Pool C

[nescac1]

Come on, man, trying to make extrapolations about likely annual NCAA prospects of entire leagues based on a single two-point upset in a single non-league game in December is beyond silly.  I could list a dozen such outlier losses suffered early in the season by better squads than this year's Eph team nearly every year.  Let's leep it real.  Indeed Smitty perhaps consider your own words just a few hours ago from a neighboring thread:

For a team who looked as good as Witt has this year to want to move them "WAY" down doesn't seem fair. I know you will bring up the lopsided score of 20ish points, but it is a single blemish to a decent team in a solid conference on the road on senior night. Teams lose, opposing teams hit shots and the ball may not bounce your way one day. I have been thinking this all season, a loss to solid/average team for all of these top tier teams are not the end of the world. Especially when you think about everybody getting their best shot. If it becomes a trend like it has for E&H or OWU I totally get it but just one loss for Witt did not make me significantly change their ranking in the poster poll.

And maybe just ONE IIAC team could make ONE Final Four before we start suggesting the league is underrepresented?  The argument about NESCAC makes sense only if just one or two Nescac teams made it to Salem in an eight year period.  When eight do, however, you can't really say that having more entrants is the only thing accounting for the league's success.

If just the top two Nescac teams made it in each year, the odds are at least 6-7 Nescac teams would stiil have made the Final Four, because typically it's the best one or two Nescac teams (often from Pool A) who end up advancing, not the arguably borderline selections ... in 2010, 2011 (2), 2013, and  2014 (2), those Nescac FF squads would all have been tourney selections under any imaginable system or schedule.   Maybe not last year's Eph team, but then again all they did was steal Midd's Salem spot in the end ... the only way to compare fairly is on a percentage basis.  What percent of a league's entrants win the title and what percent make Final Fours?   Nescac does well enough under both criteria to prove these selections are worthy. But no matter how teams actually perform once selected, which should at least be a relevant criteria to consider, folks will always complain, predictably, on an annual basis. 

[nescac1]

On another note, really interesting data Fantastic50, thanks. Surprised to see NCAC at the top of the list ... although Wooster and Wittenberg both look like legit contenders and perhaps accounted for lion's share of the wins? 

[Smitty Oom]

Come on, man, trying to make extrapolations about likely annual NCAA prospects of entire leagues based on a single two-point upset in a single non-league game in December is beyond silly.  I could list a dozen such outlier losses suffered early in the season by better squads than this year's Eph team nearly every year.  Let's leep it real.  Indeed Smitty perhaps consider your own words just a few hours ago from a neighboring thread:

For a team who looked as good as Witt has this year to want to move them "WAY" down doesn't seem fair. I know you will bring up the lopsided score of 20ish points, but it is a single blemish to a decent team in a solid conference on the road on senior night. Teams lose, opposing teams hit shots and the ball may not bounce your way one day. I have been thinking this all season, a loss to solid/average team for all of these top tier teams are not the end of the world. Especially when you think about everybody getting their best shot. If it becomes a trend like it has for E&H or OWU I totally get it but just one loss for Witt did not make me significantly change their ranking in the poster poll.

And maybe just ONE IIAC team could make ONE Final Four before we start suggesting the league is underrepresented?  The argument about NESCAC makes sense only if just one or two Nescac teams made it to Salem in an eight year period.  When eight do, however, you can't really say that having more entrants is the only thing accounting for the league's success.

If just the top two Nescac teams made it in each year, the odds are at least 6-7 Nescac teams would stiil have made the Final Four, because typically it's the best one or two Nescac teams (often from Pool A) who end up advancing, not the arguably borderline selections ... in 2010, 2011 (2), 2013, and  2014 (2), those Nescac FF squads would all have been tourney selections under any imaginable system or schedule.   Maybe not last year's Eph team, but then again all they did was steal Midd's Salem spot in the end ... the only way to compare fairly is on a percentage basis.  What percent of a league's entrants win the title and what percent make Final Fours?   Nescac does well enough under both criteria to prove these selections are worthy. But no matter how teams actually perform once selected, which should at least be a relevant criteria to consider, folks will always complain, predictably, on an annual basis.

