A Numbers Game: Taking Early Lines in College Football Betting

Last week I took a deep dive into 2018 college football game totals for 679 games! Yes, I love stats so much that I’m willing to spend hours looking at numbers to figure out the story they tell.

What I learned from analyzing all that data – from a general viewpoint – is that the UNDER hit at a 54.2 percent rate for totals over 60 and below 49 (relative to Sportsbook lines).

What about closing lines, you say? This time, I wanted to use the same data points for those 679 games but look at the closing lines and compare the results to those I gathered last week. Here’s what I did. First, I eliminated all games that saw either zero line movement from open to close and games that had line movement of just half a point from open to close. That dropped the total sample size of 679 games to just 548 with line movement of 1 point or greater.

I approached this data with the intention of comparing Sportsbook lines vs closing lines and the book’s accuracy when setting totals. Instead, the first thing that popped out to me was the number of bets being placed on a particular side of the total – OVER or UNDER.

For the Love of (UNDER) Betting

If you follow me on Twitter, you know I love betting OVERs, so naturally I thought everyone else would too. My findings were very much to the contrary. Of the 548 games I studied, 318 (58 percent) had line movement that finished below the Sportsbook total, which in part was due to the public betting the UNDER.

There were some games that the public just nailed. It didn’t matter how low the total got because the game finished much lower than the original line opened at.

Some of these were:

  • Navy (10) vs Army (17), opened at 50, closed at 38.5
  • Massachusetts (22) vs UConn (17), opened at 74.5, closed at 64
  • UCLA (14) vs Cincinnati (24), opened at 65, closed at 56.5

The lowest-scoring games from the data set where the public bet the total down:

  • Charlotte (3) vs Tennessee (14), opened at 52.5, closed at 46
  • Arizona State (10) vs Michigan State (7), opened at 44.5, closed at 42
  • Michigan State (6) vs Nebraska (9), opened at 52.5, closed at 48
  • Georgia Southern (15) vs Texas State (13), opened at 54, closed at 52

Then there were games that had their totals bet way down but went wayyyy OVER.

Two of the more memorable games were:

  • Memphis (41) vs UCF (56), opened at 72, closed at 65
  • Michigan (39) vs Ohio State (62), opened at 59, closed at 53

Others were:

  • Georgia Southern (32) vs Minnesota (35), opened at 51.5, closed at 45.5
  • NC State (27) vs West Virginia (44), opened at 49, closed at 46
  • Florida International (42) vs Charlotte (35), opened at 56, closed at 45

How ’Bout Them OVERs

To my surprise, of the 548 games I examined, only 230 (42 percent) saw totals that finished above the Sportsbook line. Honestly, I was shocked by this. Call me a thrill seeker but I’ll take an OVER bet on a college football game any day. These numbers were fun to look at because there were some games that bookies couldn’t set high enough. The Sportsbook line started high, the OVER got pounded by the public and it still hit with ease.

One of the most memorable games was Oklahoma vs West Virginia. This line opened at 80! That’s a whole lot of touchdowns but no one was scared – obviously – because it got bet all the way up to 86.5 (where it closed). 6.5 doesn’t sound like TOO much line movement but betting up and up into the high 80s? That’s some risky betting. That said, it ended up being one of those “bet it and forget it” types of games with 115 points total scored. Sixty points were scored in the first half alone! It was amazing.

Oklahoma took that game, scoring 59 points to West Virginia’s 56 to secure a spot in the Big 12 championship with both Kyler Murray and Will Grier putting on a show. Frankly, it’s what college football is all about.

  • Other games that had no trouble covering the OVER were:
  • Arizona (28) vs Washington State (69), opened at 59, closed at 63.5
  • South Alabama (38) vs Louisiana-Lafayette (48), opened at 62.5, closed at 67
  • Wisconsin (47) vs Purdue (44), opened at 53, closed at 56

Then, of course, there were some that largely missed the mark. They were bet up and finished way below the closing total:

  • Maryland (17) vs Temple (20), opened at 63.5, closed at 66
  • Coastal Carolina (12) vs Kansas (7), opened at 52.5, closed at 54
  • Marshal (7) vs Boise State (14), opened at 55, closed at 58

Line Movement: Fade or Follow?

