Reply to thread

Message: [QUOTE="auror, post: 2953104, member: 1329"] You are correct that it is even more reliable then, but the data from last year makes KenPom a lot more reliable than pure W/L records or the metric alternatives (like NET) without it at this point. It's also a daily decay, so the "preseason juice" is already a lessened influence, so I may quibble with your "a lot of data", but that perhaps depends on your perspective. I believe at this point the pre-season weights about 3 games worth of impact, and we've played 10 actual games, so that puts the pre-season influence at 23% or so. If we have ~23% influence of rank 102 (it was 5.88 AdjEM), and combined we're at 79 (8.47 AdjEM), we can solve that 9.24 would be our AdjEM for this year's play alone, rank 70th. Of course, we must bear in mind that the pre-season projections make the system more accurate at this stage by the nature of their conservatism even on teams they have underestimated (by limitation of design) ([URL='https://kenpom.com/blog/preseason-ratings-why-weight/']see his blog about it[/URL]), so somewhere between 70-79 is likely our true rank. And the ranks for other teams are likely even more accurate since they didn't have as many complicated things to project as we did, so I would expect their ranks to be accurate to +/- 5 (and the + or - can be guessed by the trend of their rank to date from season start). Now, we (and other teams) can certainly [I]improve or regress[/I] from our play to this date. Nothing is written in stone. But the rankings are already pretty good at this point. [/QUOTE]

Verification: First name of men's bb coach