I said in my original post that I know Williams is a better team, and I definitely agree it was an upset. But the take home point is that Hamline (not the class of the MIAC) is a decent squad that has the talent to win some big games. Much like Hidelberg out of the OAC, as they have some big wins this year as well. These middle to lower tier teams are still solid teams and I think they help show the depth and strength of these other conferences that only get 1-2 teams in a year.

This is absolutely nothing against the NESCAC and I am not complaining, because I get the how the process works and the NESCAC teams are absolutely deserving of receiving the pool C bids based on the criteria the NCAA puts out there (Hell I wish the MIAC go adopt the ODAC model as a chance to increase the pool c resume, but its hard to schedule more non-conference games in our D3 location of the US). I just think that the difference between the level of basketball played in the NESCAC and some of the other conferences isn't as large as the discrepancy of NCAA births for a conference (MIAC/WIAC/CCIW/IIAC) would show.


**PS: In my quote you posted from the top 25 room, what I was trying to get across was the fact that we should not hurt people for losses as much as we do (me included). I still think Williams is very deserving of their top 25 rank, even given their loss to Hamline. It says more about Hamline in the win than it does Williams in the loss. I hope I am conveying my thought process clearly here.

[Onward on, John Carroll]

Slightly off-topic, but perhaps interesting ... here are the records for each conference vs non-conference regionally ranked opponents:

NCAC 7-7 (0.500)
NWC 3-3 (0.500)
CAC 6-7 (0.462)
WIAC 5-6 (0.455)
NESCAC 10-13 (0.435)
UAA 7-10 (0.412)
CC 4-6 (0.400)
NEWMAC 5-9 (0.357)
NJAC 6-13 (0.316)
CCIW 5-11 (0.313)
ODAC 6-14 (0.300)
MIAC 4-10 (0.286)
OAC 4-10 (0.286)
LL 5-15 (0.250)
MACC 3-9 (0.250)
HCAC 2-6 (0.250)
LEC 4-13 (0.235)
SCAC 4-15 (0.211)
CUNYAC 2-8 (0.200)
IIAC 2-8 (0.200)
MWC 2-8 (0.200)
CSAC 1-4 (0.200)
E8 2-9 (0.182)
CCC 1-6 (0.143)
SUNYAC 2-13 (0.133)
SKY 1-7 (0.125)
LAND 2-15 (0.118)
NACC 1-8 (0.111)
NECC 1-8 (0.111)
MACF 2-18 (0.100)
AMCC 1-9 (0.100)
USAC 1-9 (0.100)
MIAA 2-20 (0.091)
MASCAC 1-12 (0.077)
SAA 1-15 (0.063)
ASC 0-1 (0.000)
NAC 0-3 (0.000)
PAC 0-8 (0.000)
ACAA 0-10 (0.000)
SLIAC 0-11 (0.000)
SCIAC 0-12 (0.000)
NEAC 0-14 (0.000)
UMAC 0-14 (0.000)
GNAC 0-15 (0.000)

I am no math savant so perhaps there is a mathematical explanation for this that I am missing, but I am blown away that not one conference has a winning records against non-conference regionally ranked opponents.

[toad22]

I am aware that the NCAA has no interest in expanding the size of the D3 tournament. It was a long bloody slog just to get up to 64. However, my point was that the way to construct the best national championship is to figure out ways to get the best teams in, even if that means taking more teams from the top leagues. That is where more of the really good teams live.

[fantastic50]

On another note, really interesting data Fantastic50, thanks. Surprised to see NCAC at the top of the list ... although Wooster and Wittenberg both look like legit contenders and perhaps accounted for lion's share of the wins? 

The NCAC has developed some quality depth in recent years, getting three NCAA bids more often than not (with 5 different teams appearing in the last 4 years).  Next week will be the first time since 2014 that a team other than Ohio Wesleyan will be the top seed in the conference tournament. 

This year has shown quality depth beyond Witt (15-1) & Wooster (13-3), as well. 
* Ohio Wesleyan (10-6) beat Ramapo by 29, and lost nail-biters to Whitman and at Hanover.
* Hiram (10-6) played a soft non-conference schedule, but hammered Wittenberg, and nearly split with Wooster for the 3rd year in a row (losing in OT at home)
* Denison (8-8) blew out Cabrini and almost swept Wooster, winning easily at home and barely losing on the road. 
* Wabash (8-8) won at Wooster by 13 points.
* DePauw (6-10) beat Rose-Hulman and took UW-River Falls to overtime. 
* Oberlin (4-12) nearly won at Wooster in the season opener.

[fantastic50]

I am no math savant so perhaps there is a mathematical explanation for this that I am missing, but I am blown away that not one conference has a winning records against non-conference regionally ranked opponents.

When I first ran those numbers, I thought that I had a mistake.  However, regionally ranked teams get there because they win most of their games, so we should expect few conferences have a winning record against strong non-conference teams.  That being said, I thought that at least one power conference would have pulled it off.

[KnightSlappy]

I am no math savant so perhaps there is a mathematical explanation for this that I am missing, but I am blown away that not one conference has a winning records against non-conference regionally ranked opponents.

When I first ran those numbers, I thought that I had a mistake.  However, regionally ranked teams get there because they win most of their games, so we should expect few conferences have a winning record against strong non-conference teams.  That being said, I thought that at least one power conference would have pulled it off.

Right. And since most teams are not regionally ranked, most of these games are unranked teams vs. ranked teams. In which case our expectation is that the ranked team probably wins.

[nescac1]

Smitty, the comment of yours I am focused on most specifically is: "Teams lose, opposing teams hit shots and the ball may not bounce your way one day. I have been thinking this all season, a loss to solid/average team for all of these top tier teams are not the end of the world."  I agree with that comment 100 percent.  I saw the Williams-Hamline game, and that is exactly what happened.  The Ephs didn't play their best, and Hamline simply could not miss anything at all down the stretch, they played a tremendous game - one they likely aren't capable of replicating 3-4 times in a row.  Heck, the 2014 Williams team that came within four seconds of a title in an absolutely loaded season at the top of Division 3 lost to Southern Vermont early in the year ... no one would argue that more than one team from Southern Vermont's league should EVER be in the NCAA tourney, however, despite that flukey result. 

Just as folks shouldn't blow a single surprising result out of proportion in top 25 rankings, you shouldn't make broad, overarching claims about what teams belong in the NCAA tourney based on a single result.  On any given day, sure, a middle tier team from any decent conference can beat a top 5 or top 10 team when everything goes right for them.  But that fact doesn't mean that the middle tier conference necessarily has been robbed of NCAA bids, or that its second or third place teams would likely go very far in the NCAA tournament if selected.  The NCAA selections are based on aggregated performance, not outlier games.  And the reality is that, most years, the fourth-best team from NESCAC (or WIAC, or CCIW) is far more likely to perform better in the NCAA tourney than the second-best team from most other leagues.  And we've seen that proven again and again, since runner-up teams in the WIAC, CCIW and NESCAC make boatloads of Final Fours (even when only 2-3 teams from those leagues are selected) while many leagues have never seen even their very best teams get close.  That's not to say those other leagues don't deserve a shot at the dance -- that is the point of Pool A!   

I'd say the best way to make selections is a mix of numerical criteria and human rankings, much like Division 1 employs.  If the method changed, though, there would still be multiple NESCAC teams getting selected in the average year (maybe five was excessive last year, but the league would very likely have gotten four in).  Just look at this year, when there are four NESCAC teams comfortably within the top 25 ranking.  The difference would not be MORE second and third place teams from mid-tier conferneces getting selected, but likely more one-bid conferences as the CCIW, WIAC, and UAA, in particular, and the OAC, MIAC, ODAC and NCAC, to a lesser extent, would earn almost all the other Pool C bids.   My point continues to be that while the NESCAC is a beneficiary of the current system, it would still get many teams in under nearly any fair alternative, including those that, say, weigh conference wins more heavily.  Other power leagues would just be (appropriately, in my view) in exactly the same position, whereas now they are not. 

