College Football Futures: NCAAF Championship Best Bets

College football is seen as less predictive than its professional counterpart, primarily due to the constant turnover from seniors to incoming first-year students. And there was a time that predicting a national champion was a bit of a crapshoot. Heck, look at how well unranked preseason teams fared from 1981-1990, with 3.5 national championships in 10 seasons: Clemson in 1981, Miami in 1983, BYU in 1984, and co-champion Georgia Tech in 1990. Nowadays, it’s quite different. Surprise teams seldom win it all in the modern era of college football. As Football Outsiders says:

“The strongest indicator of how a college football team will perform in the upcoming season is their performance in recent seasons.”

Compare odds from all major sportsbooks for the 2021 College Football National Championship >>

Knowing this, the best place to start in developing a sound national champion futures prediction is to “reverse-engineer” it. Look at the preseason odds for more recent national champions.

!function(){“use strict”;window.addEventListener(“message”,(function(e){if(void 0!==e.data[“datawrapper-height”]){var t=document.querySelectorAll(“iframe”);for(var a in e.data[“datawrapper-height”])for(var r=0;r<t.length;r++){if(t[r].contentWindow===e.source)t[r].style.height=e.data["datawrapper-height"][a]+"px"}}}))}();

*Based on College Football Reference Historical Preseason Odds

Here’s a quick way to visualize the preseason national championship odds. Note the “clusters.” Most recent national champions started with odds in the range of +400 to +2000, leaning heavily to the +400 side.

!function(){“use strict”;window.addEventListener(“message”,(function(e){if(void 0!==e.data[“datawrapper-height”]){var t=document.querySelectorAll(“iframe”);for(var a in e.data[“datawrapper-height”])for(var r=0;r<t.length;r++){if(t[r].contentWindow===e.source)t[r].style.height=e.data["datawrapper-height"][a]+"px"}}}))}();

Now put aside national champions for a moment. Over the past decade, something interesting has happened in this market. Look at the yearly breakdown of average payout/odds given based on “ranked odds” (i.e., “Mean 1-5” would be the five teams with the strongest odds/lowest payouts to the national championship in that season).

!function(){“use strict”;window.addEventListener(“message”,(function(e){if(void 0!==e.data[“datawrapper-height”]){var t=document.querySelectorAll(“iframe”);for(var a in e.data[“datawrapper-height”])for(var r=0;r<t.length;r++){if(t[r].contentWindow===e.source)t[r].style.height=e.data["datawrapper-height"][a]+"px"}}}))}();

The prices given for the top 10 teams—and the top five, in particular—have changed very little in the past 10 years. In contrast, the average pricing for the (least likely) bottom 15 teams have shot up significantly. The books are clearly overcompensating for a sport dominated by 5-15 teams and must offer increasingly lucrative odds for teams that realistically have no chance of actually winning at all (those outside the top 5-15 preseason teams by “assumed ranking”).

What does this mean to us? It’s a reminder to stay away from “sucker bets,” no matter how enticing. We are sharper than the typical bettor, we see what’s happening, and we’re not taking the bait. We’re staying within the top 15.

With this baseline understanding, let’s turn to 2021. Below are the odds provided by DraftKings Sportsbook on June 30:

!function(){“use strict”;window.addEventListener(“message”,(function(e){if(void 0!==e.data[“datawrapper-height”]){var t=document.querySelectorAll(“iframe”);for(var a in e.data[“datawrapper-height”])for(var r=0;r<t.length;r++){if(t[r].contentWindow===e.source)t[r].style.height=e.data["datawrapper-height"][a]+"px"}}}))}();

Rather than start with teams and raw numbers, see the chart below. It reflects the sportsbook betting odds along with PFF’s current National Champion probabilities. The result is each opportunity includes an expected value (EV). This simple mathematical formula helps measure the probability gap.

EV = (Potential Profit * Probability of Winning) – (Potential Loss * Probability of Losing)

!function(){“use strict”;window.addEventListener(“message”,(function(e){if(void 0!==e.data[“datawrapper-height”]){var t=document.querySelectorAll(“iframe”);for(var a in e.data[“datawrapper-height”])for(var r=0;r<t.length;r++){if(t[r].contentWindow===e.source)t[r].style.height=e.data["datawrapper-height"][a]+"px"}}}))}();

 

As you can see, we’re in a bit of a bind. Among those top 15 options that we know are most likely to claim a national championship, there are no obvious EV bets. However, we’ll keep digging to find the best value. It’s time to apply our strongest predictive tool: regression.

We’ve all heard the adage about history repeating itself. Frankly, all good predictive systems, be they stock trading, sports betting, etc., are based on this very notion. However, skilled bettors understand that the best predictive systems/processes are actually rooted in the idea that while history may repeat, it seldom repeats in the same way. History often repeats itself at the derivative or change. Say a team went 10-2 last year, yet some wins came by 1-3 points. In such a case, we know that with even the smallest variance (one fumble taking a different bounce, swirling winds altering a kick, etc.), this same team could have easily finished 8-4. Games (samples) less likely to be impacted by levels of variance (luck) are much more predictive. More simply put, we can identify teams that were lucky, unlucky, or as good as advertised.

Here’s the key: many markets and sports bettors still base most of their pricing/valuing on that “10-2 record” without accounting for how that team got there. That’s our sweet spot. Here’s where we can apply our understanding of regression towards the mean to maximize our likelihood of hitting while minimizing risk. So our next step is identifying teams with strong “positive regression forces” that could propel them in a good direction for the coming season.

