Want to know which Bucknell players provide the most value to the Bison when they are on the floor? Eric at Bucknell has a way of looking at the box score that goes beyond the usual categories like points, rebounds and assists.

Folks on the Bucknell message boards know Eric at Bucknell from posts. For those who don't visit those forums, a quick bit of background: Eric is not actually at Bucknell anymore. He is now a grad student at Penn State. A math wiz with too much time on his hands, he spends much of that free time breaking down the numbers to come up with interesting statistical looks at the game.

One of his projects this season has been an ongoing tally of Bucknell players' plus-minus numbers.

Basically, this is how it works: Points scored by and against Bucknell while each player is on the floor are calculated to determine each player's plus-minus ratio. That part works just like the traditional hockey stat.

Then Eric took those numbers further, calculating the minutes each player played and turning that into 40-minute averages.

It is far from a perfect system. Given the amount of zone and help defense Bucknell employs, the combination on the floor impacts the number of points scored when a player is in the game. The defensive numbers probably would have more validity if Bucknell played more straight man-to-man, though even then, a weak defender among the starting five, for example, would negatively impact the defensive numbers for the other four starters.

Eric is planning to do some calculations in the future that look at different player combinations.

A lesser flaw is the position-by=position breakdown for each player, since Eric is forced to make assumptions what spot a player is at based on the combination in the game at the time. The most glaring example of that is Abe Badmus, whom Eric has assumed has played the point whenever he is in the game, though in reality, there have been times when he has actually played the wing, especially in combination with John Griffin. For the most part the position designations are probably pretty close. We just think the totals for some players are likely more valid than the position breakdowns

Despite those flaws, Eric's calculations do offer some interesting insight into the Bison's performance and Pat Flannery's rotation decisions. For example, Justin Castleberry's plus-2.7 probably helps explain why he has taken a lot of the minutes that went to Rob Thomas (minus-7.5) earlier in the season.

We will update these again as Eric provides them, assuming we can figure out an easier way to translate his spreadsheets into Web-compatible content. If someone wants to do the computations for other teams (and can place them into a convenient html table instead of a spread sheet), send them along and we will post them, too.
  • Click here to view Eric's Plus-Minus Calculations

    Labels: