Bid Stealers 2019

This is the time of year where the bracketology world becomes focused on bid stealers. There's almost always at least 1 bid stealer each year, contracting the bubble and taking away tournament berths from the at-large pool. Last year, both San Diego State and Davidson stole bids during championship week. This year looks likely to set a modern record for the number of bid stealers, so I decided to delve into each conference tournament to estimate the likely distribution of bid stealers. For all of the conference tournament simulations I used Bart Torvik's conference tournament simulators.

The Prime Suspects

Pac-12 (58.1% chance of bid stealer)

Washington and Arizona State are the only teams with legitimate at-large chances. Washington is safe barring complete chaos, but Arizona State is right on the nose of the bubble. I am classifying anyone from Utah on down as a bid stealer, and saying that Arizona State would be a bid stealer 20% of the time they win the at-large. This makes the Pac-12 the only league that is more likely than not to have its bid stolen.

MAC (48.2% chance of bid stealer)

Buffalo is an at-large lock no matter what happens in the MAC tournament. I have seen Toledo beginning to come up as an at-large in some projections, but I'm not seeing it. They don't have a Q1 win, and their NET is still only 60th. If anyone other than Buffalo wins this tournament, they'll have stolen a bid from an at-large candidate.

Atlantic 10 (43.8% chance of bid stealer)

Davidson and Dayton have flirted with at-large contention in the past month, but late losses have doomed either's case. They're both safely in the NIT, but about a dozen teams away from the NCAA field. The Atlantic 10 has been a major bid-stealer league in recent years, and if anyone other than VCU wins this tournament they'll be one as well.

Southern (42.8% chance of bid stealer)

Despite their hugely impressive 18-0 run through the SoCon, Wofford is still only 55% to win the conference tournament. Furman and UNC Greensboro are knocking on the door of at-large contention, but I think either team would need a win over Wofford to get there. Unfortunately for them, Wofford is on the other side of the bracket. I've assigned a 2% probability that UNCG or Furman win the auto-bid and don't need it, just to be safe.

The Possible Thieves

Ohio Valley (16.8% chance of bid stealer)

This is the hardest conference to estimate- while Murray State is almost certainly out of the at-large picture, I'm not really sure what to think of Belmont's at-large chances. Right now, I have them 7 teams above the cut line, indicating they'll be a First Four team if they lose to Murray State. If there are a lot of bid-stealers elsewhere though, it would hurt Belmont. I will say that Belmont is out if they lose to Austin Peay or Jacksonville State, and makes it 2/3 of the time they lose to Murray State.

American (15.8% chance of bid stealer)

Houston, Cincinnati and UCF are NCAA locks. If Temple beats Houston again in the semifinal, they're almost certainly in so I will not count them as a bid stealer. However, everyone from Memphis on down has no at-large hopes. Memphis gets the benefit of playing this tournament on their home court, helping out their chances of thievery.

West Coast (15.6% chance of bid stealer)

Gonzaga is obviously in the tournament (and the 1 seed in the west region) no matter what happens in Las Vegas this week. St. Mary's would need a win over Gonzaga to have any legitimate shot. If anyone other than the Zags wins this tournament, they're a bid thief.

Big East (13.9% chance of bid stealer)

This is the first conference tournament I can remember where every team has at least a 1 percent chance of winning it. The Big East has been very mediocre this year, with all 10 teams falling somewhere between 25th and 100th in Kenpom. Xavier, Georgetown and Butler are far enough from the bubble that I think they probably need the auto-bid at this point, while St. John's, Creighton and Seton Hall don't. Therefore, I'm classifying the 6-10 seeds as potential bid stealers.

The Unlikely Cases

Atlantic Sun (1.5%): I'm giving Lipscomb a 1 in 20 shot at an at-large if they lose to Liberty. Their loss to FGCU a few weeks ago really sunk any realistic chance they had.

Big Ten (1.3%): Penn State, Illinois, Rutgers, Nebraska or Northwestern would be a bid stealer if they won this tournament. Roughly half of this 1.3% comes from Penn State, who T-Rank really likes.

SEC (0.8%): If Arkansas, Texas A&M, Missouri, Georgia or Vanderbilt somehow cut down the nets in Nashville they'll be a bid thief.

Big 12 (0.3%): Oklahoma State and West Virginia are the only two Big 12 teams I currently project to miss the tournament, both have a slim chance of winning this event.

ACC (0.1%): The ACC has the lowest odds of bid thievery, as any potential bid stealer would probably have to get through multiple of Duke/Virginia/UNC.

Distribution of Bid Stealers

Given these probabilities, I simulated all of the conference tournament one million times in MATLAB to figure out the distribution of the number of bid stealers. Here is the probability of each scenario:

# Bid StealersProbability
03.43%
115.76%
229.41%
328.82%
416.14%
55.30%
61.03%
70.11%
80.01%
90.00%
Expected number of bid stealers: 2.59

There's roughly a 50% chance that we have 2 or fewer bid stealers, and a roughly 50% chance we have 3 or more bid stealers. This would be a modern record- there's no year that I'm aware of with 3 or more bid stealers.

The overall consensus right now seems to be that this is a good year for bid stealers, and this analysis backs that up- there's a slightly greater than 50% chance of having 3 or more bid stealers, and an almost 1 in 4 chance of having 4 or more. Bubble teams everywhere should be wary of the fact that the cut line is likely to move upwards in the coming week.

Comments

Popular posts from this blog

Bracketology 3/12/20

Bracketology 3/7/20

Bracketology 3/11/20