[Onward on, John Carroll]

I am no math savant so perhaps there is a mathematical explanation for this that I am missing, but I am blown away that not one conference has a winning records against non-conference regionally ranked opponents.

When I first ran those numbers, I thought that I had a mistake.  However, regionally ranked teams get there because they win most of their games, so we should expect few conferences have a winning record against strong non-conference teams.  That being said, I thought that at least one power conference would have pulled it off.

Right. And since most teams are not regionally ranked, most of these games are unranked teams vs. ranked teams. In which case our expectation is that the ranked team probably wins.

Great points.  Thank you.

[AO]

What if instead of just one game being used to compare conferences, every game was considered?

The NCAA selections are based on aggregated performance, not outlier games. 

If only that was true.   Performance goes way beyond winning and losing to me.  The SoS the NCAA uses does a poor job of comparing teams from different regions.  The home and away multiplier is still broken as far as I know.  You still get a major benefit from playing poor teams at home rather than on the road.

From reading back some of the conversation last year I think the committee is aware of the limitations of the data they currently get, but I don't see any good way to correct their selections without fixing the underlying data.  They should continue to pick according to the data provided so at least they're predictable and everyone can understand why they were or were not picked.

[nescac1]

Some data sets (like SOS) are biased in favor of NESCAC or other leagues.  Others (like the one you are employing) is in my view a bit biased in favor of Midwest teams and seems to unfairly punish NESCAC and other leagues from the entire eastern third of the country.  No data set is ever going to be perfect.  SOS is, I'd agree, much worse than some alternatives, and could easily be tweaked for better results.  But in the end, I think data can only take you part of the way.  Rather, having human beings (like, say, the human beings who vote on the D3 top 25 poll) play some role in evaluating teams, beyond just numbers, would be beneficial.  But it seems unlikely that this will ever happen. 

[sac]

Slightly off-topic, but perhaps interesting ... here are the records for each conference vs non-conference regionally ranked opponents:

NCAC 7-7 (0.500)
NWC 3-3 (0.500)
CAC 6-7 (0.462)
WIAC 5-6 (0.455)
NESCAC 10-13 (0.435)
UAA 7-10 (0.412)
CC 4-6 (0.400)
NEWMAC 5-9 (0.357)
NJAC 6-13 (0.316)
CCIW 5-11 (0.313)
ODAC 6-14 (0.300)
MIAC 4-10 (0.286)
OAC 4-10 (0.286)
LL 5-15 (0.250)
MACC 3-9 (0.250)
HCAC 2-6 (0.250)
LEC 4-13 (0.235)
SCAC 4-15 (0.211)
CUNYAC 2-8 (0.200)
IIAC 2-8 (0.200)
MWC 2-8 (0.200)
CSAC 1-4 (0.200)
E8 2-9 (0.182)
CCC 1-6 (0.143)
SUNYAC 2-13 (0.133)
SKY 1-7 (0.125)
LAND 2-15 (0.118)
NACC 1-8 (0.111)
NECC 1-8 (0.111)
MACF 2-18 (0.100)
AMCC 1-9 (0.100)
USAC 1-9 (0.100)
MIAA 2-20 (0.091)
MASCAC 1-12 (0.077)
SAA 1-15 (0.063)
ASC 0-1 (0.000)
NAC 0-3 (0.000)
PAC 0-8 (0.000)
ACAA 0-10 (0.000)
SLIAC 0-11 (0.000)
SCIAC 0-12 (0.000)
NEAC 0-14 (0.000)
UMAC 0-14 (0.000)
GNAC 0-15 (0.000)

Ranked vs ranked would be much better.  I'm not sure its very useful to know how near "bottom of conference team X" faired against a ranked opponent.  That would be an expected loss in almost every instance.


