CBB: NCAA Tournament First Impressions
Posted by Neil Paine on March 15, 2010
Note: This post was originally published at College Basketball Reference, S-R's brand-new College Hoops site, so when you're done reading, go over and check it out!
Some random thoughts on the 2010 bracket...
- San Diego State over Tennessee looks like a tempting 11-over-6 upset pick. Tennessee ranked just 35th in Ken Pomeroy's ratings (our personal faves here at BBR, since they're based on offensive/defensive efficiency and the work of Dean Oliver), while SDSU was 40th. ESPN's play-by-play sim gave the slightest possible edge to Tennessee (51% to 49%), but this is close enough to be a great value pick because a 6-11 matchup shouldn't be this close.
- Another 11-over-6 I like? Washington vs. Marquette. Pomeroy had Washington 29th, and Marquette 28th; ESPN's sim actually sees UW as a pretty solid favorite. The only thing working against an upset is the pace matchup -- underdogs will typically play a slower pace than the favorite prefers, but Washington is a significantly faster-paced team than MU.
- ESPN's play-by-play simulation really likes Washington. If they make it past Marquette, it sees them as strong favorites against New Mexico as well. Then again, NM was overseeded by a pretty large amount; they ranked just 47th in Kenpom's ratings. Washington could be the 11 seed most likely to advance to the Sweet 16.
- Of course, there's always Minnesota: the Gophers (Kenpom #25) are facing Xavier (#22) in the first round, and ESPN's analysis of similar past matchups and simulation results favor Minnesota pretty solidly. Potential 2nd-round opponent Pitt ranked 30th in Kenpom and 32nd in Sagarin's Predictor (MIN was 21st), so don't be surprised if Minnesota is another 11-seed in the 3rd round.
- Super Cinderella of the 2010 tourney: ex-bubble team Utah State. Despite anxiety about getting into the field, they ranked 20th in Kenpom's rankings and will face Texas A&M (#23) in a 1st-round matchup they should be favored in (ESPN's sim has them at nearly 60% win probability). Round 2 would bring either an overmatched Siena team (Kenpom #58) or a depleted Purdue squad that's missing their best offensive player, Robbie Hummel. 12-seeded Utah St. actually looks somewhat likely to make the round of 16.
- BYU over K-State, anyone? KSU was 9th in Pomeroy's ratings... but BYU was 7th. BYU and their strong offense won 60% of ESPN's simulated matchups. K-State looks overvalued despite strong play in the B12 tourney.
- Other 1st-round "upset" specials: St. Mary's over Richmond, No. Iowa over UNLV, Georgia Tech over Oklahoma State (no bias here -- Tech is 27th in Kenpom, OSU is 44th), and Florida St. over Gonzaga
- Do I dare take Wisconsin over Kentucky in the Sweet 16? Everyone loves UK but the 32-2 Cats aren't as much better than the 23-8 Badgers as you'd think (if they're better at all). UK was one of the nation's luckiest teams in outperforming Pythagoras (20th biggest difference in the country), while Wisconsin was one of the unluckiest (235th biggest differential). Kenpom has Wisconsin 3rd in the nation, Kentucky 6th, and ESPN's sim had Wisconsin winning 54% of the time. I'm wary of Wisconsin after a similar pick burned me in 2008, but if you're feeling ballsy, the Badgers could help separate you from your pool opponents, all of whom likely took UK.
- BYU to the Final Four! Syracuse is 5th in Kenpom and BYU 7th, but 'Cuse could still be dealing with an injured Arinze Onuaku when this matchup comes around. ESPN's sim likes Brigham Young over the Orange to the tune of 58%. It's the darkest-horse Final Four pick that's still fairly reasonable.
- What to do about Duke? They've looked really good this year. They're #1 in the country by Pomeroy's ratings, #2 by Sagarin's predictor. In a hardly-threatening bracket, they should be one of the first teams you pencil in for the Final Four. But they're still Duke, ousted in Round 1 in 2007, Round 2 in 2008, and Round 3 in 2009, 2006, & 2005. I always get burned by these guys, but I can't help myself from picking them. They're a Final Four pick for me again, even though it goes against my better judgment.
- For all the marbles, I like Kansas. KU was first in Sagarin's predictor, and 2nd in Kenpom behind unreliable postseason performer Duke. They've only lost twice all year, and unlike Kentucky they weren't inordinately lucky in close games. I put my faith in Kansas in 2003 and was foiled in the championship game, but they rewarded me for picking them in '08, and it's time to go with the Jayhawks again. Rock Chalk!
March 15th, 2010 at 4:38 pm
What is the link for ESPN's simulator? SI's Stew Mandel picks BYU to the Final Four as well.
March 15th, 2010 at 4:55 pm
It looks like it's Insider-only, but the link is here:
http://insider.espn.go.com/insider/tournament/bracket/predictor
They show you a bunch of considerations like past seed differential performance (i.e., 5 seeds beat 12 seeds 68% of the time), past matchups of similar teams, and the play-by-play simulator (which I'm guessing is Accuscore or something like it).
BYU was really underseeded, btw, I guess because they lost 5 games (including in the MWC tourney) against a somewhat weak schedule. They had a great point differential, though. The Tourney is kind of a crapshoot every year, but there are a lot of good value picks this time if you have the courage to follow through with them.
March 15th, 2010 at 5:46 pm
You're talking to a die-hard BYU fan who has watched every BYU game, and most MWC games. I don't see much difference between SDSU and UNM (Aztecs were 2 OT points away from the 3-game sweep), and they're an 8 seed difference. Plus, what killed BYU (and helped UNM) was the BYU home loss. We didn't have our best player in the 2nd half (stomach virus), and we lost by one possession (Hobson blocking Hartsock). Is that one possession worth 4 seeds? Total crap. 3 of their 5 losses are by 4 points or less, and they don't have a single bad loss. Losses to only 3 teams - and they were 3 really good teams. 4 of the losses coming on their opponent's home court.
Thanks for the link.
March 16th, 2010 at 7:05 pm
Neil,
Great writeup. Was wondering what your thoughts are for strategy with more and more of these bracket contests where you multiply the seed by the round point level to get a score for a correct pick. For example, BYU that you mention above would be a great pick for the final four in my bracket since they are "expected" to make the final 4 27% of the time. Being a 7 seed they should be "expected" to garner 8.8 points (7*8*27%) where as a team like Syracuse would be expected to garner 2.3points (1*8*47%). So if you could play the bracket a million times you would come out ahead picking BYU, however 75% of the time you are going to end up with no points for that game. So should there be more weight toward still picking syracuse? I haven't seen many places give out thoughts on these types of scoring systems so i thought I would see what your thoughts were. Thanks.
March 17th, 2010 at 2:24 am
My strategy is to target the few top seeds I feel are the very best bets to win it all, and then take mid-seeded teams that were undervalued by the committee and have a good chance to win at least 1 and hopefully 2 games. So as I said above, Minnesota, Washington, and Utah St. are examples of teams that can help you maximize your expected points. Obviously you don't get to play out the bracket a million times like Sean did here, but the point-maximizing strategy over 1 million simulations will be the best strategy in any 1 simulation as well.
November 19th, 2011 at 12:38 am
I have observed that insurance companies know which objects are at risk of accidents and also other risks. Additionally , they know what form of cars are susceptible to higher risk as well as the higher risk they've the higher the particular premium rate. Understanding the basic basics of car insurance will allow you to choose the right types of insurance policy that may take care of your requirements in case you happen to be involved in any accident. Thank you sharing the ideas on the blog.