March 28, 2012

Schooling Schooler: The Problem with Polls

by Jac Coyne | | Coyne Archive | Twitter

Clemson has been bouncing around the polls this year, and currently checks in No. 11. Is there a way to help the MCLA polls achieve their mission? Jac Coyne suggests trimming it to a Top 20 while Nick Schooler says concentrate on the teams, not the records.
© Cecil Copeland

Nothing makes an MCLA fan's blood boil more than weekly polls.

Early in the season the rankings are viewed more as a novelty, but as the season matures and the polls are no longer linear – you'll have teams that have defeated higher ranked teams, but are still stuck behind them – the exasperation and bitterness ratchets up. Our boy Nick is a notorious poll troll who loves nothing more than an Internet forum flame session over the merits of one team's ranking over another.

The heightened awareness of polls in the MCLA is based on the league's history.

Up until a couple of years ago, the final poll determined who was going to earn a bid to nationals (and where they'd be seeded) and who was going to stay home. Old habits, like squabbling over rankings, die hard. Since the association has moved to a selection committee, the importance of the poll is somewhat diminished, although the final coaches' poll has been used as a guideline for seeding purposes (we'll see whether that policy continues this year).

There are issues this year that have roiled the base, most notably the lack of transparency, and I understand that gripe. But if there was one change I'd make in the MCLA poll – and it's something I've been advocating since the days – it would be to reduce the number of teams to 20 from 25. I'd do it for a couple of reasons.

First, give the coaches on the poll a break. Whereas there are about 23 teams that could legitimately round out the Top 20, the slots from Nos. 21-25 could be filled by about 50 programs. For the conscientious voter, filling in those stragglers is probably where the bulk of the time is spent on the poll. By eliminating them, you might have more coaches willing to do it on a weekly basis.

Frankly, if I were an MCLA coach and I was making roughly $1 an hour to coach my team, I'd just fill in the bottom five with teams in my conference so I could tell recruits what a strong conference we were playing in. That's probably why the weekly poll just has "The Next Five" teams receiving votes, because the laundry list would be so absurdly long if they punched them all in.

Second, it would increase the exclusivity of the poll, thereby giving it more weight. It also puts the league in step with just about every other division and sport that conducts a poll. Plus, the MCLA is only sending 16 teams to Greenville, so why list the next nine contenders? Is the association going to add an NIT bracket? Yes, there would be several teams deprived of tweeting "We're No. 24 this week!" but I think the MCLA would survive.

How about you, Nick? What's your biggest beef with the polls these days?

SCHOOLER: Aside from the glaring lack of transparency this year, the increased parity over the last few season has made it more difficult to rank teams based solely on team A beating team B. One must look at numerous angles. This has improved over the seasons with more games streaming online. However, what concerns me is that in certain cases, some teams that could make a legitimate run at the national championship are potentially getting overlooked, while teams that look good on paper are getting a free pass to the national championships.

There are three types of these teams, and they seem to make an appearance every season. I will discuss these types of teams and how I think they should be dealt with by voters and the selection committee.

The first type of team is the one who wins all of their games, but does not play a top-notch schedule. The perfect modern day example of this would be Texas. They are ranked No. 7 with a 4-0 record. The Longhorns have a one-goal win over Arizona State, a team that was ranked No. 2 at the time, but has since dropped to No. 6 (coincidentally higher than Texas, but it does not factor in their loss to Colorado on Monday) after failing to meet expectations. They also crushed LMU 17-9 in Austin.

The second type of team is like your bipolar cousin. They lose to weaker teams and win against stronger teams. They also keep in step with some of the best teams, but lose by a small margin -- better known in MCLA parlance as a "good loss." Colorado is an example of this type of team. The Buffs are currently ranked No. 18. In their first game of the season, they kept it close with No. 5 Chapman on the road, but turned around and lost by a lot to No. 19 LMU, a team unranked at the time. They had another "good loss" to Clemson, then went ahead and lost to unranked Arizona. They handled No. 22 Minnesota Duluth and followed that up with a victory over Arizona State on Monday night.

The last type is a team that had a drop-off from the previous season, but is hanging onto the success they experienced over past seasons. Arizona State is that team. They had high expectations coming into the season, but they lost to No. 7 Texas, barely beat No. 12 Michigan State, lost to No. 3 Colorado State, beat No. 17 Northeastern, lost to No. 5 Chapman, beat No. 22 Minnesota Duluth, and lost to No. 18 Colorado. So one might argue that they could be anywhere between No. 22 and No. 7, but they get the benefit of the doubt because of past success in the league.

These are the three classic cases that voters have to deal with. We see them every year, but how should they be dealt with?

In my mind, the poll should rank the best teams, not necessarily the best records.

Since the poll is being used by the selection committee to assist in picking the national championship field, it should reflect which teams have the best shot at winning the national championship. Don't reward the team that put together the cupcake schedule. Also, don't punish the team that had a few bad games paired with some good ones.

I am not arguing that Texas and ASU should be out of the picture or that they will not be contenders in Greenville. I am also not jumping on the Colorado bandwagon and riding them to the national championship. My argument is that these teams are on a level playing field, but circumstances have led voters to rank them very far apart.

In reality, these types of teams are "bubble teams," and how they finish off the season should determine who is the most deserving. All three teams have games against top 10 teams that could prove that they belong in the national championship field and not on the bubble. These tough contests are what separate them from non-contenders such as undefeated Kansas. The saving grave for these teams is that if you can't beat the best, at least you have a chance at the conference AQ.

As big of a "poll troll" as I am, I have found that complaining goes nowhere. It always seems to work itself out in the end. Aside from a few teams, most have been deserving of their bid to nationals, while the program or two that are left out could have done one or more things to get in. The selection committee and poll voters are very professional and take their roles seriously with no hidden agendas, and have been surprisingly transparent about the process in past years. They seem to get it right every time. In a season where any of the top 20 are deserving, let's hope they do it right this year.

I'm ready for playoffs to start, but before that, there are some big games this week.

To the games (where Nick is 24-16 and Jac is 22-18 and Coyne has closed the gap)...

No. 5 Chapman (11-2) at No. 1 Cal Poly (8-1) - Saturday 2 p.m. PT

COYNE: Cal Poly went on the road and beat the then-No. 1 team last week, so hosting the No. 5 team should be no problem, right? Think again. Part of it is the scheduling. Cal Poly has doubled itself down, playing San Diego State on Friday night while Chapman will come in fresh. In addition, Chapman is a different team than the one that lost to BYU in overtime and CSU up in Oregon. The defense has solidified and the confidence is sky high.

I know SLO is not a fun place to play, but Chapman races out to an early lead and parries a late charge from the 'Stangs. Panthers, 12-11.

SCHOOLER: Chaptown has been on a bit of a hot streak, winning their last seven games. Cal Poly just knocked off the top-ranked team for the second time this season. This will be an epic match-up between two well-coached, disciplined teams. This will be a great game between one of the best attacks in the country pitted against one of the best defenses, but the Mustangs finish on top, 11-9.

No. 8 Virginia Tech (10-0) at No. 15 Buffalo (3-1) - Sunday, 12 p.m.

COYNE: The Hokies' 4-3 victory over Georgia is a bit troubling, but sometimes you have to tip your hat to the opposition and it looks like the Dawgs did a nice job neutralizing Matt Giannelli and Kevin Hayden, who combined for a goal and an assist (both by Hayden). Can the Bulls shut down two quality guys like the Hokie's duo for 60 minutes? Considering the problems they had with Pitt's Tyler Novotny last weekend, I'm leaning toward no.

Buffalo does have some offensive options that will make Tech sweat, especially out of the midfield. Middies Kurt Stavdal and Alex Hultgren are extremely dangerous and will get their goals. Playing at home will be a big boost for the Bulls, but the Hokies are no stranger to road games. A late spurt lifts Virginia Tech, 11-10.

SCHOOLER: Yes, the Hokies are undefeated, but there have been some shocking scores coming out of those games. They have handled some and kept it close with others. Buffalo is just getting its season started. While I am still not sure what to make of the Hokies, I think it is about time for Virginia Tech to lose a close game. The luck can only last so long. Bulls win, 8-7.

No. 12 Michigan State (3-2) at No. 13 Boston College (5-4) - Saturday, 1 p.m.

COYNE: This is a monstrous game for these two programs. While both have a better than average shot at winning their respective automatic qualifying bids to Greenville, they both demonstrated last weekend that they could easily be thrown into the at-large pool. Sparty needed overtime to down their biggest competitor in the CCLA, Pittsburgh, and the Eagles posted a clunker in the 6-4 loss to PCLL foe Northeastern (and B.C. has Buffalo to worry about, too).

As such, the winner of this game will be a leg up on the other if the committee comes down to choosing between the two for the last pick, which is a distinct possibility. Normally, Boston College at home would be a slam dunk, but four goals against Northeastern has me spooked. I'm going with a less-than-confident pick, and taking Sparty, 8-7.

SCHOOLER: The PCLL is a better league than most people think. Boston College had a tough loss to a good Northeastern team, so Michigan State should not take them lightly. I have seen the Eagles play and they have some great players. There middies are strong and their defense is tougher. They just need to patch together a complete game. They will do that on Saturday. Eagles win, 13-10.

Coyne's Pick

No. 1 St. Thomas (5-0) at No. 3 Westminster (8-0) - Saturday, 4 p.m. MT

COYNE: The Tommies fly out to Salt Lake City with heavy hearts and without their head coach. The father of Pete Moosbrugger passed away during the week, and the UST head man is heading to Florida to be with his family. Judging by my past conversations with the Tommies about their leader, the players will be leaving it all on the field as an homage to Moosbrugger. That is certainly a check in favor of the visitors.

Another advantage for UST is they draw No. 20 Northern Colorado – a team that hasn't been terribly competitive with the top programs in MCLA-II this year – on Friday night while Westminster opens with an immensely talented No. 10 Concordia squad. While I think St. Thomas will be susceptible to an upset on Sunday when it plays Concordia, the factors mentioned above will be enough to see the Tommies through against the Griffins. UST, 13-11.

SCHOOLER: In a battle of undefeated teams, who do you chose? I have looked at all the angles and can't find a reason not to pick St. Thomas. I know Jac has some trick, but maybe he thinks I think he has a trick and really wants me to pick Westminster. I went against the Tommies once this season and it did not pay off. I won't do it again. St. Thomas wins, 15-10.

Schooler's Pick

No. 2 UC Santa Barbara (8-0) at No. 3 Colorado State (8-1) – Thursday, 7 p.m. MT

SCHOOLER: I am not going to act like I am totally confident that UCSB will win this game, but what kind of alumni would I be if I didn't pick them? Jac was happy when I asked him to throw a Thursday game into the mix because he thinks this will be a gimme.

He has a point. UCSB has not played a game in two weeks because they have been in finals. They have also had minimal practice time. On the other hand, Mike Allan and his coaching staff have had plenty of time to get a game plan together. But will they have enough time to implement and execute it?

I think they will, but they need to shut down any hope for CSU. The Rams are fired up after losing to Cal Poly. UCSB has not beaten CSU since the 2004 national championship game. Nothing would be better for the Rams than beating the Gauchos again, but that is not going to happen. UCSB wins, 10-8.

COYNE: The Gauchos are a good team, and should make some noise at nationals, but this is a terrible match-up. There are the X-factors such as limited practice during finals, playing at altitude and the Rams being loaded for bear after losing to Poly. It's also a stylistic mismatch. The Rams will be more than willing to play at the Gauchos' methodical pace, and it will mean big trouble if UCSB falls behind. I think they will, and CSU will be in control from the start. Rams, 11-7.

Schooling Schooler Archive
Week Seven: When Will Duluth Break Through?
Week Six: The MCLA's Internecine Snobbery
Week Five: Finally, the Buffaloes Stampede
Week Four: Spartans Make Their Stand
Week Three: Wither the Sun Devils?
Week Two: Nick's Bittersweet Weekend
Week One: Time to Feed the Ego

comments powered by Disqus

More Headlines