by Ken Pomeroy on Tuesday, March 3, 2015
Add the Atlantic Sun to the list of conferences spiking the single-site format and putting all games at campus sites. This means the 818-seat Hodge Center at USC-Upstate will get to host a tournament game, so citizens of Spartanburg, get your tickets fast. (Though the Spartans drew 866 to their season opener, so if you really want to go, they’ll probably let you stand in the corner of the gym.)
by Ken Pomeroy on Monday, March 2, 2015
The Horizon League features a surprise top seed as presumptive preseason-favorite Green Bay was caught by Valparaiso, who won eight of its last nine to beat the Phoenix by a game. The two are rated almost identically, but the top seed is rewarded with home court, so the Crusaders are the favorite to snag the auto-bid.
The Patriot League tips off tournament basketball Tuesday night and with that begins a series of posts with tables featuring numbers between zero and one. We’re using the low-tech log5 method and my ratings to estimate each team’s chance of winning its conference tournament. From now through next Wednesday, you’ll see periodic posts breaking down the bracket for each of the 31 conference tourneys. (Or 32 if Harvard and Yale work out an unlikely tie atop the Ivy League.)
by Nic Reiner on Sunday, March 1, 2015
My favorite aspect of win probability models is the concept of leverage. As Ken Pomeroy wrote in 2010, Leverage measures how much is at stake on a particular possession.
I’m especially interested in how leverage functions as an unorthodox stand-in for excitement level and game tension in basketball.
Considering we now have win probability graphs for five years’ worth of NCAA tournaments, I thought it would be fun to examine which of the tourney games played over the past half-decade contained the largest percentage of high-impact possessions— and thus had the most high-leverage basketball. The sample spans 332 NCAA tournament games played from 2010-2014. Before we review the games, here is more from Ken’s introduction to leverage:
The colors [denoting leverage] range from blue, where win probability is largely unaffected by the potential outcome of a possession, to yellow, where the outcome of a possession can have significant impact on the win probability (more precisely, at least a 10% swing between a 2-point possession and zero points). Leverage is not based on what happened during the possession, but is the range of win probability based on what could have happened.
And here is a relevant excerpt from a piece I did on leverage last year:
By categorizing possessions within a game by leverage, we can distill the substance of a close game to its most granular level: the range of outcomes of a single possession. It is on these graphs that we are able to see which possessions were most important, even if they did not seem so at the time. 10+ percent swings in win probability are momentous and exciting to think about, especially when they’re happening at various, seemingly tranquil points in the game.
Below are the 15 NCAA tournament games (played from 2010-2014) with the largest percentage of high-impact possessions, listed in order from lowest to highest. If you have a subscription, be sure to also click on the win probability graphs for each game listed. It’s half the fun and they feature a lot of yellow.
by Ken Pomeroy on Friday, February 27, 2015
Programming note: This is the final Week in Review for the season. Conference tournaments start next week and probabilistic coverage of those events will take precedence until Selection Sunday arrives. Thanks to all for reading this season.
Here are the most extreme things to happen in college basketball between Friday, February 20th and Thursday, February 26th…
3) #202 Northern Illinois 84, #68 Toledo 82 (12%), Wednesday. I wasn’t one that necessarily liked the idea of 96-team tournament, but I didn’t hate it, either. It’s games like this that would suddenly matter. With a 96-team field you could conceivably reward things like regular-season titles and Toledo really jeopardized its chances of having the MAC’s best record by giving away this home decision to also-ran Northern Illinois. In the real world, this loss hurts the Rockets’ chances of winning the West Division and getting a triple bye in the MAC tourney.
2) #304 Niagara 82, #152 Canisius 71 (12%), Tuesday.
1) #312 Niagara 65, #140 Rider 61 (9%), Saturday. I don’t think one team has ever earned the top two upsets of the week, but the Purple Eagles, a species indigenous to the Niagara campus, did just that. The MAAC is a weird place where home court has meant less than nothing this season. Home teams are 49-52 in conference play heading into the league’s final weekend. And with these two road wins, Niagara is now 6-13 in the MAAC while getting four of those victories away from the Gallagher Center.
by Ken Pomeroy on Thursday, February 26, 2015
Welcome to Play-by-Play Theater, the very irregular feature where I mine play-by-play data from the past five-plus seasons to discover the wacky things that happen in darkest corners of the college basketball universe.
