'Possession Evaluation' (PE) is a method of rating a team's performance in three areas of the game: offense, defense, and rebounding. The formula I employ was developed by Dean Smith and detailed in his book Basketball--Multiple Offense and Defense (Prentice Hall, Englewood Cliffs, NJ, 1981).

Offense/Defense: PE begins by counting each team's total possessions during the game. Offensive efficiency is represented by points per possessions (pts/poss or ppp). Defensive effectiveness is the opponent's pts/poss. A superior team should exceed .85 points per possession on offense and keep the opponents below .75 ppp.

"Percent Loss of Ball" (% LoB) is the percentage of possessions that end before a shot (FG or FT) is attempted. This is basically just a way to correct for the overall pace of the game when considering turnovers. %LoB can be used to indicate a team's own ball-handling skills, as well as the effectiveness of their pressure defense. A %LoB of .13 is excellent; .15 to .18 is good. You want to force you opponent into the .22 range.

Rebounding: The differential in the teams' total possessions is an indicator of rebounding. Coach Smith considers a FG attempt as an end to the team's possession; the ball is no longer in their control. If the shot misses, a possession is literally up-for-grabs. The better rebounding team should have more possessions by the end of the game.

Simply comparing total rebounds is inadequate, because rebounding figures depend greatly upon shooting accuracy. Take, for example, the 1995 game against Santa Clara.

KU's offensive- and defensive-efficiency numbers (.920 and .789, respectively) are good. What kept SC in the game was KU's terrible %LoB and poor rebounding. Even though the Jayhawks gathered two more rebounds, they should have had more.

Against a poor-shooting opponent (as SC's .353), a team should get plenty of defensive rebounds. And if the team itself is shooting well (KU's .545), the opponents will likely gather few defensive rebounds. Coach Smith recalls a game in which UNC had 59 rebounds, to the opponent's 45. Yet UNC had 11 fewer possessions, and thus were out-rebounded. The next day in practice, UNC worked on rebounding.

Total Possessions are calculated by counting the three events which end a possession:

• FG attempts. While the ball is in the air neither team controls it. A rebound of a missed shot (even an offensive-rebound) begins another possession.
• FT opportunities. This counts the entire series of FTAs, not each FTA separately. When all the FTs have been shot, the ball is given to the other team. (The only exceptions are technical fouls or intentional fouls, but these are relatively infrequent.) Only 'stand-alone' opportunities are counted here: when a player goes to the line for the 1+1 or double-bonus, or when a player is fouled on a missed shot (which means officially no FG attempt). FT attempts following a made basket are not counted, as they are really part of the same possession that produced the shot. The FG ended that particular possession.
• Turnovers.

Bob Bellotti (see Sources) has a different formula for arriving at possessions. It differs in that he counts the stats which begin a possession: opponent's FGs, opponent's turnovers, and rebounds. Instead of trips to the foul line, his system counts opponent's free-throws (divided by 2). It also includes a team's blocks (also divided by 2). It was designed for the NBA game (which has no 1+1), and I didn't think it was well suited for the college game.

The Bellotti system has two advantages, however. First, it can be calculated from boxscores or season-ending stats. One needn't laboriously count trips to the foul line. Second, Bellotti has developed a formula which, using the team's points per possession, can evaluate players' overall value to the team. It incorporates points, rebounds, assists, turnovers, blocks, etc. I'm still running players through these formulae to see how much insight they provide. Perhaps I'll make a record-list if the results are worthwhile. (I read a story in which Roy once demonstrated to Steve Woodberry and Richard Scott the per-minute, overall productivity of Greg Ostertag. It seems a logical assumption that at least points, rebounds, and blocks were factors in that equation.)

I went with Coach Smith's system in the end, at least for evaluating team performance. Obviously, his KU connection carries a lot of weight here. Furthermore, Coach Williams has indicated that his own system is very close to the one used here. In discussing the home victory over Iowa State in the 1997-98 season, Coach Williams stated KU "ended up with .84 points per possession, and we like to have ours .85 or better. We like to have (opponent's) .75 or worse, and they wound up at .68." My calculations yielded the values of .83 and .69, respectively.