Site Content Feedback and Miscellaneous Discussions

[PaulNewman]

I started this thread because I accept the complaints about clogging up the national thread, I don't want to clog the Great Lakes thread, and of course the issues are not just Great Lakes issues.

RH's latest column, per his usual, is well-done, clean and crisp, well-written, etc.  A very nice writing style that isn't wordy or gratuitous in the least.  Always enjoy them even when there are points I disagree with.  I do have a few points to make, not to counter just for the sake of countering, but honestly with an interest in forwarding the discussion.  We all now understand that the process is what it is, and no changes are going to occur in the next 3 weeks (nor likely ever because of some comments a poster on this site makes), but still....

Where RH says "in a word, scheduling," I would instead say "in a word, scheduling and how that scheduling works out for you."  As I suggested in another post with some obvious bite (sorry), in many cases we are talking about sophisticated coaches in established programs who have not just fallen off the turnip truck, who have established a pattern of scheduling over a period of years, who have served on regional cmtes, etc, etc.  [A side issue that hasn't been really vetted yet is exactly how scheduling happens and whether, for example, it typically is done season to season or 2 years in advance or whatever, and whatever politics/logistics are involved with scheduling.  Would like to see that explained/discussed as well.]  A couple more points on scheduling.  I guess most of us should have known about the AWAY games factor, but I think this is problematic, and not just because of the weighting that TJ has commented on.  Serpone at Amherst, quite  rightly, has been praised for scheduling only 5 homes games.  But should he have to, and does he even like this?  Credit to him for his use of the system as is, but in what universe (or sport) is playing less than 50% of games at home considered ideal or built into a system as an incentive???  Imagine Loras calling Messiah, and saying "we want to play you guys but we only want to play you AWAY!"  Now that just seems wrong and backwards.  And there are other reasons, including fan support (usually family and friends at this level), academic burdens, etc, etc.  Just totally counter-intuitive for me.

Which brings up another point.  Those of us questioning/complaining/challenging (even if we should have known better) aren't necessarily requesting a less numbers-driven or less "objective" approach or endorsing a more "subjective" approach.  I completely agree with RH that we want schools to play good schedules and to reward them to some degree for doing so.  There are questions about proportions, which I won't go on about here, but I think it's important to underscore that those of us perceived as potentially in the "sour grapes" category aren't arguing for a more subjective approach.  In his argument about this, RH emphasizes not "opening a can of worms" and "the minute we allow subjectivity into the system....," as though we should conclude we are clean right now in all those regards.  But we aren't (I don't think).  We've been told, without knowing the proportionality, that there IS room for the cmte members to have influence/input.  With the current system in place, are we prepared to say that there aren't any concerns of the type he noted....like a coach or team being disliked, a chair pulling in his team or a friend's team, choosing to weigh one numerical factor more than another in certain cases, etc, etc???  To bring up Serpone again, we've heard that he can NEVER win NESCAC COY.  I actually don't think this is true, partially because the others will lose credibility if they don't (if Amherst goes clean from here through the NESCAC tourney).  Anyway, you get the point.  One could read RH's column and assume what he raises as a concern has been completely accounted for and never happens, and indeed, we've been told that there still IS a subjective element.

So, back to the SOS.  Again, I fully endorse that SOS should be among several to a handful of important "objective" variables, but maybe we need a better way to derive it.  Let's consider a couple of teams that I am not a shill for....TMC and CNU, and keeping in mind the idea that coaches should be able to plan sufficiently including accounting for an occasional quirk event, cancellation, etc.  Oh, and before I get to that, schools should be held accountable, as RH suggests, but they CANNOT be held accountable for their conference schedules,which is the MAJORITY of the schedule, so, in terms of who was on top of this and who wasn't, we're really only talking about the non-conference schedules.

Here's TMC's -- Wittenberg, Case Western, Capital, Denison, Centre, OWU, Heidelberg, Transy, Marietta, DePauw

Sorry, there is NO ONE on this board who is going to look at that and tell me that isn't a MONSTER schedule.  I recall actually congratulating them a couple of months ago on playing such an aggressive schedule.  With the possible exception of DePauw, for me that's the toughest schedule in the entire Great Lakes region, and yet they came in at a paltry .516, and instead of the #1 or #2 ranking they deserved, they slipped to #5.

Here's CNU -- Randolph, Virg Wesleyan, Neumann, NC Wesleyan, NJ City, Rutgers-Camden, Methodist, Greensboro, Lynchburg

Won't go into the same detail, but again, that looks VERY STRONG, and yet CNU came in at .513.  CNU also has one of the most knowledgeable, experienced, iconic coaches in the business. 

