who plays with him.

There's a way of getting partially around this problem long-term, called WOWY (with-or-without-you), where you look at how the team does with one player and without that player, with everything else being the same and compare. But to be truly accurate with it, you need minutes played of each lineup, and we haven't been charting that (it's a pain).



Errr, I don't think Johnson is a good simalcrum of Jordan Williams. When he's on the floor, Johnson may be better than Williams, in that he's a far stronger offensive rebounder (Williams is nothing special in that regard). Johnson is even a better scorer.

The difference is that Johnson commits 6.1 fouls per 40, and thus is on the bench a lot, meaning a team can manage to relax with him off the floor. Meanwhile, Williams only commits 2.8 fouls per 40, meaning he can basically stay in a whole game.

Also, Johnson can make free throws (71% shooter), while Williams....has issues (50.5%). You can afford to hack Williams. Johnson, less so.

They're both big men who are very good. Other than that they're different players who pose different challenges, and have different weaknesses.Because of the propensity to foul more, Jordan Williams is the better player of the two.

In the ACC fantasy league, he is the top guy in total of positives like points, rebounds, blocks, assists, steals minus turnovers, and Johnson is #11, the third best center with Tyler Zeller in between and Jerai Grant just behind.

30.90 Jordan Williams Maryland
27.50 Reggie Jackson Boston College
26.90 Nolan Smith Duke
26.80 Chris Singleton FSU
26.30 Mike Scott Virginia
25.90 Iman Shumpert Georgia Tech
25.60 Kyrie Irving Duke
24.80 Kyle Singler Duke
24.60 Jeff Allen Virginia Tech
24.00 Tyler Zeller UNC
23.90 Malcolm Delaney Virginia Tech
23.80 Reggie Johnson Miami
22.50 Jerai Grant Clemson

Mason is the fifth best fantasy points center in ACC, at least until Tracy Smith gets a few more games under his belt.

30.90 Jordan Williams Maryland
24.00 Tyler Zeller UNC
23.80 Reggie Johnson Miami
22.50 Jerai Grant Clemson
17.70 Mason Plumlee Duke
14.90 Bernard James FSU
13.90 Daniel Miller Georgia Tech
13.70 Tracy Smith NC State
13.30 Ty Walker Wake Forest
13.10 Victor Davila Virginia Tech
12.80 Josh Southern Boston College
12.10 Xavier Gibson (injured) FSU
11.60 Will Sherrill Virginia
11.40 Miles Plumlee Duke

Duke's scores individually look worse than when you add them together as center by committee.

ACCBBallFan
01-04-2011, 02:04 PM
Though Jordan Williams and Reggie Johnson have different styles and skills, what they share is being a couple of the jumbo ACC centers that only Miles has the bulk to content with.

So it becomes an issue of which forward they are paired with, and whether Kelly/a Plumlee can sag off the PF to help on the center.

Fantasy Weight ACC Center
Points Height ACC Team

23.80 , 303 6'10", Reggie Johnson, Miami

12.80 , 263 6' 10", Josh Southern, Boston College
04.20 , 260 7' 00", Jon Kreft, FSU
30.90 , 260 6' 10", Jordan Williams, Maryland

06.70 , 258 6' 09", Julian Gamble, Miami
13.90 , 258 6' 11", Daniel Miller, Georgia Tech
03.10 , 255 7' 02", Catalin Baciu, Clemson
13.70 , 255 6' 08", Tracy Smith, NC State
24.00 , 250 7' 00", Tyler Zeller, UNC

04.70 , 248 7' 01", Jordan Vandenberg, NC State
11.40 , 245 6'10", Miles Plumlee, Duke
13.10 , 245 6' 08", Victor Davila, Virginia Tech
04.60 , 242 6' 08", Cortney Dunn, Boston College
10.20 , 240 6' 09", Justin Knox, UNC
14.90 , 240 6' 10", Bernard James, FSU

06.10 , 239 7' 00", Assane Sene, Virginia
07.30 , 235 7' 00", Carson Desrosiers, Wake Forest
13.30 , 230 7' 00", Ty Walker, Wake Forest
12.10 , 230 6' 11", Xavier Gibson, FSU (injured)
17.70 , 230 6' 10", Mason Plumlee, Duke
22.50 , 230 6' 08", Jerai Grant, Clemson

04.60 , 229 6' 08", Raphael Akpejiori, Miami
11.60 , 226 6' 09", Will Sherrill, Virginia
03.10 , 222 6' 08", Will Regan, Virginia
04.50 , 218 6' 10", Nate Hicks, Georgia Tech

04.60 , 200 6' 10", Berend Weijs, Maryland

Saratoga2
01-04-2011, 02:09 PM
Funny, when everyone was ragging on Miles, he was routinely putting up great plus/minus numbers, and now people are snarking on Mason and his plus/minus line is 2nd on the team. Just goes to show that the eyes don't always tell the whole story.

