Pomeroy Power Rankings with change from Tuesday games:

77 Akron -10
115 Toledo +1
122 Ohio -1
141 Ball State -5
145 Buffalo -0-
148 Eastern Michigan +5
155 Central Michigan -4
159 Northern Illinois -2
162 Kent State +9
185 Western Michigan +7
218 Bowling Green -7
283 Miami -1

(02-21-2016 10:49 AM)LongtimeFan Wrote:  Pomeroy Power Rankings with change from Tuesday games:

77 Akron -10
115 Toledo +1
122 Ohio -1
141 Ball State -5
145 Buffalo -0-
148 Eastern Michigan +5
155 Central Michigan -4
159 Northern Illinois -2
162 Kent State +9
185 Western Michigan +7
218 Bowling Green -7
283 Miami -1

Maybe this distribution of scores (actually rankings) could be used to teach elementary stats, e.g., normal curve, with most scores (ranks) bunched in near the median, 152) and few very high scores (Akron) and low scores (MIA).

The MAC is very well balanced this year, few teams which are really good or really bad.

Very hard to take Pomeroy seriously. I can't imagine anyone who follows the MAC closely that would rank the teams similarly.

I dont think anything besides conference standings makes much sense. Even that plus minus golf standing thing doesnt have the teams seperated enough IMO. Sagarins are okay, but OOC schedule will destroy you there. There is room for discussion for some teams but these rankings are pretty far off I would say. It has the top team, and bottom 3 correct?

I feel part of the problem in these rankings is that they don't take into account how much a team has progressed or regressed throughout the season. It views a win in November the same as a win in March.

(02-22-2016 10:49 AM)axeme Wrote:  Very hard to take Pomeroy seriously. I can't imagine anyone who follows the MAC closely that would rank the teams similarly.

I would contend that they aren't really that inaccurate. You can quibble about the order from 2-8 but statistically the only teams that seems out of place is Toledo and Kent State. And I would say that both of them are only slightly off.

(02-22-2016 02:40 PM)OUVan Wrote:  

(02-22-2016 10:49 AM)axeme Wrote:  Very hard to take Pomeroy seriously. I can't imagine anyone who follows the MAC closely that would rank the teams similarly.

I would contend that they aren't really that inaccurate. You can quibble about the order from 2-8 but statistically the only teams that seems out of place is Toledo and Kent State. And I would say that both of them are only slightly off.

Yep, that was (would have been) my point.

Kent State was a real surprise, but this year in the MAC other than a really good team (e.g., Akron and say Ohio) or a real bad team (e.g., Miami) it is all shades of gray.

KSU in 9th place? A bottom third MAC team? I'd be interested in hearing how someone would explain that based on, you know, actual season long results, considering that they are tied for 2nd.

(02-22-2016 05:05 PM)axeme Wrote:  KSU in 9th place? A bottom third MAC team? I'd be interested in hearing how someone would explain that based on, you know, actual season long results, considering that they are tied for 2nd.

You know I'm a fan of the KenPom rankings, in fact the 77 spot the Zips are in is more realistic than the 40 they are ranked in RPI. The biggest flaw, IMO, with his system is how much emphasis is placed on road games.

For example, Kent got less credit than they deserved for beating Akron. A Kent team in the 170 range (which, admittedly, was too low to begin with) should have jumped up way more than 9 spots for beating a top 70 team. But since it was a home win, the jump was marginal, and Akron only fell 9 places. Since Kent wasn't viewed that highly, the drop should have been more for Akron, but the system doesn't really penalize road losses.

Another example, and I only know this being the case since I'm an Akron fan. How about this one. When Akron beat CMU at home a couple weeks ago 92-88 (in a game Akron led by 20 with about 4 minutes left), Akron fell from like 63 to 68. CMU jumped from like 135 to 128.

So, it's not a perfect system, but you can pick flaws in every ranking system. I also don't think the MAC rankings are that far off. Kent probably should be ranked higher. At the same time, how much higher? Kent's only wins of note this year are all in MAC play .... @BSU, Ohio, EMU, Ohio, NIU and Akron.

I'm guessing since every MAC team now has at least four losses, all the teams in Kent's tier can pick out similar good MAC wins, too.

Really, 2-8 is a crap shoot. Akron at 1, right now, is a no brainer since the Zips also have as many quality MAC wins as anybody, and they actually have four solid OOC wins (all away from home ... @Arkansas, @Marshall, vs. Iona and vs. Santa Barbara). And neither of the OOC losses were bad (@Villanova and @Green Bay).

WMU, BG and Miami are pulling up the rear. ... The other eight, just put it in a hat and see what comes out. With two weeks left, that could change (and I'm guessing the rankings will also).

(02-23-2016 03:12 AM)Wadszip Wrote:  

(02-22-2016 05:05 PM)axeme Wrote:  KSU in 9th place? A bottom third MAC team? I'd be interested in hearing how someone would explain that based on, you know, actual season long results, considering that they are tied for 2nd.

You know I'm a fan of the KenPom rankings, in fact the 77 spot the Zips are in is more realistic than the 40 they are ranked in RPI. The biggest flaw, IMO, with his system is how much emphasis is placed on road games.

For example, Kent got less credit than they deserved for beating Akron. A Kent team in the 170 range (which, admittedly, was too low to begin with) should have jumped up way more than 9 spots for beating a top 70 team. But since it was a home win, the jump was marginal, and Akron only fell 9 places. Since Kent wasn't viewed that highly, the drop should have been more for Akron, but the system doesn't really penalize road losses.

Another example, and I only know this being the case since I'm an Akron fan. How about this one. When Akron beat CMU at home a couple weeks ago 92-88 (in a game Akron led by 20 with about 4 minutes left), Akron fell from like 63 to 68. CMU jumped from like 135 to 128.

So, it's not a perfect system, but you can pick flaws in every ranking system. I also don't think the MAC rankings are that far off. Kent probably should be ranked higher. At the same time, how much higher? Kent's only wins of note this year are all in MAC play .... @BSU, Ohio, EMU, Ohio, NIU and Akron.

I'm guessing since every MAC team now has at least four losses, all the teams in Kent's tier can pick out similar good MAC wins, too.

Really, 2-8 is a crap shoot. Akron at 1, right now, is a no brainer since the Zips also have as many quality MAC wins as anybody, and they actually have four solid OOC wins (all away from home ... @Arkansas, @Marshall, vs. Iona and vs. Santa Barbara). And neither of the OOC losses were bad (@Villanova and @Green Bay).

WMU, BG and Miami are pulling up the rear. ... The other eight, just put it in a hat and see what comes out. With two weeks left, that could change (and I'm guessing the rankings will also).

Kent state only split with Ohio, didn't sweep us. Just saying!

Toledo won at Kent State and Buffalo which is helpful in real standings and fake :)

(02-22-2016 05:05 PM)axeme Wrote:  KSU in 9th place? A bottom third MAC team? I'd be interested in hearing how someone would explain that based on, you know, actual season long results, considering that they are tied for 2nd.

This takes the whole season into account, not just the conference season and some of Kent's early season wins that looked good at the time don't look so good anymore. You want to put more emphasis on the last 14 games and ignore the first 14. Probably wouldn't be too happy if the emphasis was on the last 7. But mostly it's because statistically there isn't much difference between the top 9 MAC teams.