So my biggest disagreement is with the idea of crediting those teams who ended up high SOS as being better prepared with the suggestion that those that don't have a high SOS by definition didn't do a good job and therefore there's really nothing to challenge here.

[PaulNewman]

Let me anticipate one response in advance.....that TMC will be fine once RvR comes in.  That's true and that's great, but that's not the point.  Who is fine and who isn't going to be fine is an entirely different discussion.  This about how this particular criterion, which is weighted VERY heavily, is derived and whether it has correctable flaws.  If a gymnast gets 4 on balance beam and 10s on everything else and still wins, that doesn't mean we should ignore that she should have gotten a 9 instead of a 4.

[Ryan Harmanis]

I won't have a chance to read/respond in any depth until later, but I'm totally on board with NCAC's premise.  Feedback is welcome and the discussion on this has been really, really good.  If we can coalesce around a few thoughts/responses to today's piece (or any other piece, for that matter), I'll be happy to offer a different view next week.

[PaulNewman]

I won't have a chance to read/respond in any depth until later, but I'm totally on board with NCAC's premise.  Feedback is welcome and the discussion on this has been really, really good.  If we can coalesce around a few thoughts/responses to today's piece (or any other piece, for that matter), I'll be happy to offer a different view next week.

Thanks RH, and I will be PMing you later just fyi.

[PaulNewman]

Before I forget, just wanted to stick in this portion of FW's post in another thread about Messiah's SOS.

"Put that all together (W&L, Rowan, and E-town having unexpectently good to great seasons; two winningest opponents played away, playing best conf. foes first), and you have a surprisingly high SOS at this point of the season.  The fact that the overall non-conference schedule isn't as strong was more than offset by a bunch of unexpected or fortunate things."

[Flying Weasel]

I am all for tough schedules and encouraging, even "demanding", that tournament-aspiring teams play challenging schedules with measuring-stick opponents. But do not count me among those who get judgmental about coaches not scheduling tough enough. Sometimes it seems rather obvious that little attempt was made to toughen up the non-conference schedule, and then, yes, criticism is in order. But there are so many factors involved, and many of them out of one's control, that I'm hesitant to call out coaches for not scheduling better since I can't know all the factors that resulted in the schedule they ended up with. 

I'm very sympathetic to the difficulty of the west coast schools (SCIAC and NWC) to do much about their strength of schedule unless they are able and prepared to spend a lot of money and time on travel (which doesn't really fit the Division III ethos).  Due to lack of non-conference options, they play each conference foe twice. Obviously, it's only the top teams that would have reason to be concerned about the effect of SOS on their tournament chances, and for those teams, well, they are almost guaranteed to have a SOS below .500 from their conference play (which makes up basically 75 to 90% of their schedule) because they can't play themselves and thus play 1 more team (2 more games) from the bottom half of the table than the top half.  And somehow in the 2 to 4 non-conference dates they have they are supposed to bump their sub .500 SOS up to something that can compete with other teams with countless more options within 120 miles than they have within 1,200 miles.  It's really not fair.  Not sure what the solution is--geography is what it is.  But I feel for a Whitworth as they seem to try to do the best they can having traveled east to New Jersey in 2010 to play a pair at Montclair St in 2010 and had a home and away series with Wheaton (Ill.) in 2012 and 2014 along with weekend trips to Minnesota, and Texas other years. 

And as has been pointed out, sometimes what seemed like a tough game when originally scheduled turns out not to be (and vice versa).  Who expected York to fall so far from one year (2013) to the next (2014) and actually hurt one's SOS rather than help it?  Typically these things will balance themselves out to some degree, but not always and not for every team in a given season.  And so again I have sympathy for the teams that seem to have a solid schedule on paper before the season starts and then end up having it turn out to be more on the mediocre side by the misfortune of catching multiple teams in an down year.  However, if the schedule did turn out to be weaker than expected, than all the more reason for the team to have done well and achieved a winning percentage even higher than anticipated which should offset the lower SOS somewhat.  But given how highly it seems the committee values SOS, it's tough to completely compensate for a lower SOS with an extra win or two.  And if the SOS fell, that could mean less games against ranked teams than might have been anticipated.  A double whammy, since the record vs. ranked also seems to be very important criteria for the committee.  However, I do think it's the exception not the rule that a team will have a significant number of its opponents having significantly poorer seasons than normal.  Can and does it happen?  Sure it does, but it's still the exception.  And a very unfortunate one, for sure.  But again, I'm not sure exactly what can be done about such misfortunes.

