A Statistical Look at the FCS Playoffs: Addendum

Welcome back to the most mundane part of the FCS Playoff series. In this installment, additional background, context, and reasoning are provided for the decisions made when putting this series together. This is strictly academic for those who want to learn more about the “in the weeds” aspects and includes no charts or tables. For reference, all five parts of this series are listed below.

Part 1 – 1978 Through 2023

Part 2 – 1978 Through 1992

Part 3 – 1993 Through 2009

Part 4 – 2010 Through 2023

Part 5 – Conclusion and Addendum

Deciding Which Measures to Use

This aspect doesn’t require much explanation but it is good to provide a basic answer to the “why” question. Why did we choose the measures we did? Honestly, it came down to wanting to answer basic questions about the history of the FCS Playoffs. How much do teams win by in the playoffs (margin of victory)? Do the final games become more competitive as the playoffs go on (margin of victory by round)? Does the home team have an advantage at home? If so, how much (home record and road vs. home margin of victory)? How do conferences perform? How do conferences perform relative to other conferences?

A lot of those questions have intuitive answers but intuition isn’t always definitive. Saying “home teams have an advantage” is true but how much of an advantage is it? If you look at it by round, we found that home teams have a massive advantage in the early rounds (72% or better) compared to the semifinals where it drops off to 65%. Being able to quantify the basic questions can help understand some of the larger picture issues with the FCS Playoffs.

One debatable aspect that was not included was the average score. There are two ways to measure this: average road score versus average home score or average winning score versus average losing score. There was no specific reason for not including these other than the margin of victory was more interesting to us.

Determining the Separate Eras

When reviewing the history of the FCS playoffs, there seemed to be natural eras that evolved. For example, the expansion of the playoffs from 16 teams to 20 in 2010 and then again to 24 teams in 2013 made a natural “expanded” playoff era. The difficult part was determining how to split the other three decades especially given the early years had fewer games compared to the years after 1985. Well… It wasn’t easy.

The original plan was to have four separate eras: 1978 – 1985, 1986 – 1997, 1998 – 2009, and 2010 – 2023. This had the main issue of a small sample size of games from 1978 through 1985 (just 60 games) but it would have isolated the very early years from the expanded eras. Splitting the 16-team playoff era made sense because of how football was played in 1986 compared to 2009. At the same time, it didn’t make sense because the era would be split arbitrarily in the middle.

There are other ways to break up the data into different eras. For example, it could have been split into four eras: 1978 through 1988, 1989 through 1999, 2000 through 2009, and 2010 through 2023. The four-era method would still have only a 105-game sample for 1978 through 1988, which is nearly three times smaller than the 2010 through 2023 era making comparisons more difficult.

Ultimately, the eras of 1978 through 1992 and 1993 through 2009 were chosen because of the huge influx of teams in 1993. The FCS (then I-AA) expanded by almost one-third due to the NCAA’s rule change for athletic programs in separate divisions. In effect, this was a literally different era established by the NCAA. It’s still fair to question how much difference there is between the way football was played in 1986 and 1993, especially compared to 2009 but there was no clear cutoff point in the 16-team era that would have made sense to use.

Each method of splitting the years into eras has pros and cons. Some may even think the expanded era from 2010 through 2023 is too long and that the 20-team playoff years should be kept separate from the 24-team era. In the next iteration of this series, it may not matter because enough time will have passed to have a natural demarcation point: the 16-team 2020-21 playoffs. Speaking of that season…

The 2020-21 Season

The 2020-21 season was like no other with a majority of games ultimately moved to the spring of 2021. There are three options on how to handle that season in this data set: exclude the season entirely, keep it within the expanded playoffs, or include it in the most recent 16-team era (1993 through 2009). Given the uniqueness of the 2020-21 season, excluding it is valid because playoff games have never been played in the spring, the high number of canceled regular season games, and the makeup of the playoffs/selection criteria was altered as a result of the scheduling cancellations. Then again, excluding the season solely because it’s an outlier is a poor reason.

Including the 2020-21 season within the expanded playoff era since 2010 is another valid option because – despite the uniqueness of the season – it still reflected the game of football in the expanded playoff era. Furthermore, the selection committee has evolved in the expanded era and the 2020-21 season would still somewhat reflect updated criteria compared to prior eras.

Finally, there’s the option to include it as part of a previous era such as 1993 through 2009. The main justification would be the playoff format chosen was identical to the format used most frequently. However, the 2020-21 season wouldn’t necessarily be similar to that era because how the game is somewhat different, many rule changes have occurred since then, and the playoff selection criteria have evolved.

Ultimately, the 2020 season was kept within the expanded playoff era from 2010 through 2023. The 15 playoff games played in the 2020-21 season represent approximately 5% of the total number of FCS playoff games played since 2010 (4.97%).

CAA and Atlantic 10/Yankee Conference Debate

One of the more debatable aspects of this analysis was how to treat the CAA given the unquestionable ties to the Atlantic 10/Yankee Conference. On one hand, it was clear that the CAA was basically a continuation of the Atlantic 10. On the other hand, it was more of a hostile takeover as the CAA began to poach football-playing A10 members. That eventually led to all 12 teams in the A10 joining the CAA beginning with the 2007 season with the CAA not taking over the charter but instead establishing its own charter. Consider that the Yankee Conference merged with the Atlantic 10 as a result of an NCAA rule change but that wasn’t the case with the A10 to CAA changes. Interestingly, the CAA “claims” the A10/Yankee stats in its record book despite the new charter established in 2007.

There was a similar situation that occurred with the Missouri Valley Conference and Missouri Valley Football Conference. Like the CAA and CAA FB, the MVC and MVFC are both separate entities despite their similar name. The MVC stopped sponsoring football after the 1985 season while the MVFC (then called the Gateway Collegiate Athletic Conference) started sponsoring football in 1985. The GCAC merged with the MVC in 1992 and was renamed to the Gateway Football Conference. However, the football charter remained with the newly named GFC and retained the records of the GCAC that began in 1985.

No records of the MVC football programs are kept by the MVFC despite some overlapping historical membership and the most recent MVFC record book makes it clear the MVFC began in 1985. For the reasons outlined above, it was decided to make the CAA a separate conference beginning in 2007 despite the roots being as far back as 1978 when the Yankee Conference existed.