2023 College Football Preview, Part 8: Full Season Bad Betting Advice

At long last, the wait is almost over. On the heals of last weekend’s “Week Zero” appetizer, we are mere hours away from the main course of college football that will be Week One. 

Throughout the summer, I have conducted a math-driven deep dive into the 2023 college football season. I analyzed the preseason rankings, looked back at 2023, and taken a close look at the Michigan State Spartans, the Big Ten, the Power Five, Group of Five, and the postseason.

Today is the final stop on that journey. In many of the previous installments of this series, I have referenced various odds that my Monte Carlo simulation of the season generates. Usually, I use these numbers to predict how each team’s schedule, conference, and the full season will play out. But there is a another, potentially more interesting use for this data.

My simulation is not the only source of data that generates effective odds for the college football landscape. The various bookmakers in Las Vegas also generate odds for conference titles, total wins, and playoffs participants. If these odds are compared, I can get a sense of which bets might be more or less likely to pan out.

Today, lets explore this aspect of my simulation results, starting with an overview of the teams at the top.

The Big Picture

Table 1 below summarizes a subset of the data that my simulation generates, all of which I have presented previously, for the 30 teams that my model predicts to have the best odds to make the college football play offs. 

In addition, I have added the current money lines from DraftKings (which are of course subject to change), which translate to odds that can be directly compared to my odds. Finally, I compare these two sets of odds to generate a “return on investment” or ROI for a $100 bet on each event, assuming that my odds are correct. The bets with a positive ROI are shaded in green.

Table 1: Summary of the conference championship, playoff, and national title odds for the top 30 teams in the 2023 season. The odds from my simulation are compared to the odds from DraftKings in order to generate a potential return on investment.

The data on the left side of the table is presented in a slightly different manner that what I have shown previously in this series. In this case, it is easy to compare the raw odds that I calculate for each team to make the playoffs and to win the National Title. 

It comes as no surprise that No. 1 Georgia has the best chance to both make the playoffs (33%) and to win the National Title (11%). However, it should be noted that Michigan (29% playoff odds and 9% title odds) and Ohio State (25% and 9%) are right behind the Bulldogs in second and third place. The rest of the top eight is rounded out by USC, Notre Dame, Penn State, Alabama, and Florida State.

In my previous analysis of the College Football Playoffs, I came to a different conclusion using my disruptive scenario. In that analysis, I projected a Final Four of Georgia (No. 1 in Table 1), Notre Dame (No. 5), Penn State (No. 6), and Alabama (No. 7).

That projection is based on one very specific single season outcome (which I believe may likely share some interesting elements with the real season… assuming the teams are roughly as good as expected, which we already know is not true). The odds for each individual team in Table 1, however, are likely to be more accurate. 

My simulation puts both Michigan and Ohio State into the playoffs more times than Penn State. But the real season will effectively represent just one outcome out of an astronomically large number of potential outcomes. We will all watch together to see how things actually play out.

Shifting to the right side of Table 1, we can start to see which bets might be worth the risk. Not surprisingly, my calculated ROI for a large portion of the possible bets is negative, especially for the teams at the top of the table. Those clever folks in Vegas are experts at setting the lines just over where the probability is in reality. That is how and why they make money.

However, there are some potentially promising bets farther down in Table 1. For example, a bet on Wisconsin to win the Big Ten has a positive ROI. My computer give the Badges slightly better odds (13.7%) than the equivalent odds for a +750 money line bet (11.8%). While this is a small advantage, it does suggest that Wisconsin might be slightly undervalued.

Another promising set of bets might be on some of the borderline contenders such as USC, Notre Dame, or Penn State to win the National Title. The odds for those event are all around 5%. My calculations suggest that those three teams are all slightly undervalued as well.

Suggested Conference and Division Champion Bets

The information in Table 1 is a just a small sample of the potential bets that my simulation generates. It is also possible to estimate which specific bets look the most promising (have the highest ROI) in several different categories. Let’s now explore these potential bets, starting with the highest ROI wagers on division and conference champions, as shown below in Table 2.

[end preview]

Table 2: 2023 Conference and Division champion wagers with the highest ROIs, based on a comparison of the results of a 100,000 Monte Carlo simulation and the betting lines from Draftkings.

I should note here that “win the division” in the cases where the conference no longer has divisions (such as the ACC, AAC, Big 12, and Pac-12) translates to the odds that the team in question advances to the conference championship game.

