Saturday, August 02, 2014

The Degenerates Guide to the 2014 Season: Losing to and Scraping by Lower Division Teams

The season is slowly skulking its way to the present. In the interest of preparing you dear reader for the coming season, I will be offering a Degenerates Guide to the 2014 Season. The posts over the next few weeks will emphasize the less savory angle of college football. We’ll openly discuss the point spread and hope the feds do not shut us down. My goal will be to prepare you to make prudent thought out bets either legitimately in Las Vegas, illegitimately with a bookmaker of your choice, or when filling out your weekly parlay card. This first post will deal with the ramifications games against lower division teams.

One of the consequences of expanding the college football schedule to twelve games is that quite often teams play foes from lower divisions (typically IAA or even occasionally Division II or Division III) to add a win to the ledger and bring in some gate receipts. Of the 128 teams playing IA football in 2014, 103 will play at least one game against a IAA or other lower division opponent. Sometimes, despite the challenges before them (less money, inferior players, playing on the road), these lower division teams will rise up and win a game. Perhaps the three most famous recent examples are Appalachian State over Michigan in 2007, James Madison over Virginia Tech in 2010, and Georgia Southern over Florida in 2013. Incidentally, Appalachian State and Georgia Southern are making the move to IA this season. While those are three memorable examples, lower division opponents have actually beaten IA teams 64 times since 2005! How have those teams that were beaten fared the next week Against the Spread (ATS)? One might hypothesize that a large upset, as is typically the case when a IA team loses to a lower division team, might lead the losing team to be undervalued the following week. Is this the case? Glad you asked. I looked at the 64 times a team lost to a lower division opponent and then examined their ATS performance the following week. In a single instance (Georgia State last season), the losing team played another lower division opponent the following week and lost. That leaves us with 63 observations. In those 63 instances, the team that lost to a lower division opponent went 30-33 ATS in their next game. Our hypothesis then appears to be incorrect. These teams, on average, do not tend to be undervalued by the public in the game following their loss as they have a losing mark ATS. So then, does it make sense to always go against these teams? Not at all. Betting against these teams would have yielded a winning percentage of .524, but this is exactly the mark one would need to break even when the ‘juice’ is added to the gambling equation.

What if we break it down further? How did these teams perform as favorites or underdogs in their next game? Teams that lost to lower division opponents were underdogs 48 times in their next game. Their ATS mark was 24-24 in said games. No value on either side. Perhaps not surprisingly, teams that lost to lower division opponents were favorites just 14 times in their next game. They went 5-9 ATS in those games. Shrewd readers will note than 48 plus 14 equals 62, which is of course, not 63. That is because in one instance, a team that lost to a lower division opponent was a pick ‘em in their next game (they won and covered by the way). There may be some value in going against teams that lost to lower division opponents in the rare instance when they are favored in their next game, but the sample size is relatively small, so I would only wager with your friends money.

When teams actually lose to lower division opponents, their ATS results in the next game are mostly random and provide no value, but what happens when they have a close call, but don’t actually lose? We’ll define a close call as any game that is decided by eight points or fewer (one possession). Since 2005, IA teams have played tight games with lower division teams 91 times. 90 times they had a scheduled game after their close call (we’re disregarding Wisconsin’s escape against Cal Poly in 2008 as it was the final regular season game for the Badgers). In those 90 games, the team that suffered a close call finished 57-33 ATS for a winning percentage of .633. For those of you who don’t know any better, that is a phenomenal winning percentage for a 90-game sample size.

Let’s dig a little deeper and see if we exhume any other relevant information. How do those teams involved in close calls with lower division opponents do when they are underdogs in their next game? In 57 instances, the teams that endured close calls were 35-22 in their next game (winning percentage of .614). What about as favorites? In 33 instances, they were 22-11 (.667 winning percentage). Enduring a close call, but not actually losing appears to precede a situation where an astute gambler can take advantage of public perception. Consider that in terms of actually winning their next game, teams that were underdogs won just 14 of 57 contests (.246 winning percentage). However, they covered the spread 35 times. This seems to indicate that the lines in those follow up games were artificially inflated because of the teams’ poor performance against the lower division foe. Similarly, the teams that were favored after their close call covered nearly as often (22 times) as they won overall (27 times) in their 33 follow up games. In these instances, the line was artificially deflated by their poor prior performance.

Why does this phenomenon exist? I’ll pose a few explanations, but I’d love to hear your thoughts.

1. There is not a lot of ‘interconnectedness’ between divisions of college football. Wisconsin and Ohio State face a similar slate of opponents because they are in the same conference. Wisconsin and Cal Poly probably face zero common opponents. The computer ratings used by casinos to develop initial point spreads may have trouble determining just how good some of these lower division teams are, and by extension how tough an opponent they represent. For example, North Dakota State has won the past three IAA championships. Certainly they are better than some IA teams. But where would they rank in terms of IA quality? Are they the 90th best IA team? 80th? 50th? 30th? It’s hard to say with such a small sample size against IA opponents and very little interconnectedness. I realize this doesn’t explain why when teams lose to IAA teams there is no predictive value about their performance in the next game.

2. The general public underestimates how good lower division teams can be. While the initial point spread stems from a computer rating system, the spread is adjusted based on public perception. Before they are released, spreads are calibrated to draw action on both sides of the wager, and after release they can move up or down based on significant wagering on one side. When a team has a close call with a lower division opponent, these are probably the thought processes going through the minds of most bettors: ‘Team A barely beat William and Mary. I can’t believe they are such a small underdog against Team B. Team B will mop the floor with them.’ Or: ‘Team A barely beat Eastern Illinois. I can’t believe they are favored against Team B this week. Team B will probably win this game outright.’

3. The stigma of almost losing to a lower division opponent is the same as actually losing. When teams actually lose to lower division opponents, their results in the next game are basically random. When teams almost lose to lower division opponents, the general public treats it as a loss. Thus, the spreads are artificially inflated or deflated.

Well, that’s all for this edition. Remember, if a team loses to a lower division opponent, stay away from them the next week. However, if they almost lose, they are probably a good play in their next game. Keep in mind, if you win any cash using this strategy, feel free to float some my way.

No comments: