Thursday, February 19, 2009

A Four Pack For Thursday

A look at the four most notable games on Thursday's schedule:

Minnesota at Michigan
Penn State's win over Illinois was punch in the stomach for Michigan, who was probably just starting to feel good about its at-large chances again. Now, even after their win over Northwestern last week, the Wolverines find themselves eighth in the Big Ten pecking order. There's no way the Big Ten is getting eight bids, and we aren't even confident that they end up with seven come Selection Sunday. To have any chance at an at-large, Michigan needs to get on major roll, and that roll needs to start tonight at home against the struggling Gophers. Minnesota has lost three of its last four (all on the road) and they've seen their seed fall from a 5 all the way down to an 8 as a result. The Gophers only have one quality road win in conference (at Wisconsin in OT back when the Badgers were playing awful), and overall they've lost four of their last five conference road games. That's music to Michigan's ears as they begin a tough five-game stretch to close out their season. They have three road games left, including a return date against these Gophers, and they also have to host a Purdue team that is playing some of its best ball of the season. If Michigan doesn't do better than 3-2 down the stretch, they are going to have to do some real damage in the Big Ten tournament, and hope for some other teams to start losing, to earn a bid.

Washington at UCLA
Just when we thought the Bruins were finally starting to live up to their preseason expectations, they went out and dropped two in Arizona, including a game against the Wildcats that they were never really in. They return home this week to take on the Washington schools, and tonight they need to beat the Huskies to avoid a season sweep. A win for the Bruins would keep them in the tight Pac-10 race, where five teams are currently within two games of first place. If Washington can pull off the road upset here, they would be in great shape to win the Pac-10 title. The biggest advantage the Huskies have over the other Pac-10 contenders is that this week's trip down to SoCal is their last road trip of the season. They do still have to play the Arizona schools at home, so by no means do they have an easy schedule left, but you would have to like their chances in those games since they have only dropped one game at home all season. If they go 1-1 this week against UCLA and USC, they'd have to be considered the clear favorite to win the Pac-10.

Washington State at USC
The Trojans dropped out of our bracket this week after losing three straight road contests. They are currently 6th in the Pac-10 pecking order, which is a bad place to be considering that it's the 5th rated conference and will likely not see more than half of its teams earn bids. This matchup against the middling Cougars is obviously a must-win to keep any tourney hopes alive. The Trojans did absolutely nothing OOC so a strong finish is required to get them back into the discussion. A 5-1 finish is within reason since they have four games left at home, where they have only lost once all season. If they can manage to win five of their last six, then they will be an interesting case on Selection Sunday. That scenario would give them an 11-7 conference record, which in most years would be a lock, and a top 40 RPI. Another win in their first game in the conference tourney would make them a lock. If the Trojans can manage to go 2-0 this week at home then we would have to take another look at putting them back in the bracket this weekend.

San Diego at St. Mary's
The Gaels' tournament hopes might be resting on what happens this weekend. But before they get a shot at Utah State in Moraga, they have to take care of business at home tonight against San Diego. The Toreros started off 4-0 in conference, but they have hit the skids in a big way ever since. They've lost six of their last seven games, including a 23-point home loss to the Gaels back on Jan. 22. Patty Mills had a big first half for St. Mary's in that game, but he won't be there to help out this time around. The Gaels are just 2-4 since their star guard went down, and if they don't beat the Aggies this weekend, their chance at an at-large, which is already in major trouble, will begin to look very, very bleak. If they slip up here against the Toreros, or lose any other WCC game down the stretch, those at-large chances would be near zero.

Also receiving votes: Duke at St. John's, Wisconsin at Indiana, Xavier at Charlotte, Western Kentucky at Arkansas-Little Rock


Eric Z said...

Given the results of last ngiht, doesn't it seem more likely now that the ACC only gets 6 in instead of the 7-8 that are expected?

Va, Ga Tech and NC St are out.
Maryland - barring 2 wins over Duke, UNC or Wake in the next couple of week - is probably out
Miami - at 4-8 - has to be out now.
And Va Tech, after their loss last night and their brutal schedule coming up - aren't getting in either.

That leaves 6 for the ACC.

The more this plays out - where the weaker bubble teams in the big conferences (like South Carolina, Southern Cal, Va Teach) keep losing - I think this paves the way for 2 bids from the MVC (Creighton) and maybe the MAAC if Siena doesn't win the conf tournament.

Kevin said...

I've always thought that the committee says that there is no quota per conference - that each team is evaluated independently of how many other teams from that conference are in. And that's how it should be, IMO. If Penn State, say, is deserving of an at-large bid, why should they be penalized by the success/failure of the other teams in their conference (other than the fact that it will hurt their strength of schedule, which is already factored in to their evaluation)?

But now, Bracketology 101, the best and most respected predictor of who makes the tournament says there's no way the Big Ten gets 8, and they probably won't get 7. Doesn't that imply that the committee has some sort of quota? Is the idea that there's no conference quota just a myth?

Anonymous said...

Re: Kevin's post on quottas.

My belief is that there is no minimum quottas, but there are certainly maximums. Example: Big East. There is a fair argument that the Big East should have 10 teams in, but there is no way that is going to happen. Even 9 would be stretching it, because it is almost certain that a team with a losing in-conference record would get in.

My thought is that in-conference records can be deceiving, and that one really should only look at a statistic of record minus in-conference losses to Tourney-bound teams. This sort of a statistic would focus attention on OOC record (wins and losses), bad in-conference losses and good in-conference wins. For bubble teams, I would argue that is all that really matters -- you are a bubble team, you are expected to lose against the best in your conference (because if you did not, then you likely would not be on the bubble), so lets look at how you did OOC and did you have any unexpectedly good wins or losses in conference. This stat would say that Georgetown is 14-4 (only non-eliminated losses are to Tenn, Duke, Seton Hall and Notre Dame and wins v Memphis, Maryland, UConn, Syracuse). Va Tech (just a fellow bubble team for comparison) would be 16-6 (losses to Duke, BC and Clemson would be taken out, leaving losses to Xavier, Seton Hall, Wisconsin, Georgia, Maryland and Virginia and wins v. St. John's, BC, Wake and Miami). Try the same with Penn St. -- (eliminate losses to Wisconsin [twice], Minnesota, Mich St and Purdue, left with record of 19-3 with losses to Temple, Rhode Island and Michigan, and solid wins v. Purdue, Mich St, Minnesota and Illinois).

This sort of analysis is particularly important when dealing with non-balanced schedules like those of the SEC, Big East and Big 10. When every team plays each other team both home and away, then in-league means much more since they are on a level playing-field. However, with unbalanced schedules, you can get a team that Louisville playing all Big East teams once except for South FL, ND and West Virginia, versus Georgetown who doubles up on Marquette, Syracuse and Cincinnati. LSU only plays Tenn, KY, SC and FLA one time each, but gets 2 against Auburn, Arkansas, Ole'Miss and 'Bama. Meanwhile, South Carolina gets 2 each against Tenn, KY and FLA. Same analysis in Big 10, ACC and Big 12.

Bracketology 101 said...

The ACC is looking more and more like a 6-bid league right now. We knew they'd never get 8 (like some bracketologists were predicting), but we thought they would eventually get 7. With VT and Miami losing last night, though, 6 is looking like a very good possibility, especially given the remaining schedule that the Hokies have.

As you noted, six bids out of the ACC is very good news for the Sienas and Utah States of the world. If those teams can win their Bracketbuster games, win out, and make it to the conference tourney final and lose, they'll have a better shot at a bid than they did 24-48 hours ago.

As far as conference "quotas" goes, the committee preaches every year that conference affiliation is not factor in to an at-large selection. Their mantra is that they select the "34 best at-large teams" and never consider what conferences those teams are from. Frankly, we find this very hard to believe, and we aren't the only ones who are skeptical. It's hard for us, when we do our bracket every Sunday night, to ignore the fact that we are putting in the seventh or eighth best team from a league, and it has to be hard for the committee, too. They know what each team's conference is, who is a big conference team and who is a mid-major, and they know they are going to have to answer questions if they put in eight Big Ten teams, for example, and just six ACC teams, when the ACC has been a stronger league, top to bottom, all year. It's also tough to look at a team that went 8-8 or 9-9 in conference play and put them in over a team, from another league, that went 9-7 or 10-8 without factoring in conference difficulty. The bottom line is, we think the "quota" system is very much a factor in picking at-large teams. The committee will never admit to it, because it goes against the idea of "everyone having a fair chance," but "quotas" are real and in the end they affect who gets in and who doesn't.

Bracketology 101 said...

In addition to what Chris said about quotas...In the few weeks before Selection Sunday it is useful to look at conferences and say that this league will likely get 5-6 bids but have no chance to get 7. As things play out in the final weeks of the regular season and championship week the big conferences generally lose bids. So even though today the Big Ten looks in good shape to earn 7 bids, there is still lots of basketball to be played and if we had to put money on it we'd predict that the conference ends up with no more than 6 bids.

Anonymous said...

A-10 groups are notDiablo 3 Gold appearing to get much difficulty acquiring video games in opposition to greater meeting groups ahead of conference engage in starting Guild Wars 2 Goldoff. And C-USA will not get Memphis to sign up for the BrackerBuster.