NCAA Basketball 2011: BCS Version – Introduction
Posted by nvr1983 on December 30th, 2010For the past two years we’ve taken a look at what NCAA basketball would be like if the powers that be decided to adopt the antediluvian BCS system. In 2009 it would have yielded a championship game between UNC and Louisville along with several other less desirable match-ups. In 2010 it would have led to a championship game between Kansas and Kentucky, which could have been an interesting match-up, but both teams showed severe flaws that led to their elimination well short of the final Monday night game in April.
This season we decided that we would expand things a bit by offering our RTC/BCS college basketball rankings using a formula similar to what they use to determine the BCS rankings in college football on a weekly basis. As the season progresses, you can see how certain teams rise from relative obscurity and into the BCS picture while other teams fall from prestigious BCS games down to what would be the equivalent of pre-New Year’s Day games. With conference play about to start we thought that this would be the ideal time to start looking at the potential match-ups.
We kept the same basic rules as we used in previous years:
- We are following the BCS Football guidelines as closely as possible, but we replaced the Notre Dame rule with the Duke rule since they both have sketchy TV contracts (Notre Dame with NBC and Duke with ESPN).
- The AP and ESPN/USA Today polls are used as the human polls and ESPN.com’s InsiderRPI, KenPom.com, and Sagarin’s ratings as the computer polls. We are not including six computer rankings and dropping the highest and lowest like they do in the BCS because frankly we are not familiar with six reputable computer ranking systems. If you know of any other computer rankings leave a comment below and we might include them in the next installment of our rankings.
- We used the traditional BCS calculations for determining each team’s score weighing the two human polls and the combined computer poll average as 1/3 of a team’s total score each.
Here are your initial college basketball BCS rankings:
Before we get into the potential BCS match-ups there are quite a few things worth noting.
General Notes
- All of the rankings were taken as of the start of the day on December 29th. They shouldn’t change too much at the top, but several of the computer rankings can shift quite dramatically from day-to-day even if only a few games are played. Having said that, the computer rankings are more up-to-date by a few days than the human polls, which are only released/recalculated once a week.
- The computer rankings are calculated similar to how they are in the BCS. If you are ranked #1 you get 25 points and if you are ranked #25 you get one point with a single point difference for each spot. If you are not in the top 25, you get zero point regardless of whether you are 26th or 345th.
- Even though we only listed the top 25, there were actually 39 teams that were ranked in the top 25 of one of the three computer polls (or eight less than the number of teams that 96 human voters sought fit to allot at least one top 25 vote to in the human polls).
- The computer and humans seem to agree with almost every team in the top 8, but after that the differences start to emerge (more on this in the subsequent sections)
Somebody or Something Loves You (The computers love these teams more than human polls do)
- Kentucky and BYU, which are both fall between #11 and 16 in the two human polls, are tied for 6th in the aggregate computer rankings.
- The computers also seem to think that West Virginia is really good. In fact, ESPN Inside RPI has them as the #5 team in the country. (More on this computer ranking system in a bit).
- Washington, which came in at #26 (just off the list), isn’t ranked in ESPN Inside RPI or Sagarin, but is #5 in KenPom’s rankings
Does The Computer Know Something That We Don’t? (Teams that the human polls love, but the computers do not)
- UConn still comes in at #7 in our BCS rankings after its loss at Pittsburgh (not factored into the human polls yet), but would have still been in contention for a top five spot if KenPom had them near where the other computers had them. Instead KenPom has the Huskies at #31.
- Minnesota (15th), Memphis (28th), and Illinois (33rd) are not in any computer top 25. The following teams managed to get into one of the computer top 25s despite not having a single human vote (out of 96 voters) to be in the top 25: St. John’s, Maryland, and Miami.
- Notre Dame, Kansas State, UCF, and Michigan State are not in in two of the three computer top 25s although that may change for the Irish after their win over Georgetown
- ESPN Inside RPI has Ohio State as the…Ok. We have to address this right now.
You May Have A Virus
Let us preface this by saying that we love all the coverage that ESPN provides for college basketball (some more than others). Having said that, their Inside RPI is atrocious. I know it’s early in the season so some wrinkles have to be ironed out, but this thing is so bad that I wouldn’t have put it on our site until the rankings seemed to make some sense.