This was an interesting thing to look at. As I was sorting through my stats, organizing them the way I needed, I noticed it seemed more often than not that if a line was moving a certain direction, the opposite occurred.

Of course, I had to break it down to get a definitive answer.

Here are the results:

There were 230 games where the lined moved up from the Sportsbook total and 113 of those went UNDER.

Example: Louisiana-Lafayette (31) vs Louisiana-Monroe (28)

  • Sportsbook line: 68 (UNDER hit by 9)
  • Closing line: 71 (UNDER hit by 12)

There were 13 games that were bet up and hit the OVER on the Sportsbook line, but hit UNDER on the closing line.

Example: Vanderbilt (13) vs Georgia (41)

  • Sportsbook line: 52 (OVER hit by 2)
  • Closing line: 56 (UNDER hit by 2)

There were four games that either hit the OVER on the Sportsbook line and PUSHED on the closing line or PUSHED on the Sportsbook line but hit the UNDER on the closing line.

Example A: Rice (17) vs North Texas (41)

  • Sportsbook Line: 54 (OVER hit by 4)
  • Closing Line: 58 (PUSHED)

Example B: Arizona State (29) vs Oregon (31)

  • Sportsbook line: 60 (PUSHED)
  • Closing line: 66 (UNDER hit by 6)

If we remove the games that PUSHED and the few where if you were a genius and got the EXACT MIDDLE, that leaves 213 games total where the line direction moved UP with 96 of those games finishing below BOTH the Sportsbook and closing line. That means 45 percent of the betting market got the line movement incorrect.

Now let’s look at the opposite, games that were bet down but went OVER.

There were 318 games that were bet down. Of those games, 142 actually went OVER (45.3 percent) on the closing line.

Example: Air Force (32) vs Utah State (42)

  • Sportsbook line: 63.5 (OVER by 10.5)
  • Closing line: 66 (OVER by 8)

There were 20 games that were bet down and hit UNDER on the Sportsbook line, but went OVER on the closing line.

Example: Utah (24) vs Washington State (28)

  • Sportsbook line: 53 (UNDER by 1.5)
  • Closing line: 50.5 (OVER by 1.5)

There were four games that PUSHED on the Sportsbook line but hit the OVER on the closing line.

Example: Northwestern (29) vs Michigan State (19)

  • Sportsbook line: 48 (PUSHED)
  • Closing line: (OVER by 4.5)

As we did above, if we remove the games that PUSHED and the games that middled, then that leaves 118 games, 40 percent, that were bet down incorrectly.

What do the Numbers Mean?

If you followed the line movement during the 2018 college football season, you would have made some money.

Let me remind you that this is not the green light to bet blind. This is just a breakdown of numbers that may or may not be of any use to your betting strategy. Simply, it is for entertainment purposes and to satisfy a craving that I have to find some sort of answers to a broad idea or question.

Something to also consider is that books are smart. What works one season may not work the next. They adjust and adjust and adjust daily, weekly, monthly, yearly. There’s a reason they come out on top year after year.

This is, again, only for the 2018 college football season and does not include every single game played, so this is not a “lock” to go pound the UNDER if you see the line move in that direction. It used to be a thing you’d hear bettors say, “fade the public.” But as information – podcasts, articles, stats sites, YouTube videos, etc. – becomes more readily available, the public will gain a stronger skill set for picking not only totals, but spreads as well. We’ve seen it already in things like the Sportsbook SuperContest where the people at the top are hitting at 65-plus percent and the Sportsbook is finishing out the season with a 70.2 percent record through 17 weeks! INSANE.

I believe “fading the public” is now a thing of the past, at least more so now than ever. And if you recall how I started this piece, I mentioned how I noticed that if a line moved a certain direction, it appeared the opposite betting result occurred. Well, after breaking down the numbers, turns out that’s all it was – perception. I think of it like poker. You remember the bad beats more often than you do when you win a hand. The same could be said for sports betting. Your brain remembers when a bet is wrong more often than it would when a bet is correct.

So next time you want to generalize a trend, look at the numbers and let the stats give you the answer.