Based on my analysis, the ROIs shown in this table are the potentially best bets on the board for 2023. A bet on Florida or Kentucky to win the SEC East has an ROI close to or over $200 on a $100 bet. That said, this would mean a bet against Georgia to win the SEC East. This seems like a long shot, which it is, but my calculations suggest that the odds are better than the folks in Vegas think that they are.

In fact, the majority of the bets in Table 2 are long shot bets with money lines well over +1,000. That said, there are a few bets here that stand out to me as being a little more likely, even if they promise a lower return. 

Bets on Ole Miss to win the SEC West (+2000, ROI = $134), TCU to win the Big 12 (+1,600, ROI = $113), Duke to make the ACC Championship Game (+1,800, ROI = $101), or Texas San Antonio to win the American Athletic Conference title (+650, RO = $87) all might be worth the risk.

Playoffs and National Title Bets

Table 1 above summarized some of the returns on investment for National Title wagers. Table 3 below give the full list of promising bets for teams to either make the playoffs or to win the whole thing.

Table 3: 2023 Playoff and National Champion wagers with the highest ROIs, based on a comparison of the results of a 100,000 Monte Carlo simulation and the betting lines from Draftkings.

These bets also tend to be higher risk and high reward, and in general the ROI is lower than the returns on conference and division championship bets. In general, the National Title bets seems risky. I am not sure that any mathematical calculation would convince me to bet on Texas Tech or Iowa winning the National Title.

The playoff bets, in general, appear more promising. I can imagine a scenario where TCU, Oregon State, North Carolina, or Ole Miss make the playoffs. The bet that really catches my eye, however, is a bet on Notre Dame to make the playoffs at +500. The ROI for this bet is $27, but my disruptive simulation fully endorses this wager.

Win Total (Over/Under) Bets

Generally speaking, betting on conference, division, playoff and national title outcomes is a high risk/high reward endeavor. A lower risk approach lies in betting on season win totals. Table 4 below summarizes the best over/under win total bets on the board according to my computer.

Table 4: 2023 Season win total wagers with the highest ROIs, based on a comparison of the results of a 100,000 Monte Carlo simulation and the betting lines from Draftkings.

In this case, I am able to use my simulation’s expected win totals and win distribution calculations to make the appropriate comparison to the win total money line wagers on DraftKings. The ROIs here are in the range of $20 to $37.

I don’t think that there are many bets on this table that stick out in particular, but I would stay away from betting the over on Northwestern at 3 wins. My computer didn’t read the news this summer out of Evanston. 

I should also note that my computer tends to like bet against the top teams to win 11 or 12 games. My computer calculates a positive ROI on the under for Georgia (O/U of 11.5 wins), Ohio State (10.5 win), and Michigan (10.5 win). 

The general computerized logic here is that there is a roughly 50-50 chance that each of those teams are not quite as good as the experts believe them to be in the preseason. That may be true, but even if it isn’t my disruptive simulation has Georgia going 11-1 and Ohio State posting only a 9-3 record. As such, betting the under on these teams might make sense.

Most Promising Big Ten Bets

For Big Ten fans, it is reasonable to wonder which bets involving Big Ten teams have the best potential payback. Table 5 below provides my computer’s answer to that question.

Table 5: 2023 Big Ten bets with the highest ROIs, based on a comparison of the results of a 100,000 Monte Carlo simulation and the betting lines from Draftkings.

Some of these bets appear on the tables above, but some of them are new. Interestingly, the most promising bets in the Big Ten are for a team other than Michigan, Ohio State, or Penn State to win the Big Ten East. 

The odds are long, but if there are any Michigan State fans (or Maryland fans) who want to bet on the Spartans (or Terrepins) to win the East at +10,000 (or +5,000). I say go ahead! My computer supports your choices. Honestly speaking, a bet on either team to win the Big Ten has a solid ROI, as does a bet on Penn State to win the East, which is much more likely to happen.

Table 5 also contains a few interesting win total bets that might be worth a look beyond betting on Ohio State and Michigan to underachieve. My computer likes Indiana to win more than 3.5 games and Rutgers to win more than 4.5 games. On the other side of the coin, my computer suggests betting the “under” on Maryland to win 7.5 games and on Illinois to win 6.5 wins.

Interestingly, my computer also reports a positive ROI for Michigan State to make the over on a win total of 5.5 wins. Ironically, the raw odds for this result are just below 50%, but the money line of +135 on this wager make this a good bet. 

I think that most Spartan fans will take that optimism.

We have reached the end of my annual preseason breakdown of the college football season. This is all the advice that I have to give before the season starts. The real fun is just around the corner.