- Ranking West Virginia is a little strange, but amusing…until you see that they have Ohio State ranked 23rd in the nation! I was going to name all the mediocre teams they are ranked behind, but it would have taken too much time. Someone needs to get Andy Katz or Doug Gottlieb on the phone so they can talk with the IT people at Bristol — they may have a Trojan horse on their server.
- Things only get worse with ESPN Inside RPI as you go down the rankings. I encourage you to do so, but warn you that you may never respect their ranking system ever again.
BCS Match-Ups
And now what you’ve been waiting for. . .the college basketball BCS match-ups:
- Rose Bowl: #3 Ohio State (Big Ten Champ) vs. Washington (Pac-10 Champ) – This one has mismatch written all over it. On one side you have the Buckeyes who would have made the Championship Game if it wasn’t for ESPN Inside RPI ranking them as the 23rd best team in the country. On the other side you have Washington, which didn’t even make the top 25, but by virtue of being the highest ranked Pac-10 team (like last year) ends up qualifying to play against a team that many consider to be the best in the country (assuming a certain freshman guard doesn’t recover from a toe injury). Look for Jared Sullinger to throw up some huge numbers here.
- Sugar Bowl: #9 Kentucky (SEC Champ) vs. #11 Purdue (At-large) – Even though this may not have a very highly rated team in it, this is probably the second best BCS game of the bunch (trailing only the Championship Game). We aren’t even sure where to begin with this one. Obviously there is a bit of a rivalry between these schools based on their proximity to each other (a little over 200 miles) and there is the whole SEC versus Big Ten thing. The story lines in this game go beyond that. You have an incredible match-up at the forward position with Terrence Jones going up against JaJuan Johnson and at guard with Brandon Knight or Doron Lamb playing against E’Twaun Moore. You have the (relatively) old Boilermakers with their senior leaders (Johnson and Moore) going against the young Wildcats. Basically you have a college basketball fan’s dream match-up.
- Fiesta Bowl: #5 Pittsburgh (At-large) vs. #8 San Diego State (At-large) – We’re going with Pittsburgh here (using the 2nd at-large selection as a replacement for Kansas, the Big 12 champion, because the Orange Bowl would get first pick since their traditional representative from the ACC — Duke — is the higher ranked team) because the two other potential Big East choices here both faltered in their last game (UConn against Pittsburgh and Georgetown against Notre Dame). The Panthers with their exciting guard play from Brad Wanamaker and Ashton Gibbs should make for a pretty interesting game. Pitt’s selection (with Syracuse claiming the AQ spot – see below) also means that the rest of the Big East teams including UConn, Georgetown, Villanova, and Notre Dame are playing in lower-tier bowls while Washington gets to play in a BCS bowl despite being unranked. For their opponent, we suspect that the Fiesta Bowl (read: BCS) would put San Diego State, a non-AQ that won the Mountain West and is in the top 12, here as the last at-large. Nothing against them, but this is a team that a major bowl would not want to have to do anything with because of issues selling enough seats. Putting them in a game in Arizona should mitigate that to some extent, as we would expect at least a group of college students and some loyal fans to make the 350-mile trek from San Diego to Tempe. The fans that do show up or watch on TV should be treated to an entertaining game and will be introduced to Kawhi Leonard, who should have a big game against the Panthers on the inside. What happened to BYU? The BCS only guarantees one spot for a non-AQ and that team has to win its conference and be in the top 12 or top 16 with an AQ ranked below them. Unfortunately for the Cougars, San Diego State would win the MWC in this scenario, meaning that the Cougars would have to get a BCS bowl to invite them out of the goodness of their heart. So have fun watching Jimmer Fredette in the equivalent of a pre-New Year’s Day bowl game.
- Orange Bowl: #4 Syracuse (Big East Champ) vs. #13 Missouri (At-large) – The Orange Bowl gets the first at-large pick here and we have to think they would take Syracuse. They bring a large fan base that includes many people outside of the local fans and alumni and have a big enough name with a marquee coach to draw people’s attention. Matching them up against Missouri should provide fans with an entertaining game (think of what Missouri and Georgetown showed us earlier this year) and both teams have big enough fan bases that this game should sellout relatively easily. It would also be a small victory for human voters as the computers don’t even think that the Tigers are one of the top 25 teams in the country. On the court you would have great match-ups with Marcus Denmon and Scoop Jardine on the perimeter and Ricardo Ratliffe battling Kris Joseph and Rick Jackson in the paint.
- BCS Championship Game: #1 Duke (ACC Champ) vs. #2 Kansas (Big 12 Champ) – We are not complaining about this match-up particularly if Kyrie Irving can come back in time to play against fellow freshman phenom Josh Selby, but we can already hear the Buckeye fans screaming about how ESPN rigged their Inside RPI to protect Duke from having to face Jared Sullinger. Even without Irving, there are plenty of big names that could be expected to step up here ranging from Kyle Singler and Nolan Smith for Duke to Tyshawn Taylor and the Morris Twins for Kansas. In any event, I doubt that the bowl would have difficulty selling enough tickets with these two fan bases.
I’m not 100% sure on the total methodology fo the Massey Ratings, but they’re out there and seem legit if you’re looking for another computer system.
http://masseyratings.com/rate.php?lg=cb
To really be consistent with BCS procedures, you should only include computer rankings that ignore margin of victory. The RPI is OK, as is the “ELO_CHESS” version of Sagarin’s ratings. A third option is available at the link below (calculated by a contributor at our site):
http://dbaker.50webs.com/cbb11_rank.htm
Explanation:
http://dbaker.50webs.com/method.html
While the non-margin-aware ratings are less reliable/sensible, using them would highlight the way the BCS goes out of its way to use less reliable information in the pursuit of political correctness.
P.S. You can blame ESPN for a lot of things, but you can’t blame them for the RPI. They’re just trying to replicate the official version the NCAA uses. Here’s a free version, BTW:
http://statsheet.com/mcb/rankings/RPI/2010-12-30
About those computer rankings…
1) Don’t blame ESPN for the RPI’s atrociousness. I don’t know how the InsiderRPI differs from the regular RPI, but I assume it’s based on the same concept (record + pop record + opp opp record). If so, it was bound to fail. If you’re looking for “reputable” computer rankings, I would avoid this.
2) I notice you’re using Sagarin’s default ranking, which combines his Elo Chess (i.e. wins only) and Predictor (most accurate for predicting future games) ratings. I’m not familiar with the BCS computer rankings, so I don’t know whether they skew more towards rewarding wins or trying to accurately predict future games, but if it’s the latter, I’d go with the Predictor rating. That one, by the way, has Washington 4th, about the same as Pomeroy.
3) A couple other computer rankings I’d suggest you check out are the ones from TeamRankings.com (again, they have different variations, so make sure to choose the one that most mimics the BCS philosophy) and the LRMC ratings (http://www2.isye.gatech.edu/~jsokol/lrmc/)
I would suggest using the Massey Ratings (http://www.masseyratings.com/rate.php?lg=cb). Massey has a football rating that is actually used by the BCS, and a quick look at the basketball ratings, at least the top 25, doesn’t yield anything too atrocious.
Any RPI rating will take time to become accurate as they are heavily biased toward strength of schedule. Once teams get into their conference schedules, only those with truly horrendous schedules will be left out. Right now however, Ohio state is being punished for the bottom of their schedule. Even though they have a couple of top games on the road (which is taken into account), they have played six games outside kenpom’s (Haven’t checked the RPI on this) top 200, which is twice as many as Duke or Kansas. To most of us, there is little difference between games against a team ranked 125 as opposed to one ranked 250, as most top ten teams will crush both, but there is a huge difference to computers. Once conference schedules get factored in, this will become a much smaller factor.
you could use one of the BCS computers… http://www.colleyrankings.com/
“because frankly we are not familiar with six reputable computer ranking systems”
If you are looking for reputable computer rankings, then good luck. Almost all of these except for Sagarian would not pass an intermediate statistics course. The methodology is that bad.
I think you’ve underestimated the huge fan base BYU has when you denied them a “BCS bowl bid.” The school has fans everywhere, and at least for football is the most attended non-power-conference team in Division I.