"No matter how good you are, you're going to lose one-third of your games. No matter how bad you are, you're going to win one-third of your games. It's the other third that makes the difference." - Tommy Lasorda
I'm hoping that this is the first basketball article to begin with the wisdom of the former Dodgers manager. Lasorda was referring to the parity in the typical Major League Baseball season, but he unknowingly was describing the Atlantic Coast Conference in 2007. Every team won, and every team lost, at least a quarter of their conference games, something that occurred in only one other conference last season, the Ohio Valley. This indicates a conference in which any team can win on any given night, and indeed the regular-season co-champs, Virginia, did lose to the last-place team, Wake Forest, on the last day of the season with the stakes of claiming the outright ACC regular-season title on the line.
Much of the apparent parity in the ACC was an illusion, however, a mirage created by the schedule and an unusual distribution of close game results. Let's examine the issue of close games by looking at the point differential and Pythagorean record of the ACC teams during conference play.
Team PF-PA Diff Pyth W-L Real W-L
North Carolina 1350-1146 +204 13-3 11-5
Maryland 1235-1169 +66 10-6 10-6
Duke 1134-1085 +49 10-6 8-8
Virginia 1159-1133 +26 9-7 11-5
Boston College 1148-1128 +20 9-7 10-6
Georgia Tech 1165-1151 +14 8-8 8-8
Virginia Tech 1152-1142 +10 8-8 10-6
Florida St. 1175-1167 +8 8-8 7-9
Clemson 1108-1132 -25 7-9 7-9
N.C. State 1080-1191 -111 4-12 5-11
Miami 1125-1249 -124 4-12 4-12
Wake Forest 1090-1228 -138 4-12 5-11
The Pythagorean standings bear little resemblance to the real standings, especially at the top. If the ACC schedule were completely balanced, and you were willing to bow at the altar of point differential, then you have to acknowledge that the Tar Heels were significantly better than the rest of the league, Duke was much better than their record, and teams from Virginia were not as good as the real standings indicated. Performance in close games explains this. In games decided by five points or less, or games that went to overtime, UNC and Duke were a combined 3-6 while the Hokies and Cavaliers went 9-2.
Then there's the issue of the unbalanced scheduling. The ACC refuses to split men's basketball into two divisions, which would at least serve to give each six-team division a common schedule. Instead, each team plays the other 11 once, leaving five remaining games to be determined by a mix of rivalries and randomness. Using each team's Pythagorean win percentage, we can see how each team's slate of extra games stacked up.
Team Opp Pyth W-L Pyth W-L
North Carolina 35-45 (.433) 13-3 (.837)
Maryland 38-42 (.480) 10-6 (.634)
Duke 48-32 (.598) 10-6 (.609)
Virginia 31-49 (.384) 9-7 (.556)
Boston College 38-42 (.470) 9-7 (.544)
Georgia Tech 42-38 (.528) 8-8 (.530)
Virginia Tech 40-40 (.495) 8-8 (.522)
Florida St. 39-41 (.483) 8-8 (.517)
Clemson 45-35 (.566) 7-9 (.447)
N.C. State 44-36 (.555) 4-12 (.273)
Miami 38-42 (.474) 4-12 (.260)
Wake Forest 39-41 (.493) 4-12 (.233)
Note that the SOS covers just the five additional games on the ACC schedule, so the potential impact of schedule is smaller than that of the close-game performance. The difference between UVa's schedule and Clemson's schedule would make a difference of about a game in the standings on average.
This provides a way to recalibrate the true strength of the conference's teams from last season. UNC was clearly the best team, followed by Duke and Maryland, who had a slight advantage over the midsection of the league, where six teams were nearly indistinguishable. The bottom three are really the bottom two, with NC State having performed better than Miami and Wake during the regular season.
If the rosters remained stable, we'd expect the order of the teams to remain the same in 2008. Of course, that doesn't happen in this game. UNC's Brandan Wright and Duke's Josh McRoberts jumped to the NBA. Both are significant losses to their teams, especially on the defensive end. Maryland loses three senior starters from a season ago. Boston College, Georgia Tech and Florida State all had players taken in the first round of the NBA draft, while Virginia and Virginia Tech each saw 1500-point scorers depart. So you can see why there's more optimism around the programs at N.C. State and Clemson than there has been in a while. Those schools return the core of their 2007 squads in a year when the conference could be primed for a redistribution of wealth.
Don't get too carried away with the notion of a new-look ACC. It's not out of the question that UNC and Duke will be at the top conference again. It was easy to kick the Blue Devils while they were down last season, but their struggles were deceiving and they bring in one of the best freshman classes in the nation this season. Wright will be tough to replace at UNC, but the Tar Heels return every other player on last season's talented roster. Therefore, it's hard to imagine them having to share the regular season title again.
2007 Pythag % Returning 2008
Team Wins Minutes Prediction
North Carolina 13.2 70.4 13-3
Duke 9.7 82.4 11-5
Clemson 7.2 81.8 10-6
Maryland 10.0 48.1 9-7
N.C. State 4.5 81.4 9-7
Georgia Tech 8.5 60.7 9-7
Virginia 8.9 71.9 8-8
Florida St. 8.3 74.6 7-9
Miami FL 4.3 68.9 7-9
Boston College 8.7 50.0 6-10
Virginia Tech 8.3 39.2 4-12
Wake Forest 3.9 70.0 3-13
Ken Pomeroy is an author of Basketball Prospectus.
You can contact Ken by clicking here or click here to see Ken's other articles.