OTOH I find it fascinating the MIAA has 22 games against ranked opponents, second only to the NESCAC.

[fantastic50]

Ranked vs ranked would be much better.  I'm not sure its very useful to know how near "bottom of conference team X" faired against a ranked opponent.  That would be an expected loss in almost every instance.

OTOH I find it fascinating the MIAA has 22 games against ranked opponents, second only to the NESCAC.

Ranked-vs-ranked may slightly penalize conferences that squeeze in an extra team near the bottom of a region's rankings, but here are the ranked-vs-nonconf-ranked records...

CAC 4-0 (1.000)
NWC 3-0 (1.000)
LAND 2-0 (1.000)
MACC 2-0 (1.000)
UAA 6-2 (0.750)
NCAC 5-2 (0.714)
NEWMAC 4-2 (0.667)
MIAC 2-1 (0.667)
OAC 3-2 (0.600)
NESCAC 7-6 (0.538)
CC 3-3 (0.500)
ODAC 3-3 (0.500)
CSAC 1-1 (0.500)
NJAC 4-5 (0.444)
LEC 3-4 (0.429)
WIAC 3-4 (0.429)
CCIW 2-4 (0.333)
HCAC 1-2 (0.333)
LL 1-2 (0.333)
MASCAC 1-2 (0.333)
SUNYAC 1-3 (0.250)
ASC 0-1 (0.000)
CCC 0-1 (0.000)
NEAC 0-1 (0.000)
IIAC 0-2 (0.000)
USAC 0-2 (0.000)
E8 0-3 (0.000)
SKY 0-3 (0.000)

The ACAA, AMCC, CUNYAC, GNAC, MACF, MIAA, MWC, NAC ,NACC, NECC, PAC, SAA, SCAC, SCIAC, SLIAC, and UMAC are all 0-0, either because of having no regionally-ranked teams (last week) or those teams playing no such nonconference games.

[Ralph Turner]

I wonder if this is something that would be easy to simulate by our friend Fantastic50? What would the NESCAC Pool C situation look like if they played a double round robin schedule? I might be able to toy with some numbers but I'm not able to simulate like he is.

My gut tells me that we make too much of the single round robin. It's an advantage for the top teams, but I'm not convinced how big it is.

Yes, I could simulate it, but it would take some time to set it up; it might be something to look at after the season. 

I don't hear anyone arguing that an 11-team conference should be playing a full double-round; the only reasonably strong conference that plays more than 18 conference games is the MIAC.  However, it wouldn't be hard to play 14-15 conference games via divisional play (perhaps with the three Maine schools plus Tufts & Middlebury in one division, and Hamilton joining the five remaining CT/MA schools in the other).  Because of the trios that already play home-and-home, over half the league plays 12 games against conference opponents now. 

Going to 14-15 games wouldn't have a huge impact, but the presumed extra losses would hurt some bubble teams; the NESCAC would still get four teams in this year, but wouldn't be in the running for a 5th.  Because their SOS is already "maxed out", as alluded to, a full 20-game double round would crush WP numbers with little benefit except even higher vRRO totals; my guess is that the league would be down to 2-3 bids, like the WIAC & CCIW.

I like the idea of running the NESCAC's numbers with more in-conference games.

My suggestion is to run the numbers off the 2017-18 schedule.

Arbitrarily divide the NESCAC into 2 divisions and have single round robin inter-division and double round robin intra-division.  This format is commonly used by larger conferences, including the 13-member ASC which ends up with 17 games for the ASC West and 18 games for the ASC East. 

In the arbitrarily assigned NESCAC games, declare the winner as the team with the higher conference or higher overall season record.

Substitute the in-conference replacement games for the later non-conference games in the season, when most conferences are in full mode.  This seems to be the least "punitive/cherry picking" in eliminating non-conference games.

I would love to see the impact that it has on the NESCAC SOS, too.  +1! in advance.