Now I've had some ideas about the SOS for some time now (besides getting rid of it and developing another one from scratch). To the extent that the current simplistic SOS formula stays in use, I think they should at a minimum tweak two things.  Have a minimum opponent's winning pct. that can be used in the calculations.  If an opponent's winning pct. is below something like .400 or .425 or .380 or whatever (a statistical study could be done to hone in one a good threshold), than the threshold value is used instead of the actual winning percentage.  A cupcake game is a cupcake game for tournament-aspiring teams.  I don't like that we differentiate between a 1-11-3 opponent and a 4-9-2 opponent when comparing bubble teams for a tournament berth.  Second suggestion is get those home and away multipliers much closer to 1.0.  A near 50% value difference is absurd.  I really struggle to grasp how anyone would have thought 1.25 and 0.85 to be reasonable and sensible, much less enough people to have adopted this as official policy.  There just isn't anywhere near that level of home field advantage across the board in D-III soccer.  Maybe there's a handful of venues that are intimidating (I'm don't know of any--playing Messiah before this year was intimidating but not their venue itself).  And certainly there are some fields that due to being in poor condition, or small in size, or whatever give their home teams who are more accustomed to playing on it an advantage, but that's still a minority and is the advantage really worth 50% more?  Again, couldn't some study be done to capture the difference in home and away winning percentages across D-III and specifically among tournament-caliber teams?  And then based on that, set more appropriate multipliers.

Overall, I'd want the subjective component of the process increased.  After reading the committee's own arguments in favor of a change in the definition of a ranked team with regard to the record versus ranked opponent criteria because of how the criteria resulted in them not selecting John Carroll, it was more clear than ever to me that they don't feel they can subjectively override what the quantifiable criteria tell them.  To me the criteria should be what is to be considered, but the act of considering certain data should mean you let the data make the decision for you.  You let the data inform your decision, guide your decision, but not make it for you.  Some discernment, judgement, deeper analysis, and, yes, plain common sense should be used.  Now I am sure there is an element of subjective-ness in trying factor in five different criteria and compare teams across the five criteria, but when the committee itself seems to be saying/admitting that they believe it was wrong that John Carroll was not selected, then that tells me they don't feel enough freedom to go against the quantifiable data even when their reason, discernment, and common sense tell them otherwise.  And I find that a shame and problematic.

[Ryan Harmanis]

My thoughts in response to NCAC's (and now FW's) nice points:

(1) The use of the away multiplier, while perhaps imperfect, makes sense because it is more difficult to play on the road.  Maybe we need a new a multiplier, but that would require calculating exactly how much more likely it is to win a game at home than on the road.  The multiplier should accurately reflect the difference between home and away, correct?  Maybe that's an offseason project to look into, but figuring out how close the multiplier is to the actual difference between home and away would either show the metric is actually pretty accurate or offer a solution.  (It looks like FW beat me to the punch on this, but tweaking the multiplier would really only require someone to calculate what the multiplier should be to account for any variance in home versus away games.)

However, I don't know if I agree as much on setting a floor for cupcakes.  Yes, 2-14-0 (0.125) versus 4-12-0 (0.25) versus 6-10-0 (0.375) might all seem easy to top teams, but doesn't that balance out on the other end?  I mean, is it that much tougher to play 12-4-0 (0.750) versus 14-2-0 (0.875) versus 16-0-0 (1.000)?  Maybe, but aren't those small differences in degree probably the same as on the low end?  So wouldn't we need to do the same thing for good teams?  I wonder if it would skew things if we artificially bump the teams with bad win percentages without capping good win percentages.  The more I think about trying to calculate a more accurate home-versus-away multiplier, the more I think that's probably the better solution.

(2) I think Messiah's current SOS cuts the other way.  FW's point, if I understand it correctly, was that because he didn't use the numbers to evaluate the schedule and just eyeballed the thing, he underestimated Messiah's SOS.  But ETown, Rowan, W&L, Lycoming having better years means that Messiah's SOS should still be strong - so the metric is doing its job.  If those teams were having "normal" (i.e. worse) seasons, then maybe Messiah is in the same boat SOS-wise as some of these other teams, and I'd make the exact same point about them.  But the SOS metric keeps me (and the committee) from making that mistake and viewing teams through the lens of history, reputation, etc., rather than what's actually occurring this year.  Bottom line, I very strongly believe that if you play in a good conference, you can structure your schedule to avoid the SOS problems, even if bad luck comes into play.  How often are unexpected/unfortunate things going to make your really strong schedule result in a SOS so low that you can't get an at-large bid?  I can't think of a time that's happened, where a team had a schedule that normally would have been good enough to hit 0.540 and it drops to 0.500 based on circumstance.  Even if that happens once in a blue moon, it's not going to happen nearly as often as that same team will be rewarded for having a really good SOS in other years.