It was another epic night for #TeamBadLuck on Tuesday night. At one point, Florida missed seven consecutive free throws in the second half while seeing a six-point lead over Missouri turn into a six-point deficit. But was that really a big deal? The record for consecutive made free throws by one team is documented in the NCAA Basketball Record Book. I even wrote about it when it happened. Ah, those were the days when explaining the binomial distribution sufficed for insightful basketball analytics.
Nobody knows the record for consecutive missed free throws by a team, though. And because people were worse free-throw shooters in the ‘50s and ‘60s, it’s a safe bet that the true record occurred a long time ago, before people were producing play-by-play data. But we can check the play-by-play data that exists since the 2009-10 season, which includes over 31,000 games, to get a feel for what an exceptional streak of misses looks like.
by Ken Pomeroy on Friday, February 20, 2015
Here are the most extreme things to happen in college basketball between Friday, February 13th and Thursday, February 19th…
3) #239 Miami OH 67, #59 Bowling Green (8%), Wednesday. The regular season in the MAC features high stakes because the winner of each division gets a triple bye to the conference tournament semifinals. This year the chase for that honor is particularly interesting since the top of the MAC is stronger than it’s been in a while. There are five teams ranked in the top 100, but no clear leader of the MAC pack. Bowling Green, under first-year head coach Chris Jans, was in the driver’s seat for the East division before being stunned at home by Miami. The Falcons are now in a tie with Kent State and one game clear of Akron with five games left in the regular season. MACtion!
2) #294 Delaware 73, #89 William & Mary 70 (7%), Saturday. A Delaware win makes the list for the second straight week. The Blue Hens went to Williamsburg and handed the Tribe its first home loss of the season for the very unlikely season sweep. Colonel Ebirt sad face! This was the Tribe’s fourth conference loss and oddly, all of them were to teams in the bottom four of the CAA standings.
1) #344 Jacksonville 89, #131 USC Upstate 70 (3%), Saturday. This was the third-biggest upset of the season, behind NJIT over Michigan and Texas Southern over Michigan State, but look at that margin. This was a 16-over-1 level win that wasn’t close. Marcel White led the way with 30 points, which was unexpected considering White’s season-high to that point was 10. He didn’t even get on the floor in the previous game, and in the four games prior to that, he didn’t score at all. This was one of the most unlikely outcomes this season and it was driven by one of the most unlikely individual performances as well.
by Ken Pomeroy on Tuesday, February 17, 2015
I recently added an MVP feature to the box scores on the site, whereby the best player in each game is chosen using John Hollinger’s Game Score with a three-point bonus for being on the winning team. You could probably just choose the MVP yourself, but it means more when a computer does it.
One way to get a feel for which player is most valuable to his team is to look at who has earned the most MVP honors so far. Two players have earned 15 MVP’s each: Drexel’s Damion Lee and Northern Iowa’s Seth Tuttle. Drexel is 10-15, so Lee’s story is not going to be told here. But you should take a minute to visit Drexel’s page and check out the player stats, because Lee is having nice season without much help from his teammates.
by Ken Pomeroy on Friday, February 13, 2015
Here are the most extreme things to happen in college basketball between Friday, February 6th and Thursday, February 12th…
3) #297 Delaware 73, #127 Northeastern 68 (10%), Saturday. It hasn’t been a great year for the Blue Hens. Outside of senior two-guard Kyle Anderson—who has been around so long that I seriously thought he was drafted by the Spurs last June—the Delaware depth chart is littered with freshmen and sophomores. This was Delaware’s sixth win all season, but give them credit: They are feisty in close games. They are 6-18 overall, but 6-7 in games decided by six or less.
2) #338 Presbyterian 69, #208 UNC Asheville 65 (8%), Thursday. The Blue Hose started the season with a 69-point loss to Duke and while things have improved since then, it’s still been a tough season. Presbyterian is 9-17 with four of those wins coming over hapless non D-I competition. So that’s 5-17 against real competition. But the Hose, like the Hens of a similar color, are solid in close games, going 4-3 in contests decided by four points or less. The moral of the story is don’t let a weaker team hang around. Clutch can show up when you least expect it.
by Ken Pomeroy on Sunday, February 8, 2015
There has been recent discussion over at ESPN.com regarding the decline in home court advantage in the NBA. First, there was this piece by Tom Haberstroh, inspired by Steve Ilardi, and then additional analysis by Haberstroh and Kevin Pelton.
It turns out the same effect exists at the college level. Here’s the evolution of home winning percentage (conference games only) since the 2000 season: