Tag Archives: NCAAM

Nate Silver’s Grantland-esque website FiveThirtyEight debuted today. It includes an interactive graphic (utilizing seven different predictor variables) featuring every team’s chances to reach every round of March Madness, including their odds of winning it all. How do those odds stack up to the current (as of 11:59 pm Eastern Time) odds given by Best of all, which teams make for the best bets, even if they are unlikely to win the championship, because Vegas is giving them even longer odds than they deserve? Find out below!

Positive Expected Value Bets to Win the NCAA MB Tournament

Team Sportsbook Odds-to-One Break Even Percentage FiveThirtyEight Percent Chance to Win Bet Expected Value
Arizona 8 11.11% 13.00% 1.89%
Villanova 30 3.23% 4.00% 0.77%
Ohio St 75 1.32% 2.00% 0.68%
Creighton 40 2.44% 3.00% 0.56%
Duke 20 4.76% 5.00% 0.24%
Michigan 35 2.78% 3.00% 0.22%
Kentucky 50 1.96% 2.00% 0.04%

For a bet of Arizona’s odds to be profitable (in the long run), it needs to cash 11.11 percent of the time; Nate Silver and his team estimate that the Wildcats’ true odds lie at 13 percent. That gap produces the largest positive expected value in the field. Which teams should you avoid putting money on to go all the way?

Worst Expected Value Bets to Win the NCAA MB Tournament

Michigan St 5.5 15.38% 6.00% -9.38%
Syracuse 18 5.26% 1.00% -4.26%
Iowa St 30 3.23% 1.00% -2.23%
UCLA 35 2.78% 1.00% -1.78%
Florida 5.5 15.38% 14.00% -1.38%
Wisconsin 22 4.35% 3.00% -1.35%
Wichita St 15 6.25% 5.00% -1.25%
Kansas 13 7.14% 6.00% -1.14%

Everyone loves Michigan St, and that is precisely why they are overvalued. The Spartans are good, and it is entirely possible they could win; it is even possible that Silver’s methodology has sold them short, perhaps by not accounting for Tom Izzo. But it is also true that at the five-and-a-half-to-one odds currently offered, the Spartans have to win 15.38 percent of the time for this bet to be profitable. Even if their true probability of a championship is around ten percent, or even 12 percent, it would still not be a good idea to put money on Michigan St. With the fabled winning streaks, Syracuse and Wichita St also make appearances on this list of worst bets in the tournament.

This is not to say that these teams are guaranteed to lose. But if you place bets with a negative expected value, while you may win one or two, over time you are guaranteed not only to lose, but to lose money.


There is one more NFL game this season (no, the Pro Bowl does NOT count), but March Madness draws nigh. Out of sheer curiosity, let’s see who are the worst referees in NCAA Men’s Division I Basketball.

How does one go about that? Not by watching all the games and recording every bad call (and non-call). Such a process is no good because it is subjective. And that is fine. Who wants to scrutinize thousands of calls in hundreds of games, even for just one ref?

To determine any ranking, the objectivity and truth of the statistics at hand are crucial. Having several people judge a ref’s calls and taking the average would be a start, but again, this is not feasible, especially on the large scale required.

StatSheet has recorded many college basketball statistics going back to 1997 (for men’s Division I games), including how many fouls a particular referee has called in every one of his games. Most useful of all, StatSheet provides a home-team/away-team breakdown of all of a referee’s called fouls. Home foul margin (say, for the 2013-14 season) measures how many fewer fouls a referee calls on the home team than the away team, per game.

You can see where this is going. There is ample evidence that throughout all sports, home-field advantage manifests itself not through the athletes playing better or worse, but through the referees (unconsciously) favoring the home team, or rather, the home team’s fans.1 The more fans, the louder they are, and the closer they get to the refs, the more they affect the refs’ calls.

Ideally, referees would make the exact same calls whatever the atmosphere. They try to remain objective, and they generally do a pretty good job. But some referees may be better than others, and fall prey to the fans’ impact less than their colleagues.

There could be other reasons why Referee A’s home foul margin is substantially more than Referee B’s. Referee A could have simply called more games in which the away team actually fouled more often than the home team. Or Referee A could have refereed one team at home quite a lot; if that team played substantially cleaner than their opponents, Referee A’s home foul margin might look worse than B’s, even though he is calling fouls just as objectively.