(3) Teams being punished for a weak conference happens in every college sport, including DI sports where subjectivity plays a greater role.  Schools outside the Power 5 have no real chance to make the College Football Playoff, and very few DI schools from the weakest basketball conferences have any shot at an at-large.  So yes, I agree that Thomas More's non-conference schedule is good, but it barely does enough to make up for the conference slate.  That's just the nature of the beast.  It's also not that unfair to see TM have a lower chance of an at-large bid, because they have such an easier path to the AQ.  TM has won its conference tournament six years in a row - that's something that Loras, OWU, Wheaton, and even Messiah haven't pulled off.  So on the whole that balances out IMO.

I am much, much more sympathetic to the plight of teams (like Whitworth) that are geographically prevented from creating a stronger schedule.  On that front, I just don't know how we solve it unless we use much more subjectivity (like way more), or do something like guarantee more bids for the West region, since that's the region where this problem pops up.  Perhaps the NWC has already figured it out by using the regular season for the AQ.  As a general rule, the regular season - especially a double round robin regular season - is much more likely to ensure the best team gets the bid.  Now, you still have a problem if there are two really good/deserving teams, but it's at least mitigated.

(4) While I agree that there is some room for subjectivity, you have to be in the vicinity with the criteria for it to matter.  That's where Whitworth, Kenyon, CNU, etc. are losing out, because the gap between their SOS and the top-ranked teams is much, much greater than any difference in winning percentage.  To use your gymnastics analogy, the gymnast who goes 9-9-9-4 is going to lose to the gymnast who goes 9-8-8-8.  So if you don't have an SOS in the ballpark - setting aside whether the current metric is the right one to use or not - then any subjective view that could go the other way is useless.  Even head-to-head isn't that helpful, because of the circular "Team A beat Team B beat Team C beat Team A" problem (which is the exact Kenyon-Denison issue in the GL).

And even if subjectivity comes into play right now, its role is greatly reduced because the committee is not allowed to take teams with very weak SOS or winning percentages and put them above teams that have balanced profiles.  I can look at these criteria and see why Whitworth isn't ranked, or why Kenyon is ranked so low.  I can't look at rankings that have them at the top of their regions, above teams with much better overall profiles based on these criteria, and understand how they got there aside from the committee "thinking" that they're better than those other teams.  Because even if I agree (which in some instances I do), my view is still just a "belief" that one team is better than another.

The problem gets even worse once we get into comparisons outside the region, because these rankings are how the committee picks the at-large teams - my understanding is that they look at the top team in each region, pick the best one for a bid, and continue until the bids are allocated.  So let's say Team X is the average 2014 at-large team, which came in at 13-2-3 with a .578 SOS (so basically Coast Guard).  Team Y (Whitworth, Kenyon, TM, etc.) is 15-2-1 with a .510 SOS (the extra loss from failing to win the AQ). Team Y is very high in the polls and beat Team X during the season.  The committee puts Team Y above Team X in the regional rankings because they think that Team Y is better even though the criteria support Team X.  Come Selection Monday, Team Y will be going against the top teams from the other regions, with profiles that resemble Team X's.  And we can't say Team Y is better than these teams from other regions, because there's no head-to-head or common opponents to use.  Based on the criteria it's no contest, so Team Y keeps losing out, and teams from other regions eat up at-large bids.  Maybe, right at the end, Team Y gets a bid over a team who went 13-4-3 with a .516 SOS (Texas-Dallas).  Maybe not.  But Team X never even got considered because they were below Team Y in the region, even though their profile was flat-out better - in every respect - than a lot of the other teams that got in.  How do we explain that?  Because we "know" that Team Y was better than Team X?

To respond to FW's thing on John Carroll, maybe that's exactly what happened.  OWU passed them in the GL, and OWU was on the low end of the SOS and the win% range, so OWU probably got one of the last bids.  If that's the case, then at that point there just weren't many bids left, and the one team JC may have been above (Texas-Dallas) was the only team from the West to get an at-large.  So in that sense, JC may have been shut out because the committee decided to give each region at least one at-large bid.  Maybe the solution is to keep the RvR aggregated for teams that were ranked at any point?  JC's SOS was 0.535, which would have been the second-lowest, but not crazy low.  And if they'd kept the RvR intact, then the rest of JC's profile would have been very competitive, with 4 RvR and an above-average win% at 0.800 (average was 0.764).  The average wins versus ranked last year was 2.5, so even if we increase that to 3.5 to account for other teams getting the same bump, JC would still have been above-average.

I just don't know how we do the comparison thing for teams from different regions without the really strict criteria, because unless we just alternate regions or something, it's got to come down to subjective beliefs about individual teams or regions from coaches who have never even seen them play.  At that point, if it's subjective, I worry that talent/reputation/personal feelings might overcome what's actually going on.  Maybe that's preferable, but I'm not quite as convinced.  One option could be to put every team on the board without listing the team name or region, but then you could have seasons where one region is entirely shut out while another gets 7 or 8 bids.

[PaulNewman]

FW, I agree with your post in its entirety (convenient?  maybe, but everything you said made logical sense to me...).

RH, your post is very dense and stretches me mathematically further than I can comfortably go without really hurting by brain, and if I try to answer each thing as well as I probably should I'll probably only succeed in revealing that I had a less than ideal performance on the math portion of the SAT  .

So I may meander which fits my style better and perhaps I will luck into something that makes some sense.

First, I want us to keep in mind as a backdrop a very important thing that I think FW noted but which may get lost because he could have underscored it even more....the idea of the "D3 ethos."  Whatever the system is should parallel or at least not significantly violate that.

I was watching some of the Oglethorpe-Centre match last night, which I almost hate to admit.  But I had been wondering about other teams that might or might not have a claim about being ranked and looking at data sheets.  So I was watching and thinking maybe Oglethorpe has a case to be ranked and thinking about another win on their record.  And then I also was thinking about Centre winning or losing another game and what impact that would have on "my team's" SOS.  I sort of caught myself and thought "What am I doing?  This is D3 soccer....enjoy the game and stop thinking about how the result is going to impact everyone else's numbers."  Turns out Oglethorpe won in a really good game, 3-2, and all I could think about was that hurts my team, lol, even though the game itself seemed at least 4-5 degrees of relevance apart from my team.

I'm big on consistency and apparent contradictions, and I feel you've contradicted yourself when invoking NCAA D1.  You previously suggested that D3 can't (and shouldn't even want to) use a D1 method for selecting teams, as part of your reasoning for a more strictly formula-based system (which I think the D1 folks btw would say they also use), and then in the TMC discussion you suggest that just like in D1 it is reasonable and even desirable to credit or give extra-credit stronger conferences, and that it's OK for TMC to be punished in that regard.  Aside from the fact that their conference schedule doesn't (yet) account for how low their SOS is, as they have been nearly fully dinged on that with more conference games to go, I think you are making the mistake of confusing how things should be with mitigating likelihoods.  Yes, TMC has won 6 straight AQs, but just like Whitworth, they shouldn't by definition have to be in the position of winning it.  There are multiple problems with that reasoning: 1) TMC one of these years, and possibly this one, is going to lose to a Grove City or Geneva 1-0 or lose a PK shootout; 2) what about any other teams ever getting a bid in that conference if they have a great year and don't get an AQ; and 3) so why did TMC even bother with such a tough non-conference schedule if it's not going to matter anyway? (What's the incentive at least in terms of just this SOS discussion?).

I find the same error in your reasoning with respect to Messiah.  You've made a big point about teams being able to plan and being credited for that.  The fact that Messiah's schedule is working out so favorably should count for them for their SOS, but that's a far different conclusion that actually crediting them for the foresight when as FW explained they wouldn't have expected their SOS to fall so favorably.  I agree that most often and over time the patterns of "good planners" will work out but we're seeing this year, with multiple examples, that planning may not work out.  And, back to the "D3 ethos", do we really want our D3 schools thinking like the Power 5 in D1?  You still want to reward away games, but do you really want D3 schools micromanaging their SOS's to the point that they schedule "extra" away games to game the system?  Does Serpone really believe the ideal is having only 5 home games?  Really?  In D3?  So on the one hand you don't want our D3 cmtes to be unduly burdened (or empowered) to go beyond the numbers (and btw, we still haven't settled how much they do or don't that anyway), and then on the other hand you're at least implicitly arguing that D3 coaches and ADs plan and go through their seasons like "Joey brackets."  Anyway, I want to highlight again that planning a tough schedule and having a tough schedule play out as a high SOS are really not the same things, even if we agree that most frequently there will be a decent correlation there.

I know I'm a little all over the map here, but also in terms of the crediting (or not) coaches for planning and tough schedules you may recall that just a few weeks ago I was lamenting the cancellation of the Case game -- NOT because of the SOS (as then I wasn't thinking SOS could in my wildest dreams be a problem, but rather because of the argument that you want your team to face good competition as preparation for good competition.  THEN, I was upset about the Case game because I thought Kenyon needed it as a lead-in to the DePauw game (which in my view turned out to be correct), and NOW I lament not having it for the totally different reason of the SOS.  At the time, I believe you chimed in that playing a tough schedule was very overrated, and talked about bedding in new players, experimenting with optimal line-ups, and being healthy as concerns that trump a tough schedule.  In fairness, I'm sure worries about SOS were the furthest thing form your mind as well, but now, in this context, I think it's fair to remind you of that position.

I'm going to have to re-read the Coast Guard and Team X vs Team Y stuff over several more times before I can attempt to respond to that.  On the gymnastics thing (in hindsight a bad example) I think you're making my point.  Yes, the 9-9-9-4 will lose and that was my point that a "4" incorrectly given based on the criteria can sink a ship.  And in that regard, I' suggesting that the SOS just isn't as accurate or as trustworthy as a tool as I think it should be, and we're seeing examples of that this year (which, again, doesn't mean I don't think some measure of SOS isn't important).  On the floor/ceiling thing, my intuition is that you can be far more punished by not having floor for how bad teams get calculated vs your examples of better records.  In other words, I think a legit 12-4 team is going to beat both the 1-13 and 5-11 teams with just about the same frequency as a 14-1 team.  The punishment for a 12 or 13 or more team is just too great.  Waynesburg is too heavy of a hit and it's not countered enough with the planned tough games like CMU, which despite I think being ranked as high as #3 or #4 earlier this year hasn't done as well as expected, nor has Centre (lost again last night), nor Heidelberg, etc, etc, while a Hanover, who I would not pick head-to-head vs Heidelberg for more than draw, has a relatively good record.  You mentioned that schools ought to be able to get to .540 every year.  Well, Kenyon for at least the past several has been at or above that level.  There is nothing materially different about their schedule this year, except for a couple more home games, and as I think I referenced earlier, I think you lamented thinking that OWU's schedule was softer than normal.

Finally, I feel like I'm having a brain block.  Did the NCAA specifically address the JCU situation from last year?  Where did I miss that?

[PaulNewman]

Addendum, which relates to the floor/ceiling issue.  Maybe a tweak could be (this is gymnastics, too, right, lol?) dropping the top team and the bottom team from the calculation in the out-of-conf portion of the schedule and having some adjustment that mitigates the damage of very poorly conference teams which no team has any control over since those schedules are laid-in-stone?

[Flying Weasel]

Let's consider a couple of teams that I am not a shill for....TMC and CNU, and keeping in mind the idea that coaches should be able to plan sufficiently including accounting for an occasional quirk event, cancellation, etc.  Oh, and before I get to that, schools should be held accountable, as RH suggests, but they CANNOT be held accountable for their conference schedules,which is the MAJORITY of the schedule, so, in terms of who was on top of this and who wasn't, we're really only talking about the non-conference schedules.

Here's TMC's -- Wittenberg, Case Western, Capital, Denison, Centre, OWU, Heidelberg, Transy, Marietta, DePauw

Sorry, there is NO ONE on this board who is going to look at that and tell me that isn't a MONSTER schedule.  I recall actually congratulating them a couple of months ago on playing such an aggressive schedule.  With the possible exception of DePauw, for me that's the toughest schedule in the entire Great Lakes region, and yet they came in at a paltry .516, and instead of the #1 or #2 ranking they deserved, they slipped to #5.

Here's CNU -- Randolph, Virg Wesleyan, Neumann, NC Wesleyan, NJ City, Rutgers-Camden, Methodist, Greensboro, Lynchburg

Won't go into the same detail, but again, that looks VERY STRONG, and yet CNU came in at .513.  CNU also has one of the most knowledgeable, experienced, iconic coaches in the business. 

So my biggest disagreement is with the idea of crediting those teams who ended up high SOS as being better prepared with the suggestion that those that don't have a high SOS by definition didn't do a good job and therefore there's really nothing to challenge here.

No doubt Thomas More is severely hurt by the bottom half of their conference as they have, not one, but three conference opponents with double digits losses and just one or two wins.  That is extremely difficult to overcome and not their fault at all.  But I take issue with calling their non-conference schedule a monster schedule.  Five of the ten teams are sub .500 and a sixth is just barely over .500.  Now the other four make for a very tough slate even if in the end we determine that Denison wasn't as good as their fast start suggested.  But to call it a monster schedule seems a bit over the top. 

Thinking more comparative, Rose-Hulman was a team ahead of Thomas More in the rankings with a better SOS.  I looked at the RHIT schedule and it has fewer good teams, but less sub .500 teams and just one double-digit loss team (who they haven't faced yet).  The Thomas More schedule has more measuring-stick games and for that reason among others, subjectively, their schedule would be considered tougher.  Included in that subjective assessment are things like the realization that Greenville, though over .500 and thus not hurting RHIT's SOS too much, went 3-7-0 in non-conference play before going undefeated in the SLIAC which I think we can all agree is one of the weaker conferences.  Thomas More has some sub .500 opponents that I believe are every bit as good as Greenville (and probably better), but they hurt Thomas More's SOS while Greenville doesn't have the same negative effect on Rose-Hulman's SOS.  Now the OOWP will help provide some correction for this, but not enough.

With Christopher Newport, again, I think you exaggerate when you call their schedule very strong.  But let's break it down as I think this is partially a case of being punished for your location/geography (though not nearly to the extent of the west coast) and partially a case of a couple really bad double-digit loss teams cancelling out all the better teams played.  CNU's non-conference schedule included the teams that finished #1, #3 and #4 in the USA-South (CNU's former conference) and #2, #3, and #6 in the ODAC in 2014.  In fact their three ODAC opponents finished #1, #2, #3 in 2013 and the three USA-South teams were #1, #2, and tied-#5 in 2013. Given that scheduling an opponent is a two-way street, not just you having an open date and wanting to play a particular team, I think they did very well to play five to six of the top ten teams in their geographic region (New Jersey is NOT part of the south even if is is part of the South Atlantic region).  They added a longer trip to Rutgers-Camden to play a weekend tournament where the price of playing a good RUC was playing a horrible NJCU (1-15-4 last year and on the same pace this year).  And even RUC isn't turning out to be as good as thought during the first month of the season.  The final team, Nuemann, was never going to bolster they SOS, but they weren't supposed to hurt it going 10-9-2 last year.  As it turns out, Neumann is 3-8-2, on pace for their first losing season since 2008 and worst season since 2006. No way CNU could have anticipated that. 

The fact is that the south (not the South Atlantic) has been overall been weak for the past decade.  What's south team to do about that?  Travel more, travel further.  There's options a lot closer than for the west coast teams, but is the fight for a strong SOS encouraging teams to spend more money and miss more class time for travel?  Is that in keeping with the D-III ethos?

Both Thomas More and Christopher Newport played a handful from among the top teams in their region, but a few (Thomas More - 3, CNU - 2) really, really bad teams (one or two wins, double-digit losses) more than cancelled out the good teams they played.  I just think that's ridiculous how much one really, really bad team can negate.  Especially when they are conference opponents you have no choice but to play which is the case with all three of Thomas More's worst three and one of CNU's two one-win opponents.  And that's why I disagree with Ryan and think a minimum threshold for winning pct. is needed.  But I'll address that in more detail in a response to Ryan's post.

In conclusion, I don't think Thomas More and Christopher Newport's schedules were as good as you suggest/claim, but I also don't think they are as bad as the SOS would lead one to believe.  I also don't think we can come down hard on their coaches for not trying to schedule a tough non-conference slate as they both seem to have done a decent job of that without getting crazy with long trips. And I think (a) the SOS calculation itself is a flawed measurement, (b) it's not tempered by enough subjectivity in the decision-making process, and (c) geography can clearly facilitate or hinder the development of a strong SOS while minimizing travel costs and missed class time.  And some teams pay the price, somewhat unfairly, for all that. 

[Ryan Harmanis]

It seems like we all agree that the SOS metric should be adjusted, but disagree on subjectivity, which I think should remain pretty limited.

The last post was more of a response and less of a coherent whole on my views, which is why there might be inconsistencies.  But on conference strength, my big point is that there are going to be weak conferences, and I think that it does matter.  I stand by the idea that because the AQ is so much easier to win in very weak conferences, it usually balances out the lower likelihood of a Pool C to a large extent.  I don't want to get all mathy, but if TM has an 80% chance of winning the conference tournament and Denison (probably the 4-seed in the NCAC) has less than a 25% chance, that more than offsets Denison having better odds for a Pool C.  I do get/understand your point about two very strong teams, but I'm still trying to come up with a less subjective system (aside from just fixing the SOS metric) that won't run into the problems where the system is unclear, it's just based on opinion, etc.

The Messiah example is perhaps not the greatest one this year, but in general I'd use them as a program that has scheduled ridiculously well out of conference to ensure that a slip-up in a weak conference doesn't cost them.  If I'm speculating, Messiah just lost one of the best classes in D3 history, so maybe there was an intentional decision to lighten the schedule, and they just got lucky?  I get someone could get unlucky in the same fashion - maybe Christopher Newport, as FW just illustrated - I just don't find it that likely and I'm not sure how we avoid it aside from, again, getting a more accurate SOS metric.

I don't know if we want coaches micro-managing schedules for home/away, but I think it's fair to expect coaches to take SOS into account if their goal is to get a Pool C bid.  We're talking about a competitive sport and the postseason, so even though D3 is more pure than D1 (IMO), that doesn't change my view that if you want to be competitive for an at-large bid you should methodically set up your schedule.  If a coach thinks otherwise, then they can schedule to better fit their view of D3 (or ignore it altogether), play for the AQ, and hope for the best for a Pool C.  But I don't think it's unfair to reward the teams that increase the competitiveness of a schedule when it comes to a competitive postseason.