Specifically, football has some powerful variables that tend to regress year over year (YoY) and have proven significantly predictive. In the chart that follows, you’ll see the variables we consider, along with the abbreviations. (Note: To standardize these numbers, each variable is “ranked” from 1-130, 1 being a team most likely to improve/boost wins in this category for 2021).

Pythagorean EXPECTED Wins Lost (netPythExpWinsLOST)

Pythagorean EXPECTED Wins Lost uses basic math to interpret data and predict how performance will change in the coming season. Essentially it applies the basic logic that the best team doesn’t always win and tells us which teams were lucky or unlucky. It weighs points scored, points against, and the ratio of those points to determine the number of wins a given team should have had. Looking ahead with this data can provide quality insights.

Fumble Recovery Rates (FUmRecR)

Forcing and recovering fumbles takes skill. So does holding on to the ball. However, once the ball is on the ground, there’s virtually no skill involved. Significant recovery rates, either way, tend to move 50% over time. Hence a team with a high recovery rate in year 1 will most likely see a lower recovery rate in year 2.

Offensive Yards/Point Scored (oYPP)

Yards per point (or YPP, as Phil Steele calls it) is an oldie but a goodie. The premise is simple: No matter where a team is on the field, the challenge to gain yards is the same. However, small gains can be more valuable than larger ones. For instance, a team that starts at its 20 and gains 30 yards gets no points. But a team that’s on the opposition goal line can pick up inches and get six points. This is a clear disconnect. Another way to look at it is if an offense is productive in terms of total yards but weak in terms of total points—whether that’s a result of turnovers, poor red-zone scoring, or a bad kicking game—may be undervalued.

Defensive Yards/Point Scored (dYPP)

Same as the oYPP, but for defense. We’ve heard the term “bend but don’t break” thousands of times. Really, it’s luck and defenses that allow big yards, but few points seldom pull that off two seasons in a row.

The Regression Big Board
This averages all the effects we’ve covered.

!function(){“use strict”;window.addEventListener(“message”,(function(e){if(void 0!==e.data[“datawrapper-height”]){var t=document.querySelectorAll(“iframe”);for(var a in e.data[“datawrapper-height”])for(var r=0;r<t.length;r++){if(t[r].contentWindow===e.source)t[r].style.height=e.data["datawrapper-height"][a]+"px"}}}))}();

Put simply, we want nothing to do with the sucker bets. So here’s that same chart, but with EVs added and applied only to our relevant teams.

!function(){“use strict”;window.addEventListener(“message”,(function(e){if(void 0!==e.data[“datawrapper-height”]){var t=document.querySelectorAll(“iframe”);for(var a in e.data[“datawrapper-height”])for(var r=0;r<t.length;r++){if(t[r].contentWindow===e.source)t[r].style.height=e.data["datawrapper-height"][a]+"px"}}}))}();

Only four teams have positive EV. You can forget about USC, which is likely facing significant negative regression. That leaves us with three “best bet” teams:

North Carolina
Florida
Wisconsin

Each is a long shot for the championship. But that’s not where we’re here for. We’re looking for the BEST BET.

Now here are the PFF odds of these teams winning their conference (obviously a critical need to reach the playoffs):

North Carolina – 18% (With only Clemson at 58% ahead of them)
Florida – 3% (With Alabama at 51%, Georgia at 24% ahead of them)
Wisconsin – 11% (With Ohio State at 48% and Iowa at 17% ahead of them)

Again, they’re long shots. Each team has a powerhouse in its way. But it’s time to apply some subjectivity:

I genuinely cannot see Florida rising to the top of the SEC (even if I am a huge fan of new QB Emory Jones, who graded out at 90.5 PFF grade in limited action). Even facing an “off” year, Alabama still rules SEC football. The Gators face the Tide early, and their schedule includes some other minefields, namely Georgia and, eventually, Florida State.

Though they avoid OSU in the regular season, Wisconsin still rates 25th in strength of schedule. Even if the Badgers navigate the regular season, I simply cannot see them getting past the Buckeyes in the conference championship.

Now to North Carolina. The Tar Heels’ strength of schedule is only 58th. They have a relatively easy slate of games, save for hosting Florida State and traveling to South Bend. Most importantly, they avoid Clemson in the regular season. Of course, should they advance to the ACC title game, the Tigers will likely be waiting. Clemson will have a good quarterback this fall, but he won’t be Trevor Lawrence. So the Heels may have a shot. Much of this thought process rides on Sam Howell. Entering his third season in this offense, Howell must excel, and he must stay healthy. Quarterback play is so important in modern college football. Over the past few seasons, so many teams have been propped up by their signal-callers. In 2021, North Carolina could be that team. At least, that’s my objective rationale.

Best Bet: North Carolina is my NCAA National Champion best bet.

Join the BettingPros Discord Chat for Live Betting Advice >>

Subscribe: Apple Podcasts | Spotify | Google Podcasts | Stitcher | RadioPublic | Breaker | Castbox | Pocket Casts

Whether you’re new to sports betting or a betting pro, our Sports Betting Strategy and Advice page is for you. You can get started with our 101 section — including 10 Sports Betting Tips for Beginners — or head to more advanced strategy — like Key Numbers When Betting Against the Spread — to learn more.

The post College Football Futures: NCAAF Championship Best Bets appeared first on BettingPros.