Monday, February 05, 2007

It Ain't Easy at the Top

A lot of elite teams went down this weekend, several to marginal bubble teams making one last run at an at-large bid. Let's get it started:

North Carolina State 83, #3 North Carolina 79
The epitome of what I'm talking about. Another hiccup for North Carolina, but no real reason for the Tar Heels to panic. They're only one of three two-loss teams in the ACC, but i don't think anyone considers Boston College or Virginia a serious threat to the regular season crown. UNC is without a doubt the best team in the conference. And on the plus side, Duke fell as well, so it's not like the Blue Devils have a real shot at the title, even if they win out. As for the Wolfpack, this keeps the fans hopeful for an at-large birth, but I wouldn't give them too much of a chance. Their 3-5 conference record isn't horrible, but they're 2-7 against the RPI Top 50, with an RPI of 105. That's not even close to a bid.

Boston College 80, #18 Virginia Tech 59
Clearly there is life after Sean Williams. After a few tough games they're back with a vengeance, and are in a great position for the second half of their conference season. At 7-2 in the ACC they have a good shot to finish in the top three in the conference, which will assure them a good seed. And speaking of teams that don't deserve a top seed, I hope I'm not the only one surprised at Tech coming back to Earth. Everyone jumped on the bandwagon after those two huge upsets, but this team just isn't that great. They still will probably make the Tournament, but don't expect them to seriously contend for the ACC title. They still have an RPI of 33, and remain fourth in the ACC at 6-3. They have enough winnable games to get the 9-7 record they probably need to get in the Tournament. A huge test will be the home-and-home with Virginia, another bubble team and their big rival. Both those teams could see themselves fall out of the Tournament if swept. Should be some good ball.

Georgia Tech 80, #21 Clemson 62
Just like NC State, this win keeps Tech's season alive. After a good start to the year, the Yellow Jackets had really stumbled with four straight losses that dropped them to 2-6 in the ACC and out of everyone's bracket. 3-6 isn't all too impressive either, but atleast it keeps them in contention. They have a surprisingly good RPI of 61 with wins over Duke and Memphis, in addition to this one. They have the resume wins, but just need to stop blowing the easy games. Speaking of a team with big wins and some blown easy games, what has happened to Clemson? Five losses in their last six games, it's like they forget how to win. Without a doubt they've played good teams - their other four losses were against Duke, Maryland, North Carolina and Virginia. But they're now 4-5 in the ACC with an RPI that has dropped out of the Top 20. Of course, if the season ended now they'd be in the Tourney. But if they finish 8-8 in the ACC, will the Selection Committee be willing to put them in? Tough question. Clemson needs to win five of their last seven conference games to put that issue to bed.

#8 Texas A&M 69, #6 Kansas 66
A gigantic win for a school without much history of basketball success. This team really needs big nights from Acie Law IV to win big games, and they've been getting those a lot this year. In addition, due to the unbalanced schedule, this is the only meeting between these two teams this year. The key will be the two games against Texas, starting with a huge one tonight. They can take one more loss and still win the Big 12 title, but a Texas sweep will do them in. A Big 12 title should give them a #2 seed, a remarkable achievement for Billy Gillespie.

Kansas State 73, #23 Texas 72
A huge win for Bob Huggins that validates the run that this team has been on lately. That's now seven straight wins, moving them to an impressive 17-6 and 6-2 in the Big 12. An RPI of 39 and a decent 6-6 vs. the RPI Top 100 has them in a spot where, if the season ended now, they'd almost definitely be in the Big Dance. Do they have the talent to keep this up? Maybe not. They still have a home-and-home with kansas, and another game at Okie State. I see a final record around 10-6, which will be enough. As for Texas, another huge day out of Kevin Durant, but he needs more support from his teammates if Texas is going to be a real elite team. Getting 32 points is great, but it means more when it's done through the flow of the offense, and not because the team just watches him do his thing.

Brigham Young 90, #25 UNLV 63
I'm not too surprised BYU knocked off the Rebs, but rather the gigantic margin of victory. This game was never close. They are now looking pretty good for March, with an RPI inside the Top 30. More importantly, this moves them ahead of UNLV in the standings. It's not that likely that the Selection Committee wants to give the Mountain West three bids, so it's imperative for these two teams to grab second place from the other. In fact, BYU can still think about winning the conference outright. They already beat Air Force at home, and the sweep would probably do it. As for UNLV, they continue to have a great RPI (11), but they are far from a sure thing. They are only 6-3 in the conference, so can probably only lose two more games. A win over Air Force in a few weeks would probably be enough to get them in, though.

Villanova 57, Louisville 53
This game was much more important for Nova than Louisville. The Cardinals are still in good shape in the Big East, but a loss here would have dropped Nova to 3-6 and with a possibility of dropping out of the Big East Tournament, let alone the other Tournament. With their good strength of schedule, and several good resume wins, they should have enough to get in with a 9-7 record. Their schedule gets easier, but not that much easier. It won't be that trivial to finish 5-2. Louisville, meanwhile, remains at 6-3 and in contention for one of the four byes in the first round of the Big East Tourney. In a lot of ways, they're the anti-Virginia Tech, with no big wins (an awful 0-5 versus the RPI Top 50) but without a lot of bad losses. Their RPI remains weak (62nd), but that should improve due to an increasing schedule alone. Besides, they still have South Florida, St. John's and Seton Hall at home. One more win brings them to 10-6, and it's hard to see the Big East getting denied a 10-6 team.

#22 Arizona 84, Washington 54
I don't know how many people noticed this, but entering this game the Wildcats had lost six of their last eight games and still had an RPI of 8th in the country. That has to be unprecedented. They've had a ridiculously tough schedule, so losing some games shouldn't be held against them. It's ridiculous that some people have dropped Zona to a ten seed, or worse. No way does this team get anything worse than a six seed. If they can finish strong and get something like 11 wins in the Pac-10, they could get as high as a three seed. They're still in sixth in the conference, but there's a real logjam ahead of them. A few wins and they'll scoot right back up near the top of the conference. As for Washington, they're now 4-7 and 8th in the conference. Add that to a 1-7 road record and an RPI of 75, and this team is pretty much done. Their last gasp will be this week against California, currently holding down 7th in the conference. A loss there will end the season officially.

Southern Illinois 54, Wichita State 46
Another big win over a tough opponent for the Salukis. I know that a lot of analysts still consider Creighton the class of the Valley, but no one has been more consistenly good than SIU. They don't have many marquee wins, but you can't ignore their RPI of 9. The longer they keep winning, the further they'll climb up everyone's bracket. As for the Shockers, another missed opportunity to establish the good conference record they need to stay in the Tournament. Their recent struggles have been compounded by the fact that those wins over LSU and Syracuse aren't looking as hot as they did a couple months ago. The RPI has creeped back up to 54th, and will continue to rise if they can improve their conference record to something over .500, which is what they need to make the Tournament anyway. I can't see a 9-9 team getting out of the Valley. That means the Shockers need four of their last five conference games. The good thing is that three of them are at home, where they've only lost twice this season. And one of their two road games is against Drake, which is very winnable. If they do blow one of those games, however, the season could come down to the season finale at Creighton. That could be a monumental game - for both teams, if Creighton is still in the running for the conference title. The key will be getting the end of their season started on a good note, by taking care of their next two games at home against Illinois State and Indiana State. They have the experience necessary to win the winnable games, with their top seven minuter-grabbers, scorers and rebounders as seniors or juniors.

Florida State 68, #10 Duke 67
A massive win for an FSU team hoping to avoid another #1 seed in the NIT. At 17-6 and 5-4 in the ACC, along with an outstanding RPI of 18th in the country, they have the numbers. But this gives them another marquee win, to prove that the early win over Florida wasn't a fluke. Even better, they are done playing ranked teams, knowing that four wins in their last seven games will get them in the Tournament for sure. As for Duke, this is the first time in years that they've lost this many home games in one season. They're still probably looking at a three seed in the Tourney, maybe a four, but this loss probably ends any hope at the ACC title. It's really possible that both Duke and UNC lost this weekend because they were looking ahead to their big matchup this week in Durham. Add that to the motivation both those teams will have coming off bad losses and we might have an outstanding game. Actually, UNC/Duke is always great basketball. I, for one, can't wait.


D. Moore said...

Louisville -- Their RPI...should improve". I think I'm realizing I've been misunderstanding how the RPI works. Does it really matter more who you play than how you do against them? Beating 4 top 150 teams and losing to 2 top 25 teams won't actually make your RPI go up from 61, will it? The frightening thing is that I think you really are implying that, and that you're probably right. The only other game before the conference tourney they have is GTown (top 50 RPI), and I'd predict the win because LVille is at home. How big a swing can one game make?

NC State -- If they continue on a run, people will scream extenuating circumstances because so many of their losses were without Engin Atsur. If the bubble is weak, it might be a valid argument.

GA Tech -- "this win keeps Tech's season alive". I always thought of the NIT as worse than death. Not a single road win all year long. Not even against Miami and Wake Forest. I know you weren't saying it was likely, but this team's just weak.

Jeff said...

Oh, I don't seriously think Georgia Tech has a real shot at the Tournament. I'm just saying that their fans can still have hope. But they pretty much need to win out - which they're not going to do.

As for NC State, they also need a big run to be seriously considered, and I just don't buy it. They really just had one big win (Virginia Tech isn't that good), so I need more evidence before I accept that they can win six of their final eights games, which is what they need for the 9-7 record they need to have a chance Selection Sunday. It doesn't matter who you've had injured during the year if they can't break .500 in conference overall.

And back to the first point, remember that the RPI is 3/4 strength of schedule. It's 1/2 the record of your opponents and 1/4 the record of your opponents' opponents. Louisville's strength of schedule should have a solid improvement from here on out. Their RPI really got killed over Christmas Break (they played five opponents between 12/20 and 12/30, all which an RPI worse than 150), and it's slowly improved as they've played their Big East schedule.

D. Moore said...

NC State--
Hmmm. I was about to disagree with you, and say that NC State doesn't have to go 9-7 in conference for the NCAA to overlook their mid-season injury slump. But I do think they'd need to go 8-2 in their final 10 games to get that consideration, and after looking at their schedule I realize that would put them at 9-7 anyway.

So, can they go 6-2 in their last 8 games?
Wins -- Miami, Wake
Losses -- UNC
TBD -- Maryland(2), GA Tech (away), VA Tech (home), FSU (away)

Yeah, the away games should do them in. But it was still a cool theory.

OK. I did the math (mostly because I don't want the RPI to work the way it really does), and Louisville's schedule isn't as good as it looks from here on out. The trick is that despite Pittsburgh's RPI of 3, and Marquette's RPI of 15, 4 of their 7 remaining opponents are really, really bad. Louisville's RPI today is at 59, and if you average the RPI of their remaining opponents, it's 78. They'll get a boost from the opponents' opponents portion of the RPI, but not from the opponents themselves. They've actually gotta beat people to move up.

This is the reverse of the trick that the MVC has used in the last few years to boost the RPI rankings of teams in their conference. If you want a strong RPI, you don't have to play great teams, but you do have to avoid playing really bad ones. If you can schedule road games against teams you might beat but not everyone else can (e.g., teams with an RPI between 75 and 150), you can have a shiny RPI come tourney time.

Creighton has 6 losses (out of 7 total) to teams like that, and they still have an RPI of 25, mostly because they only play 2 teams outside the top 150 all year. Louisville's already played 8 teams like that, and 7 of them were home games.

Jeff said...

You're right, the major fault with the RPI is how it rewards a team that beats two teams ranked 95th and 96th more than a team that beats a team ranked 25th and a team ranked 325th. It shouldn't be held against teams if they put in a few patsies as they get ready for their year (as long as they throw some tough teams in there as well). And teams should get more rewards for beating elite teams. I don't think a team like Creighton or Southern Illinois would go any better than 6-3 in the Big East.

But like I said, that's why rankings like the Sagarin Ratings are being used more and more by the Selection Committee. It's almost more worthwhile to look at those rankings than the RPI. For the record, Sagarin has the Missouri Valley as the 7th best conference, narrowly behind the Big East and well ahead of the Mountain West.

D. Moore said...

I have to disagree with you on part of what you are saying. Yes, teams should be rewarded more for beating an elite team, but I think the selection committee looks at that when they consider the marquee wins that a team has.

However, I love the idea that the RPI punishes teams for playing bottom feeders. I see this as the same as in college football, where games against division II schools don't really count in meeting the required # of wins for bowl eligibility. When Virginia plays Longwood next week, they have no more chance of losing that game than they do during a BYE week. Why should they get any credit for that win?

Also, although I don't think it's a particularly effective way of achieving this, how else can the selection committee encourage major teams to play lesser known teams outside their conference that might actually beat them?

Jeff said...

I'm not saying the committee should reward teams for beating up on Longwood. What I'm saying is that for an elite team with a good record in a good conference, they actually receive a substantial RPI hit for playing a couple bad teams. That's what happened to Louisville, for example. Their RPI would be higher if they hadn't played those games.

A team has a right to play a couple of bad teams to warm up during the out-of-conference season in order to get prepared for conference play.

That's why the Sagarin rating is superior to the RPI - it takes into account the score of the game. A team shouldn't be penalized for beating Longwood by 40 points. And they also should be REWARDED for beating at team like Bradley by 20. It's not only about who you play, but how you play in those games.

No system is perfect, which is why no one should take any one rating system as the gospel. Take a combination of several in order to properly judge a team. Which is, of course, what the Selection Committee does.

Jeff said...

I can actually give you a good example of what I'm talking about right now. I was just watching the RPI ratings for Wisconsin, because they had a game at Penn State tonight. They won by 13 points at Penn State, which is a respectable result. But the Nittany Lions are ranked barely in the top 200 in the RPI. Wisconsin's record is so good that another win doesn't really change their winning percentage, but the hit to the strength of schedule knocked them back. Their RPI dropped from .6589 to .6578. I don't think anyone should have their rating get worse for winning a BCS conference game on the road.