Over a large sample, and given that the NCAA schedules their referees a variety of teams, both at home and on the road (as well as across different conferences), such random differences should even out. Over lots of games, and lots of teams, and lots and lots of fouls called (or not called), any differences in home foul margin are likely attributable to the referee himself.

This data comes from the previous five seasons of NCAA Men’s Division I Basketball, the 2008-9 through the 2012-13 seasons. It includes only referees who have done 50 or more games in each of those five seasons.2 Once again, home foul margin indicates how many fewer fouls a referee calls on the home team than the away team, per game.

10 Worst NCAA Men’s DI Basketball Referees, 2008-9 through 2012-13

Rank Referee Common Conferences Home Foul Margin (per game) Net Home Foul Margin (all five seasons) Games
1 Antinio Petty Atlantic Sun, SEC -3.10 -942.6 304
2 Doug Shows Big East, SEC -2.99 -1157 387
3 Rick Hartzell SEC, Sun Belt -2.97 -1210 408
4 Bernard Clinton A-10, ACC -2.65 -772.4 292
5 John Gaffney Big 10, Big East -2.63 -972.3 369
6 Bo Boroski Big 10, MAC -2.62 -1034.8 395
7 Lamar Simpson A-10, SEC -2.61 -781.6 299
8 Terry Wymer Big 10, MAC -2.61 -976.7 374
9 Les Jones ACC, Big East -2.57 -1122 436
10 Tom O’Neill Big 12, Mountain West -2.57 -1121.3 436

Among the 68 referees in the sample, the average home foul margin is -2. In the last five seasons, Antinio Petty has called one more foul on the road team than the home team, per game, than the average referee of his kind. This is not 100 percent conclusive; perhaps by chance, in his games the road teams really did deserve more fouls than in those games refereed by others. But it is certainly suggestive, especially over five full seasons. Likely, a hollering home crowd affects Petty’s judgement more so than others’. Likely, the above referees (unconsciously) not only favor the home team a bit, but a bit more so than other referees.

And here is the other end of the spectrum:

10 Best NCAA Men’s DI Basketball Referees, 2008-9 through 2012-13

Rank Referee Common Conferences Home Foul Margin (per game) Net Home Foul Margin (all five seasons) Games
1 John Cahill Big East, SEC -1.04 -420.8 404
2 Anthony Jordan SEC, SWAC -1.07 -305.2 285
3 Dwayne Gladden A-10, CAA -1.21 -390.6 322
4 Wally Rutecki MAAC, Big East -1.23 -408.7 331
5 Jeff Clark A-10, Big East -1.30 -421.2 325
6 Gene Steratore Big 10, Big East -1.30 -398.9 306
7 Pat Driscoll Big 10, Big East -1.42 -503.1 354
8 Mike Eades ACC, Big 10 -1.49 -676.5 454
9 Glenn Mayborg Big 10, MAC -1.50 -412.9 275
10 Sean Hull ACC, CAA -1.53 -482.7 316

Notice that all referees call more fouls on the road team than the home team. This may indicate that all referees are biased (by home crowds) or that away teams objectively foul more than home teams. Neither notion is ridiculous; both are probably true to some degree.

Was John Cahill really the best referee in college basketball the last five seasons? There are lots of different ways to answer. It seems he was the least swayed by the fans, the biggest distraction a referee faces in his job. These results suggest that Cahill kept his focus on what actually went down on the court, more so than any ref. Often in the same arenas, refereeing the same teams as Petty (they both worked in the SEC), Cahill was less susceptible to the crowd to the tune of two fewer home net fouls per game. What a guy.3

  1. See Scorecasting: The Hidden Influences behind How Sports Are Played and Games Are Won, by Tobias J. Moskowitz and L. Jon Wertheim, “Comforts of Home” and “So, What Is Driving the Home Field Advantage?”. 
  2. Why five seasons? Why 50 games? The number of seasons is mostly arbitrary. The number of games is a little less so. The busiest referees–the top two or three nationwide–tend to do about 100 games a season. Fifty is half that. These are the refs doing the lion’s share of the work, working at least half as many games as the referee who works the very most. 
  3. Apparently, Mr. Cahill retired at the end of last season. Road teams everywhere must be missing his fairness. 
%d bloggers like this: