MBB: North Coast Athletic Conference

[Pat Coleman]

That's right - the limited n/c opportunities can affect the data early on, but by now it should all be washed through. Good teams will show themselves, bad teams will be exposed, no matter who they have 'touched'.

Same with football, by week 8 or so it normalizes itself.

And yes, it's possible that a 'mediocre' team could be rated highly, but most likely the 'mediocrity' is a perception. Still, using an algorithm based rating system that uses elements of offense / defense, home / road, SOS and MOV is the best way to determine how good team A from the NE is compared to team B in the west.

Take time to understand the data and the reasons behind what the data shows, just don't pooh-pooh it out of hand like out-of-work political pundits.

What do you think of the case I cited on the Pool C board last week, then? Not Week 8 but the end of the season, where 14-11, 10-10 St. John's was rated 30th two years ago? I understand St. Thomas won the title out of that league but six MIAC teams in the top 30 is more than excessive.
http://masseyratings.com/rate.php?lg=cb&yr=2011&sub=11620

[smedindy]

The MIAC was a meat-grinder and had a great non-conference record. The Johnnies took care of the teams they could and beat a team that beat Mankato. Yes, it's D-2 but that shows quality and distinction. Their only bad loss was Macalester.

What this also shows is a house effect of St. Thomas winning it all, since it's a post-season rating. I wonder where the Johnnies were before the Tommies began their run? Since everything is interlocked, the Tommies results help the MIAC, especially the teams like St. John's which had a very very close loss to St. Thomas earlier in the year. Post-season power rankings always show this effect, but in 'perception' others also will tend to lift up the conference of the winner, anyway. So I can't say it's unexpected to see St. John's this high with those data points coming from St. Thomas. Again, everything is inter-related. It's a universe of hoops.

I also tend to shy away from looking at the power rankings after I do my tourney projections. The NCAA tourney is a fickle-beast - the one and done aspect of it hurts and helps teams. However, a total picture of the entire season is a really good measure. I again wonder where St. John''s was before the MIAC tourney, or right after it?

This is also a result of the MIAC having limited non-conference games. Yet every measure would also show the same effect, including the 'official' rankings the NCAA uses. But winning these non-conference games, or staying in the neighborhood of very good teams, isn't an aberration. It shows quality and distinction, even if early in the year, even if they're .500 in conference. As it is in football, you take a mid-pack OAC or WIAC team and plop them in most every other conference, they win or compete. Who is to say that this St. John's team, plopped into most every other D-3 conference, couldn't have competed or won it? This was a big ol' slog that year.

If they're over-rated it may be by just a few points. Can you be so certain St. John's couldn't beat Emory six times out of 10 that year? Remember, it's SPORTS! Things happen!

This all reminds me of the early debates of perception vs. data when Bill James started his abstracts. The Enos Cabell debates, as it were. I guess the debates on Derek Jeter's defense at short in the 2000's were also the same, since the data flew in the face of perception, and the starry-eyed didn't see the flailing on the dives to balls a good shortstop would get. Because, you know, narrative!

Ken Pomeroy has Belmont, Middle Tennessee, SF Austin and ND State rated ahead of Illinois even after their win last night? Perception says that's false, since Illinois has beaten a lot of really good teams. Reality is that the data shows it thus, because Illinois hasn't been super-spectacular in other games. He also has a 14-11 Northern Iowa team at 66, but that's where the data points.

At any rate, again, it's just a tool. The probability measure for each game on Massey are fun (and pretty accurate), and of course, because of SPORTS! things happen. Like Wabash beating Wooster.

[Pat Coleman]

Because, you know, narrative!

Guy in the cube next to me is probably wondering why I'm laughing.

[Dave 'd-mac' McHugh]

smedindy... I get your points and I actually enjoy your breakdown of things - even if some of it needs rereading to keep it from going over my head . However, you do point out where there could be flaws... there are times when teams are rated below teams we all know from watching games is not the case. Sure, the data says that this is how it should be - but the reality and the actual watching of games says otherwise.

For example, Massey has UW-Whitewater #3 in Division III and UW-Stevens Point #4 despite the fact the Pointers have beaten Whitewater twice by an average of 9.5 points. (Yes, I know that isn't a gross example, but it still shows some questions.)

This example is a bit more abstract, but when I talked to Mark Edwards on Hoopsville the other night, he stated that Rochester was the best team they had played this year. However, Rochester is rated behind Illinois Wesleyan - who Wash U has also played.

(By the way, can someone tell them at Massey that it hasn't been Salisbury St. in a long time ).

I get that Massey is just a place for information and I do glance at it when I need some extra information as I look at teams for my Top 25 ballot or otherwise. But while this doesn't apply to you, smedindy, there are many people who put so much stock in Massey that the ratings are used as the only source for an argument or to try and prove a team is better than others think. This drives me a bit crazy for many reasons.

Yes, Massey has certainly improved, but as Pat points out, it isn't infallible. I just hope others will appreciate that instead of telling those like me that we are crazy for not putting more stock in the numbers.

[smedindy]

But Dave, we watch A game and get a perception. We don't watch EVERY game. When a coach says "X" team is the best they've faced it also has quite a bit to do with matchups and style of play and how well his team played on that day. Coach-speak is kind of noise to me anyway. Gotta cut through it.

Also, like any sporting event, there's luck involved at times.

Head to head's not as important over a universe of games. Take my Northern Iowa example in D-1. Their two worst losses were to Evansville (they were swept). But all of the data, and in fact reality, shows Evansville's not as good as UNI. It's just matchups and style and luck. In the UW-W and UW-SP example, how do you reconcile that UW-SP, while sweeping WW, lost to teams WW beat? It's not transitive at all.

As for Rochester, their ratings slip is due to the 14-pt loss at NYU. Not a horrible loss, but still, losing by 14 to that team, after beating them by 15 earlier, probably set some data points in a tizzy.

Besides, how much of a difference is being ranked 6th versus 3rd when we're dealing with a 400 team universe? Not much...

[Titan Q]

This example is a bit more abstract, but when I talked to Mark Edwards on Hoopsville the other night, he stated that Rochester was the best team they had played this year. However, Rochester is rated behind Illinois Wesleyan - who Wash U has also played.

But two things...

1) IWU played Wash U without starting SG Pat Sodemann (11.7 ppg, 3.2 rpg, 25-26 FT) and starting PF Victor Davis (10.8 ppg, 5.2 rpg) - those two will probably both be all-CCIW players this year.  (Those are CCIW stats.)

2) Teams evolve from December 1 to February.  IWU is a great example.  The Titans graduated their entire perimeter from last year (Eliud Gonzalez, Jordan Zimmer, John Koschnitzky, Stephen Rudnicki) and replaced with completely new guys - JV players and a transfer.  November and December was truly a "work in progress" as they say.  IWU's perimeter is dramatically different right now than it was on 12/1 -- in fact, most would agree it's second in the CCIW now to only Wheaton.


So I'm not sure that's a great example at all, Dave.  And I'd say that's an example where Massey filters out that subjective perspective that might be a little flawed.  In other words, it just knows opponents, scores, and margins over the course of the entire season.  It doesn't know that the team Mark Edwards saw way back on Dec 1 is playing much better now - actually playing about as well an anyone (11-0 in CCIW) - and is healthy.

[smedindy]

IWU's improvement is definitely shown in it's ratings, rankings and perception. Over time, teams will find their true levels. You can't really hide after 16 games or so.

[KnightSlappy]

Besides, how much of a difference is being ranked 6th versus 3rd when we're dealing with a 400 team universe? Not much...

It's like WAR in baseball -- a fairly terrible metric to cite if you're trying to prove Robinson Cano was better than Miguel Cabrera (7.8 fWAR to 7.1 last season, a relatively small difference), but pretty solid if we're comparing Mike Trout (10.0 fWAR) to Cabrera.

Massey won't be good at perfectly sorting out a clump of relatively equally talented teams (and as others have pointed out, perception isn't either), but it does put teams in the right neighborhood better than any human could. There will be some outliers from time to time, but 2011 St. John's shouldn't cause us to throw out the entire system.

I haven't investigated the 2011 MIAC very closely, but I would think that leagues with larger round-robins have a higher potential for variance in their Massey rating. If a number of teams "randomly" perform well in their few pre-conference games, the whole league will see a boost (or the latin phrase for the other way around). The bottom six teams in the league standings were 21-10 in the non-con that year and then went 40-81 in league play. That likely served to squish the bottom MIAC teams closer to the top MIAC teams (relative to D3 overall) than we might otherwise expect. Fewer non-conference games means there are fewer scores to link them back to the rest of the division which means we're likely so see more noise in their specific rankings.

[sac]

St. John's beat D2 St. Cloud St. that year in a real game.   Any D3 win over a higher level opponent will greatly improve their rating in massey.


These 3 games pulled the highest rated D3's up the ladder more than normal.

St. Johns beat St. Cloud
UW-SP beat St. Johns  ---the very next game for St. John's
St. Thomas beat UW-SP


This was probably a once in a decade scenario.