Mason may have been on the floor when other players went off. He was active defensively, but only 2 points leaves any conclusions based on these plus and minus numbers suspect.

Notice also that Andre's numbers are very low. That doesn't make a lot of sense, since he shot very well and had 4 steals. Again, I find the plu and minus numbers very suspect.

-jk
01-04-2011, 02:47 PM
Mason may have been on the floor when other players went off. He was active defensively, but only 2 points leaves any conclusions based on these plus and minus numbers suspect.

Notice also that Andre's numbers are very low. That doesn't make a lot of sense, since he shot very well and had 4 steals. Again, I find the plu and minus numbers very suspect.

No regular stat records how many times a player rotates over to cut off a drive. A hedge to push the point guard well beyond the hash mark, forcing a reset. Set a really solid pick to open a teammate. The pass that sets up the assist. Blocking out the other team's best rebounder. Shutting down their best shooter. Plugging the passing lanes. None of those gets captured in stats we see, but are all important to team success. Because in the end it's a team sport - a team victory - not an accumulation of individual stats.

That's the whole idea of +/-, that individual stats simply cannot show a complete picture. It's not perfect (and fails on players with too few or too many minutes), but it paints a picture with broad strokes over time. Coaches watch tape and can see the discrete things a player does, and the coaches know what a player was told to do. We don't have that option. We have a box score and +/- and the "eye test".

Personally I think Mason is more suited to Kyrie as PG, and Miles more so to Nolan. Miles better understands moving under the basket in a slower paced game. Mason is more mobile and shows better in a faster paced game. I'm sure the coaches are working hard on Mason's footwork and positioning in the half-court, but those things take time and understanding.

-jk

loran16
01-04-2011, 03:25 PM
Because of the propensity to foul more, Jordan Williams is the better player of the two.

In the ACC fantasy league, he is the top guy in total of positives like points, rebounds, blocks, assists, steals minus turnovers, and Johnson is #11, the third best center with Tyler Zeller in between and Jerai Grant just behind.


Well yes. Mind you using, a fantasy league total thing is just silly. (Also, Johnson's advantage over Jordan is in OFFENSIVE REBOUNDS, not total rebounds, though he probably leads there too).

As I said, the strategy with dealing with either of them changes because of the foul thing and Jordan's FT shooting. Thus saying facing one is good preparation is probably a bad idea.

Of course, when Johnson is ON THE FLOOR, he's more dangerous than Williams. It just happens he's often not on the floor.

Newton_14
01-04-2011, 07:54 PM
I'm not wholly disagreeing, but I'd hesitate to conclude such a thing from this statement. 8 points of Mason's +/- come from the line on the floor while Nolan went on his run, and I don't recall Mason's play being especially big in that. Of course, I'd need to rewatch the game to check.
Remember, there are four other guys on the court, and if their play results in a run, the fifth player, who may have done nothing, gets credit as well.

1 game is a small sample size for conclusions using +/-

You should rewatch that stretch. Seriously.

Here is how it basically broke down.

Mason defensive rebound, Nolan jumper
Mason defensive rebound, Mason screen, Nolan 3
Miami made basket
Nolan 3 (assist by Ryan)
Nolan defensive rebound, Mason assist and screen, Nolan 3

Mason defensive rebound, Mason screen, Nolan down the lane for the dunk.

I think it is more than fair to give Mason some credit in that stretch.

loran16
01-04-2011, 11:07 PM
You should rewatch that stretch. Seriously.

Here is how it basically broke down.

Mason defensive rebound, Nolan jumper
Mason defensive rebound, Mason screen, Nolan 3
Miami made basket
Nolan 3 (assist by Ryan)
Nolan defensive rebound, Mason assist and screen, Nolan 3

Mason defensive rebound, Mason screen, Nolan down the lane for the dunk.

I think it is more than fair to give Mason some credit in that stretch.

I'm not saying he doesn't deserve some, I'm just saying to not take it too seriously.

For example, defensive rebounding is somewhat of a skill, but it's also one that's one of the most easily replaceable...it's very possible that Miles could've been in the game and gotten all of those boards. The screens too.

When I say I'd need to rewatch, I'd more need to see Mason's defense during the stretch and how he was playing.
----------------------------------------------------------------------
Anyhow, +/- really likes Mason because of that stretch. That's all i was pointing out in my caution...I'd say the same about anyone else on that line (Kelly or Curry for example) who wasn't directly involved with the scoring run (Nolan).

MarkD83
01-05-2011, 01:54 AM
---------------------------------------------------------------------
Anyhow, +/- really likes Mason because of that stretch. That's all i was pointing out in my caution...I'd say the same about anyone else on that line (Kelly or Curry for example) who wasn't directly involved with the scoring run (Nolan).

But Mason was directly involved in this scoring run as documented by the rebounds and screens he set.

I think the criticism/angst about Mason is that we expect him to score 20 points a game and in the new offense he is not doing that. One test to see if we are judging Mason's performance vs expectations is to substitute his name with Brian Zoubek's.

Zoubs defensive rebound, Nolan jumper
Zoubs defensive rebound, Zoubs screen, Nolan 3
Miami made basket
Nolan 3 (assist by Ryan)
Nolan defensive rebound, Zoubs assist and screen, Nolan 3

Zoubs defensive rebound, Zoubs screen, Nolan down the lane for the dunk.

In that scenario we probably would talk about how critical Zoubs was to this run by Nolan because our expectation last year was that his role was to get rebounds and set screens.

Basketball is a team game and when one player is hot the role of the team is to help the hot player and that was what Mason did.

loran16
01-05-2011, 02:11 AM
But Mason was directly involved in this scoring run as documented by the rebounds and screens he set.

I think the criticism/angst about Mason is that we expect him to score 20 points a game and in the new offense he is not doing that. One test to see if we are judging Mason's performance vs expectations is to substitute his name with Brian Zoubek's.

Zoubs defensive rebound, Nolan jumper
Zoubs defensive rebound, Zoubs screen, Nolan 3
Miami made basket
Nolan 3 (assist by Ryan)
Nolan defensive rebound, Zoubs assist and screen, Nolan 3

Zoubs defensive rebound, Zoubs screen, Nolan down the lane for the dunk.

In that scenario we probably would talk about how critical Zoubs was to this run by Nolan because our expectation last year was that his role was to get rebounds and set screens.

Basketball is a team game and when one player is hot the role of the team is to help the hot player and that was what Mason did.

Perhaps I'm not making my point properly.

The best way to use +/-, and a way that's useless with a single game sample size, is to look at a lineup of players with a player (in this case Mason Plumlee) and then look at a the same 4 other players with a substitute for Plumlee. You can do the same thing with Smith or anyone else. Of course, this doesn't work for small sample sizes.

When you don't have a large sample size of multiple games, instead I hesitate to use +/- as a great indicator of any result UNLESS a player's +/- is positive clearly from his use on multiple lines.
For example, in this game, Ryan Kelly was on 4 lines that outscored Miami, and ZERO that were outscored. There was one line that had a good performance that didn't have Ryan Kelly, for sure, but Kelly was the only player who wasn't on a line that was outscored by the opponent. Thus I'm more inclined to believe that Kelly's +/- result is clearly a real product of his great play, rather than simply being a lucky part of a lineup with 4 hot players.

That isn't the case with any other player on the team in this game, and thus I'm very hesitant to make any conclusions about the other players strictly from their +/-.

Kedsy
01-05-2011, 09:39 AM
I'm very hesitant to make any conclusions about the other players strictly from their +/-.

I agree with this. Frankly, I wouldn't use a one-game sample to make a conclusion about Ryan Kelly, either (although I do think he played well).

However, I'm less hesitant to use one or two game samples to dispel conclusions about a player. In other words, if someone says "player X was terrible" and then I see player X had the best plus/minus on the team, I'm willing to say perhaps he wasn't so terrible after all.

And obviously that's what I think about Mason's play in the past two games.

Neals384
01-05-2011, 10:07 AM
Not a good game to go small:

2 guards, Singler, 2 bigs +18
3 guards, Singler, 1 big -7