I've mentioned elsewhere that we were six guys, a couple of which left, only to be quickly replaced by others. We each drafted a team of 23 players: 14 forwards, 7 defensemen and 2 goalies. Before each week, all the guys told me which three players
they would bench for the coming week, so that they had a roster of 12 forwards, 6 defenders and two goaltenders for the week's games. They also had to divide the skaters into four different lines and three defensive pairs, while they also had to designate
their starting goaltender. In cases where people didn't update their information, I just used their current line-up for the next week as well.
All six rosters were put into individual matrices, the stats for the whole team were calculated at the bottom,
and then compared to the other teams. The top team in each statistical category got 2 points in that particular stat, the worst team got 0, and all the others got 1 each. In case of a tie (and there were LOTS of those), the top lines and the first defensive
pairings for the teams in question were compared (of course the starting goalies were compared for the goalie stats). It may seem that this allocation of points was strangely uniform, but tests suggested that any other allocation (I tried 2, 2, 1, 1, 0 and
0) got weird or strange results more oten than not (lots of shutouts, for instance). I'll give you an example later, where I construct six teams, that for convenience sake will be called Bos, Chi, Det, Mtl, NYR and Tor. Maybe someone will help me out by "drafting"
six teams before the start of the season in a bit more than 25 hours? If so, I can use the first week of the season as my example week.