The computer age has bestowed some unforeseen blessings on the sports world. Math whizzes are now using the personal computer to track game results to a level never before possible, and sports are the better for it. The book "Moneyball" chronicled the way Billy Beane, general manager of the Oakland Athletics used obscure statistical analyses to scout and deal for talent. Now, a variation of this method has spilt over into basketball with a number of NBA teams using statistics in a similar manner. These methods were described in a recent Sports Illustrated article titled Measure of Success. Described as simply as possible, these statisticians track how well a team does when a player is on the floor, versus how well they do when they arent.

These measures disregard a players contribution as represented by their scoring, rebounding and assists totals, and simply asks, "Does the team do better or worse when this player is on the floor?" They also in many cases contradict the conventional wisdom concerning the relative worth of certain players. Jason Collins, a fifth-year center for the Nets of little renown gets ranked as the fourth best defensive center in the league. Conversely, high-scoring Michael Redd of the Milwaukee Bucks is such a disaster on defense that his teams tend to lose with him on the floor, and win while he is on the bench.

While amateur coaches are sure to find these analyses interesting, is there any way in which they can employ these metrics for their own use? Amateur coaches might be able to get team assistants to gather the data needed to maintain these stats, but there is a much better way these concepts can be applied.

Virtually all teams rely heavily on scrimmage play as part of their practice routine. By adopting what I call a Roster Round Robin format during scrimmage play, coaches can get a much clearer picture of how and when individual players perform better than others. The format simply requires that sides be set to 3v3, 4v4, or 5v5, depending on available players. Players are issued reversible jerseys, and score is kept for each side. Stages are set to 3 or 4 minutes, and at the end of each stage, a team score is recorded, with each player earning points for themselves on the basis of their team result. Two players are then directed to exchange sides by flipping their reversible jersey, and a new stage is played. This pattern is continued until all possible roster combinations have been used.

For a 3v3 contest, this would equate to 10 possible combinations, thus requiring a 30 to 40 minute game. Each player earns a plus/minus score across all stages. As everyone plays under all roster combinations and points can only be earned on the basis of team results, an individuals result reflects their team contribution across all stages. In order to maintain game continuity and ensure rapid roster rotations, the roster rotations are predetermined and printed on a grid used for scoring.

Some may notice that the one thing seemingly not taken into account in this format is position play. A roster rotation schedule that disregards position play will likely result in some oddly balanced sides think 5 guards versus 5 forwards and centers that wouldnt reflect anything resembling a real-game matchup. This difficulty can, however, be overcome by taking positions into account when devising the roster rotations. For a 4v4 game, with players restricted to either a center/forward or guard position, there would be 9 different roster combinations required. For a full-sided game, a center position could be added, but 18 stages would be required. With three minute stages, this would be a 54 minute scrimmage game. Coaches may wish to split this size of scrimmage across multiple practices.

Would the results from any particular scrimmage mean much? For one game, probably not, as we all know the ball can bounce funny for a time. But if this format were used on a regular basis, the results should reveal which players are contributing the most to their team. These results will either confirm or deny a coachs sense of who his best players are, but there are two even bigger benefits that can come from the use of this format. First, as players come to understand this as the new measure of their play, they will be more receptive to a coachs instruction on team play. As well, players will intuitively respond to the demands of the game, and adjust their play accordingly. Simply put, the ability to measure team play translates to a better ability to teach and learn it.

The second benefit may be even more important. A system that clearly and demonstrably measures a player on the basis of their team contribution fosters team chemistry better than any other. The talented offensive player who lets down on defense can be brought to the table when the impact of his lackluster defense can be shown. Players who are frustrated because they think they should be playing more can either makes their case in the round robin practice, or be shown theyre not there yet. Regardless of the case, team chemistry is advanced.

The only real barriers to this practice format are logistical. Running this practice format requires pre-numbered reversible jerseys, and scoring/rotation grid sheets.