[smedindy]

Oh, and then St. Cloud beat Mankato who was 25-2 so that yanked the whole chain, including St. John's, up even more! It's an entirely connected universe.

[sac]

Oh, and then St. Cloud beat Mankato who was 25-2 so that yanked the whole chain, including St. John's, up even more! It's an entirely connected universe.

Exactly, not only did St. John's beat a D2 but they beat one with a winning record that went on and beat other good D2's.

How often has that really happened for a D3 school?

[AO]

Oh, and then St. Cloud beat Mankato who was 25-2 so that yanked the whole chain, including St. John's, up even more! It's an entirely connected universe.

Exactly, not only did St. John's beat a D2 but they beat one with a winning record that went on and beat other good D2's.

How often has that really happened for a D3 school?

The only bias I see in this discussion is of the East Coast variety.      

well I suppose in Pat's case maybe it's an attempt to mask his Minnesota bias.  like that ref that is buddy buddy with a particular head coach and to prove his lack of bias makes a lot of bad calls in favor of the other team.

[John Gleich]

For example, Massey has UW-Whitewater #3 in Division III and UW-Stevens Point #4 despite the fact the Pointers have beaten Whitewater twice by an average of 9.5 points. (Yes, I know that isn't a gross example, but it still shows some questions.)

Interesting you brought this one up... Because I certainly think that Whitewater is playing better basketball right now than UWSP, especially in the last 3 (Point is 1-2, Loss at Platteville #31, Loss at La Crosse #33 Win vs. Eau Claire #177; Whitewater is Win vs. La Crosse #33, Win vs. St. Norbert #24, Win vs. Platteville #31).

The injury bug has hurt Stevens Point even more than you and Coach Semling alluded to last night on Hoopsville... Stevens Point dressed 9 guys on Wednesday vs. Eau Claire and one of them (Ritchay) got hurt right before halftime and didn't play in the 2nd half. So of UWSP's 14 guys on the roster, 6 of them were/are hurt, including Tillema and two ADDITIONAL starters (Ryf and Ritchay).

Hopefully with the week off, a few of these injured guys will be back... but in many respects, if they're not, it doesn't make a whole lot of sense to my why Tillema would come back just for a few unguaranteed games (lose once and you're done), especially if he's only at about 80%. Now, certainly, Tillema at 80% is better than probably 90%+ of the players out there and would certainly help... but would it be enough help for a deep run? I'm not so convinced.

And I'm also not convinced that Point can't made a fairly deep run even WITHOUT Ty Tillema. Lost in all of the shuffle of injuries and the two setbacks last week is the fact that Point is only letting up fewer than 57 points per game in WIAC play this year, only turning the ball over 10 times per game (though that was up to about 13 without PG Ryf against EC) and they shoot 77% from the FT line. Those are all ingredients for success...

But without the players, it'll be really tough... as shown with recent results.

Yes, Massey has certainly improved, but as Pat points out, it isn't infallible. I just hope others will appreciate that instead of telling those like me that we are crazy for not putting more stock in the numbers.

This is a bit of a straw man here though... because nothing is infallible. How many upsets do we see of top 25 teams every week?

The top 25 takes a subjective look at the teams. Massey attempts to be more objective, based on certain criteria.

The great thing about basketball, though, is that, on any given night, a team like Illinois can beat #1 Indiana or TCU can beat #5 Kansas. Does it mean that Illinois is better than Indiana or that TCU is better than Kansas?

No, it doesn't. All it does is muddy the waters in terms of where Indiana or Kansas really should be. There's underachieving and there's overachieving... and it's possible for any team to do either.

Anybody can come out like Edgewood did last year at River Falls in the first round of the NCAA tournament last year and just shoot the lights out and win a game. The final score doesn't tell the tale of how Edgewood took a tie game with 7 minutes left in the first half to 7 at halftime... and then extend it all the way to 24 in the second half.  UWRF made a run and actually cut it to 5 with :30 left in the game, but the upset was not averted.

[smedindy]

Well, we have had some interesting results:

OWU beats Hiram 63-61 on a three pointer with 4.4 seconds left.

Denison beat Allegheny 90-75.

Wabash belted Kenyon 58-45, hitting 28 free throws.

Oberlin (!) shocked DPU 51-49 on a last second layup by Randy Ollie.

Wow. The middle is now totally mired in muddle.

[Ralph Lee Wilson]

Wabash should have won by one more, but the ref decided to grant Kenyon an extra free throw after a rhynie waved a towel behind the basket.  The ref said (and I quote) "I already told you no!". Needless to say, the ref was booed by all Wabash fans for the rest of the game.