Tentatively there is now an actual algorithm that will be used when deciding the player-of-the-week. So far it seems to hold up well and has been checked against most short term variables such as player position, wins/losses, blow-outs/close victories, and tempo of offense; it was checked against SFU’s prolific offense, as well as UW’s display of defense against SDSU, and then a middle-of-the-pack starting line-up, and various other performances of the week.
The good news is it also makes our conference look fairly respectable -it hasn’t been tested against the line-ups of any of the current top 5 teams, but it should give players in context, rather than the league itself because it negates a lot. The idea is essentially hockey weighted basketball stats and thus player effectiveness rating.
Also, here’s where there’s some crow-eating about Riley Carel: he was the most effective guy on his team during the UAF game, BUT compared to other effective guys that night he was middle-of-the-pack. He can play better and he knows it.
It’s been tested quite well for how brief of a time it’s actually been played with (roughly five hours), but here are some major questions with answers that should be delivered over the next couple of months:
- Trends -will the same guys win all the time and does it matter if they do?
- Does it take into account stylistic differences among teams?
- What are the long-term lurking variables and can they be nullified?
- Do teams that go deeper inevitably suffer because things are more spread out?
With the first question: that happens anyway -guys are called starters and/or play major minutes for a reason. The second, so far so good, but again: long term trends. Obviously some of the lurking variables are known, but it’s the more esoteric that there aren’t answers for yet. Teams that go deep still have guys that are scoring most of the points to get them that deep; this was checked against a prep team that played a majority of blowouts (90% of regular season games).
Player-of-the-week will be posted soon.