Cumulative Plus/Minus

[Jumbo]

Now updated through the Florida State game...

[Saratoga2]

Now updated through the Florida State game...

The numbers have become more meaningful as the number of games played have gone up. Individual play and performance during the season has been very interesting. Scheyer had a magnificent game last night. I belive the game before it was Nelson with his own magnificent game. We have also seen Singler, Henderson and Paulus have really excellent games and Taylor King has racked up points/minute like none of the others.

I wonder if the totals for the players really obscure the great game one night, so so game the next. JJ was a special player and more often than not, he had an excellent game without the variability we see on this team. Is the defense what is causing the variability of our individual performances, or is it some combination of defense, illness, crowd impact or other reasons?

[ArtVandelay]

Hey Jumbo, do you know if anyone tracks possessions per game stats and offensive/defensive efficiency on a per game basis? I know that KenPom tracks aggregate team stats (and has us averaging about 74 possessions per game, 22nd in the country), but it doesn't seem like he tracks this stuff from game to game. It would be interesting to see how our tempo varies based on opponent, and how our efficiency lines up with my subjective perceptions of how the team is playing. I know Paul over at Blue Devil Hoops offers HD boxes, but these seem only to track individual tempo-free stats, not team.

Also, ever heard of anyone tracking overall team data historically (I mean, aside from KenPom)? His stuff goes back to '04 - I'm guessing maybe nobody even thought to keep track and/or evaluate this stuff before that? It would be interesting to see tempo and efficiency stats for some of the kickass teams, i.e 1999, 2001, 2002, etc. Ah, the effect of sabremetrics on basketball...

Thanks.

[mus074]

you just gotta look for it:

http://kenpom.com/expsked.php?team=Duke&y=2008

As for historical info, let me know if you find it anywhere. This stuff is still a pretty new way of thinking about the game. How many times do we still here Steve Lavin talk about the best offense or defense based on points per game? Please. Someone give that guy a coaching job. For the love of the game. Get him out of the studio.

[ArtVandelay]

Ah, fabulous. I sort of thought it might be there. Thanks.

The fact that the historical data goes back only to 04 makes me think that it is just b/c it's a new way of thinking about the game. And yes, agreed on Lavin. Sort of like baseball commentators talking about things like wins, RBIs, and even batting average. Sadly, the people that are smartest about analyzing the games are too busy typing on message boards...

[mus074]

Ah, fabulous. I sort of thought it might be there. Thanks.

The fact that the historical data goes back only to 04 makes me think that it is just b/c it's a new way of thinking about the game. And yes, agreed on Lavin. Sort of like baseball commentators talking about things like wins, RBIs, and even batting average. Sadly, the people that are smartest about analyzing the games are too busy typing on message boards...

I have to agree. You know what I would like to see? Individual stats on likelihood of making a shot after a make, a miss, a made 3, a made FT, etc., just like OBP when facing a 2-1 pitch.

How's the import/export biz going?

[ArtVandelay]

Well, I've decided to quit and become an architect. Have you seen the new addition to the Guggenheim?

I'm assuming you mean essentially the offensive outcomes based on prior defensive possession? That would provide some interesting insights, particularly with regard to the whole "our offense feeds off our defense" meme, or "the patented Duke run." Sadly, I suspect that's way too detailed for anyone to have compiled. The only reason sabremetricians can do what they do is because people get paid to sit and watch baseball games, entering pitch-by-pitch data. That must be unbelievably boring, if you ask me. Although no more boring than my job, at times, I suppose.

[mus074]

Actually, I meant the same player's last offensive play, looking at the theory of streaky shooters and the notion of keep shooting to get out of the funk. We do have those stats in the play-by-play records of games at espn.com, although again it is quite new.

[Jumbo]

Now updated through the Clemson game...

[johnb]

Jumbo--this is great, and I imagine it must take more than a few minutes. thanks.

[Jumbo]

Jumbo--this is great, and I imagine it must take more than a few minutes. thanks.

No problem. The tracking is remarkably easy. It doesn't distract from the game at all -- it's purely a matter of logging each score and noting the lineup when Duke subs. So, a partial game log might look like this:

Paulus-Nelson-Henderson-McClure-Singler
0-2
0-4
2-4
2-6
Paulus-Nelson-Henderson-Singler-Thomas
2-8
4-8
Smith-Scheyer-Nelson-Singler-Thomas
6-8
9-8
9-10
11-10
11-12
13-12

It takes a little while to add everything up after the game, but that's not a big deal either.

[Jumbo]

Now updated through the Virginia Tech game...

[Jumbo]

Now updated through the Maryland game. Should I start doing ACC-only stats?

[cspan37421]

Is there a link to how +/- is calculated? Thanks.

[Jumbo]

Is there a link to how +/- is calculated? Thanks.

It's just a measure of how many points the team scores vs. how many it allows when a given player is in the game. So, all you have to do is keep a running score (super-easy), note when Duke's lineup changes (also super-easy) and at the end of the game, add up the total score for each player (still easy). Then, Voila! -- plus/minus.

[Bryan]

It's just a measure of how many points the team scores vs. how many it allows when a given player is in the game. So, all you have to do is keep a running score (super-easy), note when Duke's lineup changes (also super-easy) and at the end of the game, add up the total score for each player (still easy). Then, Voila! -- plus/minus.

So, if there is a sub after the first FT when shooting two, who gets the +1 (or -1) if the shooter makes the second FT, the player on the floor when the foul was committed or the player who subbed in?

Also, do you give players + and -'s for technicals?

[Jumbo]

So, if there is a sub after the first FT when shooting two, who gets the +1 (or -1) if the shooter makes the second FT, the player on the floor when the foul was committed or the player who subbed in?

Also, do you give players + and -'s for technicals?

During FTs, the points are assigned to the guys who were on the floor when the foul occurred. Technicals are part of the game, so if points are scored off a technical FT, they count.

[Jumbo]

Now updated through the NC State game...

[Jumbo]

The stats are now updated through the Miami game. If you get a chance, take a look at the way the numbers are starting normalize. I'll try to post some ACC-only plus/minus stats if I get around to it.

[Jumbo]

Now updated through the UNC game...