Home Field Advantage (HFA) for college football may be one of the more difficult values to quantify as there is no singular method to correctly assign an HFA value. There are countless approaches and all have their strength and all have their flaws.
You can take a look at stadium size, attendance and the preconceived perception of the stadium, but just because you have a huge stadium and a packed out house doesn’t mean you are going to cover the number. Tennessee has a stadium that seats 102,455 people and last year they had 93.5% capacity according to Phil Steele, yet they went 1-6 against the number at home last season.
You can look at past performance, but as Adam mentioned, there is so much turnover every year between players and coaches that the sample size college is extremely small and even then, there are still outliers. Take Oklahoma for example, looking at their results SU and ATS over the last 3 years resulted in an HFA of 4.5 with Adam’s method, the highest of any school. But the last 3 years all had Baker Mayfield under center and now he is gone. Should Oklahoma’s HFA be so high now that Baker is starring in Hard Knocks on HBO?
You can assign an HFA number based on the market and the season win totals. You can assign an HFA number based on your own experiences if you can. The latter might be my favorite method but it is a very expensive and time-consuming method. I currently have 19 college stadiums checked off and will be looking to add 1 more this year. If I keep this up, I could potentially have a first-hand HFA number for every school by 2094…if I’m lucky.
With these methods and others not mentioned above to go along with all the resources we have at our disposal, which method do I use?
The answer is all of them.
This season, I started by using the same method Adam mentions by assigning an HFA value based on performance for the last 3 seasons. Teams that perform well at home should be rewarded, regardless of their stadium or crowd sizes. On the flip side, why reward a team like Hawaii with a strong home field when they have only covered 2 games at home over the last 3 years(Ralph Michaels had some great insight on his HFA thoughts and Hawaii specifically on the BangTheBook podcast on August 7th if you missed it).
After talking with Adam, curiosity got the best of me and I calculated what the HFA would be using the same method but for the last 5 years’ worth of data. While doing this provided more data points, it normalized the HFA for a lot of teams. In the end, I scrapped it for this year because I think it reduced the edge I was creating and the same turnover problem that faces college programs is magnified even more.
So now I have an HFA number based on the performance of the last 3 seasons. The next step I took was to compare that number with the market’s season win total for a team. What I look for are signs that a team will regress or show improvement that could potentially drop or boost their HFA value.
A good example is Western Kentucky. Based on their performance for the last 3 years, their HFA is 4.0. However, their season win total is only 5.5. After digging a little deeper, you find they only return 3 starters on offense, they lose their multi-year starter at quarterback and the program is in the middle of a transition period with the head coach entering his 2nd year at the program. This is enough for me to drop them down to 3.0 for this year.
Arizona State, Boston College, FAU, and UCF are some of the schools where this came into play.
Next, I adjust teams based on perception (a.k.a the eye test) and my own personal experiences when applicable. For me, I adjust Mississippi State up a point. Performance results gave then an HFA of 3, but having been to Starkville for a couple games, their HFA could be 10 because of those cowbells. Imagine sitting on a plane for 4 hours and there is a kid kicking the back of your seat the entire time. That’s what it is like in Starkville for a game, only instead of a kid kicking your seat, the kid is banging a cowbell nonstop for 4 hours and there isn’t just 1 kid, there are 60,000+ kids. It is deafening and you can’t hear anything on the field. All of this happened during a day game too. It was a great experience but there isn’t enough Tylenol to fix that headache. After the flashback, I settled down and gave them a 4.0.
So now I have an HFA value ranging from 4.5 down to 2 for every team and this number is what I use as a baseline for the season.
As Adam mentioned, HFA values need to flexible and may change week to week for any given team based on the matchup and the time of the game. With this in mind, I went ahead and I assigned a minimum HFA and a maximum HFA for each school. Now there are plenty of schools (mostly Group of 5 schools) where their min/max HFA remains unchanged but in some cases, such as LSU where there are significant changes. I gave LSU a base HFA of 3.5, but I also assigned a minimum HFA of 3.0 and a maximum HFA of 5.5 (which is highest I am willing to go).
If LSU is playing a noon game against an FCS opponent (or Rice on November 17 this season), then I would give them an HFA of 3.0 for that game. If LSU is playing a big conference game at night and has the chance to replicate “The Earthquake Game” of 1988, then they would get 5.5 for an HFA value for the week.
Finally, I am finished. My end product includes a baseline, a minimum and a maximum value for HFA for all 130 teams.
Is this perfect? No. Is this accurate? Time will tell. Will any of these numbers change? Oh yeah. This may seem like overkill but there is some HFA value included in the line which means there may be an edge to gain and that’s enough motivation for me to put in the effort.