Saturday, December 29

UNLV odds & McNeese St stats


First the important stuff, odds for today's game:

UNC 81.11508, UNLV 78.49534
Pace 80.86226
HomeWinP 0.5932798
Check out that pace! Could be a real barn burner, which I feel would only help Carolina. More possessions against a poorly set up defense can only be a good thing. It'll also be interesting to see who draws Bennett on the defensive end. McAdoo might be the most similar player but he typically reserves his effort for offense. Could it be a task for Bullock, who is a hair shorter but is frequently lauded for his defense? I don't know if Joel James is quick enough or experienced enough, time will certainly tell.
Now the results from McNeese St (my apologies, the laptop didn't make it home to Charlotte) arguably the Tar Heels' best performance of the season (at least measuring as compared to KenPom predictions). It also came in a home game, hopefully the shooters continue to see a big basket today. Most notably, and I can't emphasize this enough, the Carolina offense is more inextricably linked to Field Goal shooting than ever this year. The performance against McNeese State did nothing to dissuade me of that notion. Yes, all teams are beholden to their shooting percentage but with more 3 point attempts and less ability to rebound and play defense this Tar Heel team is particularly at the mercy of their shooters (much to Roy's chagrin "If you want to give me something, that means that you think it’s to your advantage. I want to take what I want.")
One of the most encouraging signatures from the graph is the location of the point guards. Both Strickland and Paige used fewer possessions and had higher ratings than their season averages. I didn't watch this one, but it would indicate to me that they didn't turn the ball over, or take too many shots, instead relying on McAdoo and thewing players to generate offense. Assuming I'm correct this would indicate that the offense was operating well with fewer broken plays and forced shots by those two players. To use a sports metaphor, imagine a QB like Cam Newton able to find many different receivers and less reliant on scrambling out of trouble.
Jackson Simmons and Joel James both broke the possession meter. Each had more offensive rebounds than shots attempted and thus created possessions for the Heels and had "negative" possession used numbers.Also, PJ Hairston's name is really really big, which can happen when a blowout limits his playing time. He used 39 percent of the team's possessions while on the floor, a comfortable first place over McAdoo at 29%. 
$`Four Factors`
        OPP      NAME ORTG perEFG perORB  FTR TORATE
1 McNeeseSt     Total 1.28   0.56   0.51 0.30   0.18
2 McNeeseSt oppTotals 0.83   0.38   0.31 0.26   0.13

$`Last Game`
           POSS  ORTG   USG perFTM perFGM perFGM.3 perORB perDRB perTS  FTR perEFG ARATE
Bullock      11  1.55  0.24    NaN   0.60     0.62   0.04   0.13  0.85 0.00   0.85  0.19
McAdoo       13  0.92  0.29   0.33   0.45      NaN   0.08   0.16  0.44 0.55   0.45  0.14
Hubert        3  0.67  0.16    NaN   1.00      NaN   0.18   0.23  1.00 0.00   1.00  0.00
Strickland    8  1.00  0.16    NaN   0.67     0.00   0.00   0.09  0.67 0.00   0.67  0.26
Paige         6  1.00  0.13    NaN   0.33     1.00   0.00   0.00  0.50 0.00   0.50  0.41
James         0  -Inf  0.00   1.00   1.00      NaN   0.18   0.23  1.06 2.00   1.00  0.00
Hairston     12  1.67  0.39   0.88   0.45     0.38   0.24   0.19  0.69 0.73   0.59  0.14
Johnson       7  1.43  0.26    NaN   0.50      NaN   0.20   0.22  0.50 0.00   0.50  0.00
McDonald     10  1.10  0.25   0.67   0.33     0.43   0.00   0.04  0.53 0.33   0.50  0.16
Tokoto        3  1.00  0.11   0.33   0.50      NaN   0.07   0.11  0.45 1.50   0.50  0.16
Davis         0   NaN  0.00    NaN    NaN      NaN   0.00   0.00   NaN  NaN    NaN  0.00
Simmons      -1 -4.00 -0.09    NaN   0.67      NaN   0.67   0.12  0.67 0.00   0.67  0.00
Total        76  1.28  1.00   0.64   0.47     0.46   0.51   0.67  0.58 0.30   0.56  0.83
oppTotals    76  0.83  1.00   0.42   0.36     0.18   0.31   0.49  0.39 0.26   0.38  0.46

$`Season Totals`
         NAME POSS ORTG  USG perFTM perFGM perFGM.3 perORB perDRB perTS  FTR perEFG ARATE
1     Bullock  114 1.38 0.19   0.82   0.49     0.48   0.07   0.13  0.64 0.15   0.62  0.15
2      McAdoo  183 0.98 0.28   0.62   0.47      NaN   0.10   0.17  0.50 0.38   0.47  0.06
3      Hubert   13 0.92 0.06   0.00   0.50      NaN   0.09   0.10  0.40 0.58   0.50  0.02
4  Strickland  117 0.91 0.20   0.59   0.46     0.24   0.02   0.08  0.51 0.38   0.48  0.23
5       Paige  112 0.69 0.21   0.80   0.36     0.32   0.01   0.06  0.45 0.06   0.43  0.22
6       James   31 1.39 0.09   0.64   0.53      NaN   0.10   0.16  0.56 0.44   0.53  0.04
7    Hairston  118 1.14 0.32   0.88   0.37     0.33   0.11   0.13  0.54 0.31   0.47  0.10
8     Johnson   79 1.32 0.25   0.60   0.62      NaN   0.08   0.26  0.62 0.13   0.62  0.03
9    McDonald  102 1.11 0.24   0.69   0.41     0.45   0.05   0.07  0.55 0.17   0.54  0.12
10     Tokoto   47 1.13 0.18   0.33   0.55     0.14   0.11   0.15  0.53 0.43   0.56  0.08
11      Davis    6 1.33 0.05   1.00   1.00     1.00   0.00   0.03  1.39 1.00   1.50  0.40
12    Simmons    7 2.29 0.07   1.00   0.43      NaN   0.18   0.18  0.51 0.29   0.43  0.05
13      Total  924 1.09 1.00   0.65   0.45     0.37   0.40   0.67  0.53 0.26   0.51  0.63
14   oppTotal  924 0.90 1.00   0.68   0.38     0.33   0.31   0.61  0.47 0.26   0.44  0.50

Thursday, December 20

UNC @ Texas, Post Game Wrap


Well, no one can say KenPom didn't warn us. On one end the game was largely what I expected it to be given Texas' ratings. The Heels' offense has been hot and cold this year, centering largely around the 3 point shot without which it seems to flounder. And last night flounder it did, greatly. There was the all too familiar 4-6 point hump (similar to Butler, couldn't quite get close enough to lead), and if a few more 3's drop (team 3FG% was 16, normally it is 36) Carolina would have briefly seized the lead. As an exercise, just check out the game graph, every UNC player except Brice Johnson under-performed their season offense. Texas is a good defensive team, and those numbers are the evidence to back up Bullock no having open looks at 3 and McAdoo getting bullied out of the post.
The other end of the game was much more troubling. Texas isn't a good offensive team. Much like a teacher who knows which material is going to be on the test, I'm going to repeat that in case you missed it, TEXAS ISN'T A GOOD OFFENSIVE TEAM (98.7 adjusted, 174th after last night). Upcoming on the schedule is UNLV (45th), UVA (104th), Miami (25th), FSU (96th) and Maryland (72nd). These are all teams better than Texas, a bad offensive team that UNC refused to guard at times. To me this is as alarming a problem as the offense, defense is based on communication and effort and should "gel" a lot faster than the offense.
Now, if you've made it through the two doomsday paragraphs above I have some positive notes. In general these problems seem fixable. I wasn't super impressed with Texas' defense, they were good but the Heels gave the ball away multiple times, you remember the ball bouncing off McAdoo's hands, James' hands, through Bullock's leg, off Tokoto's hands out of bounds. Also, the issues overall are more of consistency than ability (which frequently makes it all the more frustrating). Take the 11 minutes spanning halftime, where UNC used a 30-15 run and closed a 39-20 deficit to 54-50, and did it only using 2 3-pointers (of 3 made in the game :-/).
Consistency problems have always plagued Roy Williams' teams, this isn't something that should come as a surprise to no one. Even in 2009 the team had ups and downs (Loss to Boston College anyone?), but no one notices when you're 15 to 20 points better than your opponents. When you're basically even to start with and you under-perform you lose basketball games. I don't think expecting 30-15 runs for 40 minutes is realistic, but somewhere in that 10 minute span exists the Carolina team that the media, coaches and fans went into the season expecting.
-Ryan
Cason, here's some +/- stats for you, that 30-15 run consisted of 10m51s here is the breakdown:
Bullock 10:51
McAdoo 9:01
Strickland 9:01
Paige 7:30
Tokoto 6:53
Hairston 5:19
McDonald 3:46
James 3:15
The name that jumps out to me is Tokoto, the others are all in proportion to their minutes played across the whole game, while Tokoto's minutes are elevated. Roy made the choice to start Tokoto in the second half and it payed off as the run continued.