The Problem with Subjectivity
My previous section on the inter-region comparisons is where I think subjectivity really runs into the biggest problems.  While it's fine to say (as my personal rankings do) that Kenyon is better than Denison, particularly after they've played, can we do the same thing with Kenyon and, say, Washington U. (0.750 win%, 0.574 SOS)?  I can do that shooting from the hip because I've seen Kenyon play, know the program, and above all my rankings have no consequences, but purely on paper how do we do that comparison in a way that's fair to both schools?  The national committee might have 2 guys who have seen either team play once or twice, and it's unlikely anyone has seen both play, so it just becomes (IMO) too hard to make that call without strict adherence to the criteria.

One possibility is to use more subjectivity in the regional rankings before using the criteria for Pool C bids, because the coaches know the teams better and so are maybe informed enough to make that call.  But that only exacerbates the inter-regional comparisons because for Pool C bids the committee is going to compare the highest ranked teams from each region.  So even though Denison has a stronger resume than WashU on the criteria (0.769 win%, 0.578 SOS), they never even get brought up because they're behind Kenyon.

Seems we all agree that changing the SOS metric might/should solve the problem.  I would also throw out there using RvR as any team ranked at any point, although you could be getting credit for beating a team that lost its last 6 games and finished 9-8-4.  I just have a hard time thinking that the benefits of increasing the subjectivity outweigh the costs.  Within the region, sure, we can all agree that the GL doesn't comport with our current view, but once we go outside the region to actually give out Pool C bids, I think it's nearly impossible to make that call in any fair, explainable, predictable way.

[PaulNewman]

FW, I'm going to disagree with you on TMC.  Maybe "monster" was too strong but I was comparing to other teams in the Great Lakes which is what is relevant to their status at the moment.   I'm also going to employ the thinking that you suggested with Messiah in terms of being accountable for what you planned vs how it pans out.  Capital, for example, has beaten both OWU and Kenyon in recent years.  Their overall record is horrible compared to recent years, and even this year Capital is surging in the OAC and now 4-1-1 in conference.  Compare them to a Hanover who has a better record and we can't assume that Hanover is stronger.  Transy and Centre also year in and year out would be considered solid to above average opponents.  They have 2 cupcakes IMO, Marietta and Wittenberg (and both I think are doing even worse than usual).  Otherwise, OWU, Case, DePauw and Denison are basically all of the top GL teams except for Kenyon, maybe ONU, and arguably RHIT.  Would you settle for strong to very strong?  Compare those 4 games to CMU ONLY playing Messiah and Kenyon and yet landing at .541.

As for CNU, I'm not sure the South in general gets treated fairly which made me think about Oglethorpe last night.  Every time I watch a Randolph or Lynchburg or Roanoke play a heavyweight like a CNU or NJAC team they look quite good and results are pretty good.  I've asked recently how Centre fares well against top NCAC teams and bombs out in the SAA.  Maybe the SAA teams are better overall than they are credited?  Do you think CNU made a reasonable effort to have a good schedule?  Should they be in a position to have to worry about SOS???

RH made this point...."(4) While I agree that there is some room for subjectivity, you have to be in the vicinity with the criteria for it to matter.  That's where Whitworth, Kenyon, CNU, etc. are losing out, because the gap between their SOS and the top-ranked teams is much, much greater than any difference in winning percentage."  Yes, the gap is much greater but does it make sense??  Even if concede that OWU has a better schedule than Kenyon (which I do concede but barely), is the gap really a .494 to .564 gap?  I don't think so, and in this case the W-L % as a number is what it actually is.

[PaulNewman]

This is as much a question as anything, but RH, you keep raising what happens in comparing teams in different regions for who gets the next available Pool C on the board.  I'm not sure how they do that, and are you sure that it isn't objective?  And how do we know how a .545 in one region jives with a .545 in another, like in your Denison vs Wash U example.  I finally did get your Team X vs Team Y stuff but maybe that should involve an adjustment in their procedure as much as dealing with the numbers.

BTW, RH, I don't necessarily disagree with you about subjectivity, IF there is a way to better trust the "objective."

[Ryan Harmanis]

Yea, seems like we just need someone to bite the bullet and play around with the numbers to see what the multiplier actually should be.  I don't have the time right now, but could definitely look into it in the offseason.

[Mr.Right]

They will never change but the one reasonable change would be the multiplier for Home and Away and Neutral.

Home-1.00
Away-1.25
Neutral-1.15


IMO