match
The National Collegiate Athletic Association men’s round ball tournament match starts in March of every year and is conducted over three long weekends. The event, commonly mentioned as March Madness, attracts a lot of interest and revenue. The tournament also creates a large betting market like other major sporting events. As per an American Gaming Association press release, Americans were projected to bet $8.5 billion on the 2025 event, with 4.1 million people betting legally, and another 7.6 million projected to be betting illegally. These figures are likely to grow even higher soon after the legalization of sports match betting in states other than Nevada.
There have been many tests of the efficient market hypothesis that employ sports betting marketing. Prior research investigates whether specific markets are efficient in general, as well as searching for conditions under which market prices are biased. Because they are the ones most likely to receive the most action, betting on the largest professional sports leagues – the National Football League, the National Basketball Association, Major League Baseball, and the National Hockey League – has been heavily researched. The two highest-revenue-producing sports of the NCAA – football and men’s basketball – have been studied as well.
Background and data
The format of the NCAA men’s round ball tournament match is single cancel. In 1985, the size of the tournament field was increased to 64 teams. These teams are selected and put in by a selection committee that pairs teams for each of the four regions. The 16 teams within each region are put in, with the 1 seed being the top team in the region and the 16 seed being the weakest.
These two teams face in the first round of games, with the 2-seed matching up with the 15-seed, and so on. In this way, the higher seeds in the tournament match begin with what is looked for to be the best chance to advance. The tournament bracket is fixed, and there is no re-seeding based on the result of each round. In the second round, the 1–16 game winner will face the winner of the 8– 9 game.
The first round of the tournament held all day on both Thursday and Friday of the first weekend, march madness bracket, march madness 2025 bracket, march madness, march madness 2024 bracket, march madness tournament consists of 32 games. The second round matches the 32 winners and consists of 16 games. In total, each tournament consists of 6 rounds and 63 games. The information for the analysis is obtained from the website www.oddsshark.com, which offers closing-line betting data for all of the NCAA tournament games played from 1996 to 2019.
The website states that the reported line is the sportsbooks. The data for this analysis consists of all the games played by differently-seeded teams in the tournaments during the period. Most betting market research utilizes the home team as the reference, but these games are hosted on a neutral field. Here the higher-seeded team is utilized as the reference to fit the analysis framework into a favorite vs. longshot setting. The higher-seeded team in each contest is, in theory, supposed to be the better team based on the selection committee of the tournament. There were 1512 games in the 24 tournaments, and 1494 of them are covered in the sample.
Testing overall market efficiency
Another method of thinking about match marketing efficiency is to look at how accurate the betting line is relative to the outcome of the game. A 0 constant and a coefficient of 1 for the Line variable jointly indicate that the market prices are correct, something that is true for the NCAA tournament match during the time under study.
In both of the models, the coefficient for the betting line is not significantly different than 1. In unshown results, I also fit a Profit model in which the outcome variable is 1 if the higher seed beats the spread, and 0 otherwise. As in the descriptive results, this does not provide any indication of inefficiency by the latest tournament round ball or by the direction or magnitude of the line.
Testing for bias by tournament seed
While very little research has been dedicated to the betting match market for the tournament, a significant amount has centered on other areas of the event. For instance, there have been some studies of the process of choosing the field and seeding the tournament.
Coleman et al., discover bias in both seeding in favor of major-conference teams and in choosing participating teams. They also discover bias in favor of teams that have a representative from the school on the selection committee.
Paul and Wilson revisit the problem and discover that the previous results of bias can be explained as an artifact of political correctness in the selection process. Although the official model for the selection committee did not vary with the margin of victory in regular season games, march madness 2024 bracket, latest madness tournament 2025, madness tournament 2025 update, latest match, including this in empirical models removes the indication of bias against major conferences or against teams that have representatives on the committee.
Stone and Arks discover that the selection committee fails to adequately factor in team momentum leading up to the tournament when making seeding decisions. The market does an exceptionally effective job in predicting the score of the more highly seeded teams in tournament contests with a mean error of just 0.05. The score of the less highly seeded teams is predictively low by about 0.5. Paul and Wilson revisit the disparity is possible because the market is in general more apt to be aware of the highly seeded teams, who are frequently members of more highly regarded conferences, played by higher TV exposure nationally, and so on.
Testing for bias by conference affiliation
Colquitt et al., in a college basketball betting match market study, find that the betting market is more efficient for better-known conferences. For instance, intra-conference games in the Southern Conference have more forecast errors than such games in the ACC. As mentioned before, the market in the NCAA tournament is unique in that it attempts to price the market for matchups between teams from 32 different conferences that may have no previous history with each other.
Rather than pricing a game between two teams in the ACC, say, march madness bracket the market is now attempting to price a game between a team in the ACC and a team in the Southern Conference. There could be a possibility that market participants are biased either for or against teams in the more famous conferences in such matchups since they might be better known to them.
Humphreys et al. address the potentiality that sportsbooks would rather not set and adjust betting lines to get fair money on both sides of a bet in the best college basketball market. Rather, they might utilize the biases in bettors to make a greater profit. In such situations, it may be achievable for an unbiased player to decide on a profitable strategy.
Conclusion
The NCAA basketball tournament match generates a big market for betting. The relatively high number of games compared to postseason play in other sports allows for observation of the efficiency of a market for betting that has an unorthodox amount of interest. One might make the argument that such a market would be extremely efficient, with such a large level of activity. Still, the market is also likely to attract interest from casual players. Such bettors will be less knowledgeable than players who wager throughout the regular season and are likely also to be more susceptible to bias from channels like past performances.
Based on information from approximately 1500 NCAA mens basketball tournament games from 1996 through 2019, this research has some evidence that the betting market for such games does function efficiently. This is in agreement with several studies that get this finding for betting on standard season games. The line to bet on in the tournament closely captures game outcomes, and I don’t find proof that heavily favored teams cover the spread at less than-optimal rates.
This second result is contrary to what has generally been demonstrated in earlier research on regular-season college basketball games Although the tournament injects uncertainty into the market by bringing together teams that are not familiar with one another, the data gleaned from the regular season and the heavy activity level seem to be the prevailing forces, leading to latest efficient market pricing.