The best passing team in college hoops last season was Boston College. Illinois was widely admired for their ability to pass the basketball, but they had nothing on BC. I’ll define the best passing team to be to be the team that can best make the difficult passes that produce points.

One of the stats I posted last season was assist percentage. It measured the percentage of a team’s field goals that were assisted on. A lot of the stats I posted during the year were meaningless, and assist percentage certainly could be placed in that group. It seemed that assist percentage was mentioned more by TV guys than it ever was in the past. It was all the rage when describing the crisp passing that Illinois often exhibited. But it turned out that across college basketball, assist percentage correlated poorly to an efficient offense.

Not that there wasn’t any relationship between the two, but assist percentage told us more about a team’s style than their substance. The other obvious thing was that this metric was somewhat dependent on how often a team made three-pointers.

That was evident in the Final Four. In the three games in St. Louis, there were 52 made three-point shots, 45 (87%) of them received assists. On 110 made two-point shots, 49 (45%) received assists. This provides some anecdotal evidence to supplement the circumstantial numerical evidence that so many three-pointers are assisted.

This is a good time to review our Statisticians’ Manual so we are all on the same page with respect to what as assist means…

A player is credited with an assist when the player makes, in the judgment of the statistician, the principal pass contributing directly to a field goal (or an awarded score of two or three points). Only one assist is to be credited on any field goal and only when the pass was a major part of the play. Such a pass should be either (a) a pass that finds a player free after he or she has maneuvered without the ball for a positional advantage, or (b) a pass that gives the receiving player a positional advantage he or she otherwise would not have had.

Philosophy. An assist should be more than a routine pass that just happens to be followed by a field goal. It should be a conscious effort to find the open player or to help a player work free. There should not be a limit on the number of dribbles by the receiver. It is not even necessary that the assist be given on the last pass. There is no restraint on the distance or type of shot made, for these are not the crucial factors in determining whether an assist should be credited. [NCAA Basketball Statisticians’ Manual, 2005]

If you muddled through all of that, you may have learned something. For me it was, "It is not even necessary that the assist be given on the last pass." I wonder how often that happens.

It is not too surprising that three-pointers tend to get more assists. After all, a portion of two-pointers are "un-assistable," the result of put-backs and tip-ins, while few threes are shot off of the dribble.

At the same time, it’s safe to say that a lot of three-point assists are not very skillful. Any player with a modicum of coordination is able to make a perimeter pass. Hence, the high assist rate observed. And really, the most important role in the passer-shooter combination in this case is the shooter. The ability to make the three is harder than setting the shooter up. That’s not to say there aren’t situations where it takes skill to find the open shooter. A skip pass and the drive-and-dish situation are two examples that come to mind where the passer deserves some credit. But those situations are a fraction of the assists awarded on three-point shots.

So getting back to Boston College, they were second in the nation in raw assist percentage, yet they were fifth-worst in three point production. That tells you enough – they probably got more assists on two-pointers than anybody else.  But maybe they were a good shooting team that didn’t score often on put-backs, so there were more two-point assist opportunities.

Not quite. They were fourth in the nation in offensive rebounding percentage. So chances are they had more than their fair share of un-assistable baskets. (Footnote: five, count ’em, five Big East teams ranked in the top seven nationally in offensive rebounding percentage.)

Another way to look at this is by using something I’ll call "assists over expected." Assume that the average team gets assists on 80% of its successful threes and 47% on its twos. For each team you can compute their expected assists:

.8 x successful threes + .47 x successful twos

Then compare this number to their actual assists. Boston College comes out tops nationally in this difference. Their assist percentage of 68.6 dwarfed their expected number of 52.4.

Since I’m on the subject of BC, it’s a nice time to point out that the Eagles should be a factor in the ACC race next season (assuming Craig Smith comes back), especially considering that seemingly half of the ACC is going pro or graduating.

They’ll have Smith and Jared Dudley as scorers, plus an adequate senior point guard in Louis Hinnant. The two guys that will see their minutes increase most are two role players that will fill in nicely around this group.

Rising junior Sean Marshall, the only shooter on the team, will get more minutes at the expense of departing slasher Jermaine Watson. Marshall raised his three point shooting from 29% as a freshman to 39% last season. Bump that up three or four more percentage points and Marshall could be the most accurate sniper in the league on a team that needs some perimeter production. And Marshall should get plenty of looks, with Dudley and Smith occupying the opposition’s attention. There will be some defense sacrificed with the loss of Watson, but that will be more than compensated by the other guy that picks up minutes…

The biggest reason BC will improve is that rising sophomore Sean Williams will finally get more time. Nate Doornekamp, a seven-footer who couldn’t shoot 40% from the field and didn’t average one blocked shot per game, inexplicably kept Williams out of the starting lineup all season. (Pomeroy’s Summer Challenge ’05: find a player that deserved his playing time less than Doornekamp.)

Williams was mainly a defensive force, averaging 5.4 blocks per 40 minutes. (By comparison, Shelden Williams, who led the majors in blocks per game, averaged 4.4 per 40. Then consider that BC played slightly slower than average.) Sean Williams mostly stayed out of the way offensively, but shot 65% when he did get involved. Williams was impressive on the offensive boards – he had 60% of Doornekamp’s minutes, yet had 17% more offensive rebounds. He’s certainly one guy to keep in mind for having a breakout sophomore season.

Duke will be the favorite to win the ACC, but don’t count out BC as a legitimate contender for second place, despite their late-season swoon in ’05. And expect at least one anticipated game involving opposing centers named Williams.

[Update: BC blogger Bill has a prediction of Williams’ block numbers next season.]