Feeds:
Posts
Comments

Posts Tagged ‘Western Carolina’

“Head-to-head” is a terrible tiebreaker

Posted in Rant, Uncategorized, tagged , , , , , on November 28, 2017| 2 Comments »

When sports decide which teams make the playoffs, there are often tiebreakers used to sort-out teams with identical records.  One of the most common tiebreakers is the “head-to-head” record.

Head-to-head record is a poor tiebreaker, especially compared with, say, strength of schedule.

Let’s NCAA basketball as an example.  Suppose Wake Forest (my alma mater) and Western Carolina (the school where I teach) both go 29-1 in the regular season.  Suppose Western Carolina’s only loss was to Mars Hill, and Wake Forest’s only loss was to Western Carolina.

Image result for western carolina basketball

The diehard “head-to-head” people would say “Western Carolina should be ranked higher!  They won the head-to-head matchup!”  But I say: what about the other 29 games?  Wake Forest plays in the ACC, and has multiple games against Duke, UNC, Miami, NC State…you get the picture.  Western Carolina has a much weaker schedule against the likes of Wofford, the Citadel, VMI, Chattanooga.  Sure, Wake Forest lost to Western Carolina, but in turn, Western Carolina lost to Mars Hill!  Which loss looks worse, hmm?

Image result for wake forest logo

The problem with head-to-head is that it’s only one data point.  Any team can lose a single game.  Maybe Wake Forest was missing their starting 5 due to the flu for that one game against Western Carolina.  (Clemson’s loss in football to Syracuse, earlier this year, was partially due to their starting quarterback being out).  A tie-breaker should be as broad as possible, and take the entire year into consideration.  This is especially true in basketball or baseball, when the long seasons make head-to-head records well-nigh meaningless.

I’m thinking about this now because college football is nearing its end for the year, and some arcane committee will have to decide who gets into the college football playoff and who stays out.  There are those who have complained that an undefeated Wisconsin team (currently #3) should be ranked higher.  But the Badgers have played an easy schedule, compared to, say, Auburn, who has faced the toughest road possible.  (Look at strength-of-schedule rankings here.)  In fact, I personally rank Auburn (10-2) higher than Wisconsin (12-0) for that very reason.  We only have 24 total data points here, but if Auburn and Wisconsin swapped schedules, then Auburn would be 12-0 easily and Wisconsin would be, at best, 7-5.

I can’t say it enough.  Head-to-head is a garbage tiebreaker.  It’s only one data point.  Strength-of-schedule incorporates literally dozens of data points and should always take precedence.

Read Full Post »

The NCAA should adopt a Swiss-system for college football

Posted in Rant, Uncategorized, tagged , , , , , , , on November 7, 2016| Leave a Comment »

Here’s something that will never happen, but it would be awesome:

The NCAA should go to a Swiss-system for college football.  And I don’t mean for the playoffs; I mean for the entire season.

First of all, here’s a brief primer on what the Swiss-system is.  I don’t think I can explain it better than the hive mind on Wikipedia, so here’s a quote:

“A Swiss-system tournament is a non-eliminating tournament format which features a predetermined number of rounds of competition… In a Swiss tournament, each competitor (team or individual) does not play every other. Competitors meet one-to-one in each round and are paired using a predetermined set of rules designed to ensure that each competitor plays opponents with a similar running score, but not the same opponent more than once. The winner is the competitor with the highest aggregate points earned in all rounds.”

Such systems are very common in chess tournaments, and also used in backgammon, squash, and eight-ball tournaments.  I’ve never heard of them being used in team sports, which is a pity.

Image result for bern

If I were Emperor of the World, here’s how I would implement the Swiss-system for college football.  At the beginning of the season, I’d rank the 128 FBS teams (teams that normally are bowl eligible) from #1 to #128.  (Well, I probably wouldn’t rank the teams personally, but I’d have a computer and/or a committee rank the teams much as the BCS does now.)  The great thing is that a ranking of #1 vs. a ranking of #5 (say) at the beginning of the season wouldn’t matter much at all.

The first week of the season, #1 would play #65, #2 would play #66, and so on.  For illustrative purposes, if we based seeding on the current NCAA rankings (as of Nov. 7, 2016), we’d have Alabama (#1) playing Southern Mississippi (#65), Michigan (#2)  vs. Texas Tech (#66), Clemson (#3) vs. Georgia (#67), Washington (#4) vs. NC State (#68), and so on, down to California (#64) vs. Florida Atlantic (#128).  Every higher-ranked team would be favored of course, but you’re going to get plenty of upsets: every one of the matchups I just (arbitrarily) presented would be a decent game.  Gone would be the days when an Alabama would play a non-FBS Western Carolina for their first game and win 49-0 to pad their resume.

Starting with week #2, things are already interesting.  Every week after the first, each team plays another team with the exact same record (if possible).  Continuing with my example, and assuming that all the higher ranked teams won in week 1, you’d already have on the table Alabama (#1) vs. Troy (#33), Michigan (#2) vs. Tulsa (#34), Clemson (#3) vs. Minnesota (#35), etc.  None of these games are cake-walks by any means (for perspective, the current records of Alabama, Michigan, and Clemson are all 9-0, but the current records of Troy, Tulsa, and Minnesota are 7-1, 7-2, and 7-2, respectively.)

Here’s the thing: starting with week 2, every single game in college football is a competitive game.  And starting around week 4, every single game is almost evenly-matched.  We’ve eliminated the all-too-common problem with the current system: that the top teams really only play 2 or 3 meaningful games a year.

Suppose we were using the Swiss-system, and we were making the matchups for the coming week’s games (Nov. 12).  What games would be on tap?  Well, there are currently 5 undefeated teams, which in a Swiss-system would be very unlikely after 9 weeks.  Just for fun let’s assume that it’s possible, but let’s ignore Western Michigan (no way they’d go 9-0 if they faced a few good teams).  With Alabama, Michigan, Clemson, and Washington all 9-0, this week’s marquee matchups would be Alabama vs. Washington, and Michigan vs. Clemson.  It’s likely that next week you’d have Alabama facing Michigan.  This, in early November!

The good matchups continue all the way down the line.  One-loss teams would all face each other, and you’d perhaps have games like Louisville vs. Ohio State.  Even at the bottom of the barrel, with a Rice playing a Florida Atlantic, the games would be evenly-matched.  This would be great for fans, because as it stands, when a Rice fan attends a game, they fully expect a loss; but with a Swiss-system, that same fan can be hopeful for at least a 50-50 shot at winning.

At the end of the season, an undefeated team would be almost impossible.  It’s likely you’d have 3 or 4 teams that were 10-2, and they’d all have already played each other.  That’s when a playoff would kick in.

For the playoff, we’d have the 4 (or better yet, 8) teams with the best records play each other in a standard elimination format.  At this point, it wouldn’t matter if they’d already faced each other in the regular season; rematches at this point would be desirable.  The great thing is that these teams would all be excellent teams.  In a Swiss-system, if you go 10-2, facing tougher opponents every single week, no one can argue you aren’t one of the best teams in the country.  Built into the Swiss-system is an important feature, which is that basically, every team at the end with a similar record faced a similar strength of schedule.

This is important, for in the current system, teams which are 12-0 can be left out of the playoff discussion if they’d didn’t play any good teams.  That’s never struck me as particularly fair.  If my team goes 12-0 and doesn’t get to the playoffs, then that means the team never even had a theoretical shot at making the playoffs to begin with.  What’s the point, then?  It’s a sordid fact that in the current system, there are only 30 or so teams that can ever even theoretically make it to the playoffs in a given year.  I’m sorry, Florida Atlantic, but if you go 12-0 next year you ain’t playing in a major bowl game.

There are obviously a few objections one could raise to my brilliant scheme.  Let’s address them.

  1. What about logistics? How in the world could you have teams flying around the country, facing each other, planning trips on only a week’s notice?  Well, as Emperor, it wouldn’t be my problem.  But in any case, it’s the 21st century for Xenu’s sake, so I think with some 747’s and the internet, it could be done.
  2. What about revenue? If Western Carolina doesn’t get to play and get crushed by Alabama, then Western Carolina loses out on some big time TV money!  OK, sure, but the games will in general be much, much more competitive, and many more fans will go to see WCU home games since they finally have a chance to win.  I really believe TV revenue would be up across the board.  We could even implement a TV revenue-sharing scheme, but that’s a topic for another day.
  3. What about rivalries? Well, what about them?  The current system doesn’t give a fuck about them in any case.  I can’t even keep track of who’s in what conference these days.  (Syracuse is in the ACC?  When did that happen?  They’ll always be a Big East team to me.)  With a Swiss-system, conferences become meaningless, and I say good riddance.

Image result for army navy game 1963

As for the handful of actual rivalries that still exist, and haven’t become jokes, such as Army vs. Navy (come to think of it, that has become a joke) or Alabama vs. Auburn, you could have the teams play an extra game in mid-December that doesn’t really count for anything.  If you think it’s unfair that a team has to play an extra game compared to other teams, well then, why not have every team pick a greatest rival that they have to play every year?  Hell, these games might even be the first games of the year, a sort of pre-season-type game, and the results don’t affect the Swiss-system per se, but the results might inform the seedings.  So Navy plays Army (a joke game, usually) but if Navy loses, we might initially rank them a little lower than otherwise.

Would any of this ever happen?  Not one chance in a million.  But it’s fun to speculate about.  And mark my words, in the universe(s) where I actually do become Emperor, initiating an NCAA football Swiss-system is one of my first magnanimous acts.

Read Full Post »