$`Four Factors`
    OPP      NAME ORTG perEFG perORB  FTR TORATE
1 Texas     Total 0.84   0.34   0.39 0.48   0.22
2 Texas oppTotals 1.06   0.45   0.38 0.35   0.16

$`Last Game`
           POSS ORTG  USG perFTM perFGM perFGM.3 perORB perDRB perTS  FTR perEFG ARATE
Bullock      14 1.29 0.23   0.83   0.35     0.25   0.16   0.20  0.46 0.35   0.38  0.13
McAdoo       18 0.78 0.28   0.60   0.36      NaN   0.07   0.16  0.45 0.91   0.36  0.12
Hubert        1 0.00 0.06    NaN   0.00      NaN   0.10   0.00  0.00 0.00   0.00  0.00
Strickland   11 0.82 0.18   0.75   0.43     0.00   0.00   0.06  0.51 0.57   0.43  0.06
Paige        12 0.58 0.21   1.00   0.25     0.17   0.00   0.00  0.39 0.25   0.31  0.34
James         3 0.00 0.21    NaN    NaN      NaN   0.11   0.00   NaN  NaN    NaN  0.00
Hairston     12 0.75 0.27   1.00   0.17     0.14   0.14   0.08  0.33 0.33   0.21  0.09
Johnson       3 1.33 0.25    NaN   0.67      NaN   0.00   0.29  0.67 0.00   0.67  0.00
McDonald      6 0.33 0.17    NaN   0.14     0.00   0.04   0.10  0.14 0.00   0.14  0.00
Tokoto        4 1.00 0.11   0.33   1.00      NaN   0.04   0.23  0.55 6.00   1.00  0.00
Davis        NA   NA   NA     NA     NA       NA     NA     NA    NA   NA     NA    NA
Simmons      NA   NA   NA     NA     NA       NA     NA     NA    NA   NA     NA    NA
Total        80 0.84 1.00   0.69   0.31     0.16   0.39   0.57  0.41 0.48   0.34  0.52
oppTotals    80 1.06 1.00   0.73   0.41     0.35   0.38   0.55  0.50 0.35   0.45  0.40
$`Season Totals`
         NAME POSS ORTG  USG perFTM perFGM perFGM.3 perORB perDRB perTS  FTR perEFG ARATE
1     Bullock  103 1.36 0.18   0.82   0.48     0.45   0.07   0.13  0.62 0.16   0.59  0.15
2      McAdoo  170 0.99 0.28   0.65   0.47      NaN   0.10   0.18  0.51 0.37   0.47  0.05
3      Hubert   10 1.00 0.05   0.00   0.45      NaN   0.08   0.09  0.36 0.64   0.45  0.02
4  Strickland  109 0.90 0.20   0.59   0.45     0.25   0.03   0.08  0.50 0.41   0.47  0.22
5       Paige  106 0.67 0.22   0.80   0.36     0.28   0.01   0.07  0.44 0.06   0.43  0.20
6       James   31 1.26 0.10   0.58   0.52      NaN   0.10   0.16  0.54 0.39   0.52  0.04
7    Hairston  106 1.08 0.31   0.88   0.36     0.33   0.10   0.13  0.52 0.26   0.46  0.10
8     Johnson   72 1.31 0.25   0.60   0.64      NaN   0.07   0.27  0.64 0.14   0.64  0.03
9    McDonald   91 1.12 0.24   0.69   0.42     0.46   0.06   0.07  0.56 0.15   0.54  0.12
10     Tokoto   43 1.16 0.19   0.33   0.55     0.14   0.11   0.15  0.54 0.38   0.56  0.07
11      Davis    6 1.33 0.05   1.00   1.00     1.00   0.00   0.03  1.39 1.00   1.50  0.41
12    Simmons    8 1.50 0.08   1.00   0.36      NaN   0.12   0.19  0.47 0.36   0.36  0.05
13      Total  848 1.07 1.00   0.65   0.45     0.36   0.39   0.67  0.53 0.25   0.51  0.61
14   oppTotal  848 0.91 1.00   0.70   0.38     0.34   0.31   0.62  0.48 0.26   0.45  0.51

