(01-31-2017 12:50 AM)MaddDawgz02 Wrote: Also this notion that no one is close to Akron is not supported by Kenpom either,also note that Akron has been dropping in these rankings even though they keep winning. That is something you might want to email Kenpom about to complain.
99 Akron
101 Ohio
112 EMU
I still think KenPom is the best data-driven ranking system, but it's hardly one that is flawless.
Let's just look after tonight:
No. 99 Akron beats No. 189 NIU on the road by four. Akron gains two spots.
No. 101 Ohio loses to No. 222 WMU on the road by five. Ohio drops four spots.
No. 112 EMU loses to No. 157 Kent State at home by 6. EMU drops five spots.
That really doesn't make sense, especially since Akron actually dropped 5-6 spots by beating that same WMU team on the road just a week ago by 10 that the Bobcats lost to. But the teams dropped essentially the same number of spots? That is despite Akron winning by 10 and Ohio losing by 5. And Akron scored more points and gave up less against WMU than Ohio did.
Back to UofToledofans' point, seems like winning is secondary to the analytic stats. That's something, even as a guy who thinks there is a value in analytic stats, I have a problem with. While I think sites like KenPom (and even Maddawgz' rankings if you were to tweak it a bit) do add value, there comes a point where you can't look mainly at pure offensive/defensive/SOS ratings, and ignore wins and losses, to get a true picture. If that was the case, why even play individual games?
With all that said, I do think there is a lot of things you can take from pure analytical rankings, just think it's a flaw to make that 100 percent of your argument on why a team is ranked where it is.