Anyone who has ever placed a bet against a point spread on a college football game knows that there are parallel games being played. One game is being played by two opponents to win. Another game is being played between a favorite and an underdog to cover points. The two games have different turning points. A game that is played for the win might come down to the last play. A game that is bet against the spread might be decided much earlier in the contest. Sometimes the action is concurrent. In college football sometimes the team that wins doesn’t win by enough, and that affects their postseason fortunes. The question is not really if Vegas affects college football polling, it’s really a question of how much.
From the opening game to the last conference championship two major polls frame a season-long debate. The Associated Press and The USA Today/Amway Coaches Poll rank the top 25. Unlike the NFL, whose postseason is decided by records and rules, style points count in college football. The College Football Playoff is determined by committee. The deliberation and judgment of 13 human beings decides which four teams get the privilege of playing for the championship. The Committee tries to outline its criteria as specifically as possible:
The committee selects the teams using a process that distinguishes among otherwise comparable teams by considering conference championships won, strength of schedule, head-to-head competition, comparative outcomes of common opponents (without incenting margin of victory) and other relevant factors that may have affected a team’s performance during the season or likely will affect its postseason performance.
Nowhere in this dissertation do you find a reference to point spreads or polls. But it defies human nature to believe that 13 people would not be influenced by collective wisdom from the college football public. The caveat “without incenting margin of victory” is very sportsmanlike, but not realistic.
The ubiquitous presence of betting lines cannot be denied. Nearly every preview and prognostication makes a reference to the point spread. Every weekend TV personalities, reporters, and bloggers engage in a media frenzy to boost their expertise credentials. Some people will tell you how to gamble, some people will post their predicted scores; but there is no way around the point spread. It hovers over every game as the most legitimate gauge for any given team’s performance.
In 2007 three economics professors Rodney J. Paul, Andrew P. Weinbach, and Patrick Coate published a study in Journal of Sports Economics. Their article was titled “Expectations and Voting in the NCAA Football Polls: The Wisdom of Point Spread Markets.” Using a complicated mathematical formula they tried to determine the most significant factors affecting the two major polls – the AP and (at the time) the ESPN/USA Today Coaches Poll. In their empirical study they found that point spreads and national television coverage were the two most influential factors during the 2003-2004 season. Times have changed over a decade. Television coverage is no longer an obstacle for top 25 contenders. Anyone with a reasonable cable package can watch pivotal games involving any of the top 25 teams throughout the season. Point spreads, however, may have increased in influencing public perceptions of top 25 teams.
Television coverage has changed in other ways. Networks are no longer afraid of gambling talk. TV has amplified the public discussion of point spreads. Early in this year’s season ESPN, the $7.3 billion partner of the College Football Playoff, was airing “cover alerts”during their broadcasts and College Gameday was openly discussing betting lines.
According to The Washington Post nearly $4 billion is bet on sports legally in Las Vegas yearly, and an estimated $80 billion to $380 billion is wagered illegally through a shadow industry of offshore online betting houses, office pools and neighborhood bookmakers. That’s a heavy price paid for contributions to public wisdom, and it’s not ignored. Paul, Weinbach, and Coate credit the phenomenon of the point spread:
The point spread is essentially a price determined in a large market. Research has shown that sports betting markets consistently produce point spreads that are optimal unbiased predictors of the outcomes of games.
Some might ask a legitimate question: Should point spreads affect the college football polls? The answer might be found when we compare the efficacy of financial markets to sports betting. Paul, Weinbach, and Coate write:
Prices in financial markets incorporate all relevant information brought together by a large group of agents with a vested interest in the asset or event in question. In his book The Wisdom of Crowds, James Surowiecki (2004) notes that speculative markets, such as the point spread markets in sports, are extraordinarily useful to society. These markets combine all useful information dispersed throughout society and ultimately result in an aggregated valuation that is generally superior to the valuations of any one expert alone.
The point spread is designed to promote gambling, but it has also proven to be a sharp indicator of teams’ performances. College football fans, whether or not they gamble, have to accept the influence of the point spread. Voters in the AP, and even the Coaches Poll pay attention to betting lines. We might not want to think we live in college football world where sportswriters and coaches are conscious of point spreads, but they are. Paul, Weinbach, and Coate explain the magnitude.
The impact of the point spread differential variable is the main focus of this study. Voters respond when a team performs in a manner different than expected. Empirically, when the score differential (favorite score minus underdog score) is different than the point spread, voters adjust their opinions of teams. Point spread differential was found to have a positive and significant impact on vote points in both polls. Teams that cover the point spread, thereby exceeding voter expectations, receive increases in vote points, because voters adjust their valuations of the teams upward. These results also show that if a team wins the game but does not cover the point spread, the failure to meet expectations will result in a decrease in vote points. The effects of performance relative to expectations will not only change the number of vote points a team receives but will also affect team rankings in the polls. These changes in rankings that occur could have a profound effect on bowl match-ups, the national championship game, and payouts to conferences and individual colleges.
The economics professors published these findings in 2007 based on 2003-2004 data. Sure, it’s one study, and every study has its flaws no matter how empirical. But common sense informs college football fans that the point spread affects the two major polls. Just last week Utah was a 6.5 point underdog to unranked USC. The Tojans smashed the Utes 42-24, and Utah dropped out of the top 10 in both polls. Clemson was a seven point favorite and destroyed Miami 58-0. They jumped to number three in the AP Poll, leaping TCU and LSU.
The BCS is gone, but the human element of the season’s most definitive poll is not. The College Football Playoff Committee can spew its platitudes and criteria. It won’t ignore the polls, and the polls don’t ignore point spreads. So, whether the high priests of college football like it or not, Vegas leaves its sticky fingerprints all over the Dr. Pepper National Championship Trophy.