McNeese St prediction(107.4,90.2,75.7,97.5,104.2,65.3)
$HomePts
[1] 85.78433
$AwayPts
[1] 63.7419
$Pace
[1] 73.45037
$HomeWinP
[1] 0.9681834

Saturday, December 15

ECU @ UNC 121215


$`Four Factors`
  OPP     NAME ORTG perEFG perORB  FTR TORATE
1 ECU    Total 1.16   0.55   0.29 0.36   0.15
2 ECU oppTotal 1.09   0.49   0.24 0.40   0.12

wtf, these shapes are way to similar
































































$`Last Game`
           POSS ORTG  USG perFTM perFGM perFGM.3 perORB perDRB perTS  FTR perEFG ARATE
Bullock       7 2.00 0.13    NaN   0.67     0.67   0.17   0.11  0.78 0.00   0.78  0.18
McAdoo       20 0.95 0.31   0.90   0.36      NaN   0.00   0.21  0.52 0.71   0.36  0.00
Hubert        0  NaN 0.00    NaN    NaN      NaN   0.00   0.00   NaN  NaN    NaN  0.00
Strickland   11 1.09 0.17   0.75   0.67     0.50   0.00   0.12  0.77 0.67   0.75  0.37
Paige         9 0.67 0.18   0.50   0.33     0.50   0.00   0.08  0.44 0.33   0.42  0.19
James         4 0.50 0.25    NaN   0.25      NaN   0.00   0.12  0.25 0.00   0.25  0.00
Hairston     11 1.64 0.28   1.00   0.60     0.50   0.17   0.14  0.77 0.40   0.70  0.00
Johnson       5 1.00 0.17   0.50   0.50      NaN   0.00   0.25  0.51 0.50   0.50  0.00
McDonald     10 0.70 0.23   1.00   0.33     0.00   0.00   0.09  0.37 0.11   0.33  0.05
Tokoto        4 2.50 0.20   0.00   1.00      NaN   0.22   0.30  0.92 0.20   1.00  0.12
Davis         0  NaN 0.00    NaN    NaN      NaN   0.00   0.00   NaN  NaN    NaN  0.00
Simmons       0  NaN  NaN    NaN    NaN      NaN    NaN    NaN   NaN  NaN    NaN   NaN
Total        80 1.16 1.00   0.79   0.51     0.43   0.29   0.71  0.60 0.36   0.55  0.59
oppTotal     80 1.09 1.00   0.78   0.42     0.37   0.24   0.74  0.55 0.40   0.49  0.54

$`Season Totals`
         NAME POSS ORTG  USG perFTM perFGM perFGM.3 perORB perDRB perTS  FTR perEFG ARATE
1     Bullock   89 1.37 0.17   0.82   0.51     0.47   0.06   0.12  0.65 0.12   0.63  0.15
2      McAdoo  152 1.01 0.28   0.67   0.48      NaN   0.10   0.18  0.52 0.32   0.48  0.05
3      Hubert    9 1.11 0.05   0.00   0.50      NaN   0.08   0.10  0.38 0.70   0.50  0.02
4  Strickland   98 0.91 0.20   0.57   0.45     0.27   0.03   0.08  0.50 0.39   0.47  0.24
5       Paige   94 0.68 0.22   0.67   0.37     0.30   0.01   0.08  0.45 0.04   0.44  0.19
6       James   28 1.39 0.09   0.58   0.52      NaN   0.10   0.16  0.54 0.39   0.52  0.04
7    Hairston   94 1.12 0.31   0.86   0.39     0.35   0.09   0.13  0.54 0.25   0.49  0.10
8     Johnson   69 1.30 0.25   0.60   0.64      NaN   0.07   0.27  0.64 0.15   0.64  0.03
9    McDonald   85 1.18 0.25   0.69   0.44     0.47   0.06   0.07  0.59 0.16   0.58  0.12
10     Tokoto   39 1.18 0.20   0.33   0.54     0.14   0.13   0.13  0.54 0.23   0.55  0.09
11      Davis    6 1.33 0.05   1.00   1.00     1.00   0.00   0.03  1.39 1.00   1.50  0.40
12    Simmons    8 1.50 0.09   1.00   0.36      NaN   0.12   0.20  0.47 0.36   0.36  0.05
13      Total  767 1.10 1.00   0.64   0.47     0.38   0.39   0.68  0.54 0.23   0.52  0.61
14   oppTotal  767 0.90 1.00   0.70   0.38     0.34   0.30   0.63  0.48 0.25   0.45  0.52

Tuesday, September 18

Monday, February 27

UVA, Duke, and a short treatise on college basketball...

Ok, this might be a long one, I watched a lot of great basketball yesterday (and currently OSU-Wisconsin) so I'm primed and ready to go for March Madness. I'm going to do my typical recap of the UVA game first, which will likely lead into a larger commentary on college basketball (thanks to some awesome graphs from KenPom). That will then be followed by a Duke preview, because I'm travelling again Thursday-Saturday and won't have time to get one out later this week. Alright, time to follow me down the rabbit hole!

Yesterday the Heels seemed to follow an all too familiar formula when playing on the road. They started out good or even great, got a little lazy and trailed early in the second half, and finally put some solid defensive possessions together for the win. The mild difference was that I'm not 100% sure the "solid defensive possessions" were that and not just some luck. Some will say luck is the residue of design, though I'd disagree and say that UVA missed some fairly open shots. Then again if the Heels designs were to be ahead, that fact likely forced Mike Scott into an ill-advised 3 on one of the Hoos' last possessions. In brief summary, UNC played decently and shot poorly, luckily UVA played a little worse and shot poorer. Looking over the Four Factor numbers the Heels took care of the non-shooting aspects of the game, decent TO numbers, good rebounding, good number of trips to the line. Similarly, the Heels did the same on the defensive end of the floor, kept UVA off the glass and the free throw line (UNC eschews forcing turnovers as a defensive mechanism, if you don't agree with me or are unsure why I have a book recommendation for you). The Heels did all the little things right, but their shots didn't fall. Because they did those other things the Heels left Charlottesville still in control of their own ACC (and NCAA?) destiny.

Favorite Number: 10 FTR (Free Throw Rate, FTA/FGA) - This is key to the UNC defense, I'm sure you've heard me say it, but I'll hammer on it again. The height the Heels have, and the method of defense they play, relies on the ability to affect shots without fouling. In games the Heels have posted under 25 FTR (19 total this year), the opponent has only exceeded 1 point per possession once (Jan 29th, hosting GT) and the Heels are 19-0 in those games. Not to get ahead of myself but this will be a huge key to the Duke game.


Least Favorite Number: 35.3 eFG% - I'm not a fan of the number itself, but it is more how the Heels earned this metric that was upsetting. There were a number of poorly selected shots (largely 3's) and possessions where the ball never visited the paint. I don't mind a cold shooting night, but only when it is earned by the opposition through defense and not as a result of lazy or misguided offensive flow.

Of the Four Factors, the one that has the most bearing on efficiency is Field Goal percentage. This isn't a debatable fact, you can play a rotten all around game and be bailed out by good shooting. Alternatively you can have a cold night shooting and lose despite be on point everywhere else. That is what makes college basketball exciting and unpredictable, and what makes March worth experiencing. On Saturday, the Heels produced a poor shooting night of epic proportions, only making 17 field goals. This is arguably the bottom of the Heels' efficiency curve, they won't have many nights worse than this, and a lot of that can be attributed to random shooting noise, especially when a team like Virginia slows the pace down to 60 possessions. Fortunately the Wahoos had a poor shooting night of their own, and had an offensive output similarly on the bottom end of their spectrum. I've graphed (below) the offensive efficiency outputs of UNC and UVA over the course of the year to illustrate my point:

Inline image 1

As you can see, the UNC curve exists at higher efficiency values than the UVA one, meaning that the Heels' worst game is better than the Hoos' worst game. I felt that it would be safe to ignore defense here, since UNC and UVA have similarly skilled defenses. Though one could graph each curve with raw data first, and then account for each team's defense after the fact.* As an example, graphing Duke and UNC's curves would show Duke to the right of UNC, but would be ignoring the fact that UNC's defense is measurably better than the Devils' defense this year (yes, I just wrote that sentence, tongue far from cheek, crazy huh?).

Any college basketball game can be simplified to this degree, the point per possession output of one team graphed against the other, and will be fairly accurate as long as you take home court and defense into account. (You might notice that UVA is bi-modal, which complicates things, I'm assuming most teams have offensive outputs that fit a normal curve). This is a really neat way to think about basketball, and in a way provides insight into March Madness. When you look at a given matchup there are a pair of curves. For any two teams that make the NCAA tournament their curves overlap. The only ones that don't (using historical data) are the #1 and #16 teams, which is truly saying something. Of the 67 games that will be played in March (& April), 63 of them could go either way and the large majority of those will contain curves like the above, where the teams only differ by 5-10 points/100 possessions. The largest swings in the curves happen as a result of field goal percentage, something that can come and go in any given game, making March all the more mad. KenPom has talked a bit about this lately on his blog (see the link above), mainly within the context of three point shooting being a force that serves to increase entropy within the system of college basketball.

Now, are there exceptions to this rule? Of course. The 2009 Tar Heels weren't within 10 points of any of their 6 NCAA opponents. There is the occasional team and bracket breakdown that leads to an overwhelming champion, but this year doesn't appear to have an overwhelming favorite. If the chips fall the right way, Kentucky would be the one team to run the table, but they aren't worlds above the other teams in the bracket and they have won a few close games against a lackluster SEC.

So, since UNC falls into the category of "not overwhelming favorite" their chances will be bettered by a safer path to the Final Four. Depending upon how the remainder of the season plays out, UNC seems to be destined for between 3 and 10 on the NCAA tourney "S-curve" (where 1-4 are 1-seeds, 5-8 are 2-seeds, Lunardi currently guesses UNC is at 6 http://espn.go.com/blog/collegebasketballnation/post/_/id/50096/joe-lunardis-latest-bracketology-update-10). For simplicity I'm assuming a win against Maryland, but depending upon the Duke outcome and ACC tournament performance the Heels could end up as high as the 3rd or 4th #1 seed, but as low as a #3. That said, what matters more than any positioning between 4-7 is the ability to avoid the best teams and the teams that are underseeded. For instance, this year it'd be best to stay out of Kentucky's region, and away from a dangerous Buckeye team that seems destined for a 3-seed. Funny that given any small change in the last Duke-UNC game and the entire storyline of the season would be different (see here for the original publishing of this idea http://www.basketballprospectus.com/article.php?articleid=2087). As it stands, Carolina has finished the season strong and in position to set themselves up for a #1 seed regardless of Austin Rivers' heroics.

Next Saturday the ACC crown will be up for grabs (safe to say, because even if either team loses mid-week, the regular season title will still be in the balance). There are a couple of keys to the game in my mind, one that I've touched on already.

1) Free Throws - Duke cannot get to the line as often as they did in the Dean Dome. This may see like an impossibility, and it very well may be, but the point is that if Duke is only taking shots from inside the arc on half their possessions, UNC can't foul them on and additional half of those. Let Duke take contested 3's, or let Rivers drive, but don't let Duke beat you from behind the arc and the stripe.



2) Rebounding, Offensive - Duke isn't a horrible defensive team, just a bad team relative to Dukes of the past. One of the most stark contrasts in this game is UNC's ability to board its own misses (5th in the country/1st in the ACC) with Duke's inability to prevent it (162nd/9th). Even if the shots aren't dropping, the Heels must trust their own inside game and their ability to get multiple shot attempts in a given possession. Last game Duke had 10 offensive rebounds while UNC had 14, if the Heels are going to withstand a hot shooting night from Duke, they'll need to get 2 and 3 attempts off their own misses.

3) Additional Production - In the first game, UNC's key players played well, but no one else stepped up. Bullock and Hairston had 0-fers from behind the arc and McAdoo had 6 points. This was balanced well by the key scorers pouring in above average nights, but that can't be expected to happen. Bullock, Hairston, or even McAdoo getting hot would go a long way to a Carolina win.

The way that Duke and UNC are designed are in such a complementary fashion that each has the other's number. If you were to create an offense UNC would have difficulty with, it would be one with a potent slasher (Rivers) and several role players capable of hitting three's when the defense gets sloppy. Similarly, Duke has problems defending opponents in the post (they're 2nd in the country in opponent's 3PA/FGA, no one shoots 3's against them) and preventing opponents from getting second looks. UNC played an excellent game offensively on February 8th, but Duke's 14 3's were too much to overcome. I doubt Duke will hit 14 3's again (they hit 14 one other time this year, and took over 30 in one other game), but the Heels will have a difficult time reaching 84 this time around as well.


*There are actually even further assumptions that are being made to simplify the situation. One could even make further assumptions about how defenses and offenses interact. Overall UVA and FSU have very similar defensive efficiencies, but they each arrive at the numbers differently. FSU uses their height and length to pester opponents into turnovers, while UVA doesn't focus on them and is shorter overall. Thus the same defensive metric has grossly different effects on UNC. So, much like physics, college basketball theory works best in a vacuum