College Football Playoff: How much do big wins impact Top 25 ranks?

Sometimes emphasized, sometimes diminished, margin of victory is impossible to escape in football. How does it seem to impact the College Football Playoff selection committee?

With only one week remaining in the regular season for most teams, it is coming down to the wire in the race to secure position in the College Football Playoff or one of the New Year’s Six bowl games. And the big question for teams as they try to jockey for position is just how much impact it might have to run up the score.

Sometimes it seems like the committee views such actions favorably. Other times, it seems like they are looking for any justification to dock a team relative to other teams with similar records.

Should a team look to run up as big a scoreline as possible? Or should they aim to impress the selection committee with their sportsmanship? It is a dilemma that has never really been fully answered.

Even this morning, as we use the space in this week’s Sunday Morning Quarterback to investigate the question, we will never be able to come up with entirely conclusive answers. That is because, unlike the Bowl Championship Series, there is no specific formula that defines what delineates a Top 25 team from another that might sit outside the rankings.

So getting into the mind of the committee is an exercise in correlations, and nailing down causation for the selectors will always be a murky proposition. That, however, does not mean we should refrain from trying to at least find some logic in the chaos and confusion.

Generally, one would expect teams with better records and rankings to have better margins of victory. They sustain fewer defeats that most heavily drag down averages, and when they do lose those defeats tend to be by closer margins. Their victories tend to be more eye-popping and substantial as well. That is by no means a hard and fast rule, but it tends to generally hew more true than not when evaluating the majority of top-ranked teams over time.

With that in mind, let’s look at the margin of victory for Top 25 teams in both the first five years of the College Football Playoff and the last five years of the Bowl Championship Series to see whether the selection committee is more prone or less prone to select teams because of (or despite) their margin of victory.

Evaluating the BCS Baseline

The BCS era was hardly a clear-cut and linear process. The formula for determining the top teams in the country altered over time, as the furor of the public led to changes in how margin of victory was emphasized or deemphasized by the algorithm.

When it first launched in 1998, margin of victory was explicitly incorporated into the formula through the inclusion of multiple computer rankings that included a margin of victory component in their calculations. After a few years, however, several of the computers were cycled out in favor of other algorithms that weighted margin of victory less heavily into their final values.

Throughout those first years of the BCS, there was near-constant tinkering as the general public complained about this or that team being included or excluded. The formula was jiggered to hew more closely to the human polls until, in 2004, human polls were given two-thirds of the impact over the system.

In 2004, the AP Poll served as the counterweight to the Coaches Poll; after 2004, when the AP opted to disassociate themselves from the BCS, the Harris Interactive Poll was created out of whole cloth to serve the function the AP Top 25 formerly served in the formula. The basic formula, though, remained in place for the remainder of the series’ existence through 2013.

Over the last five years between 2009 and 2013, that was the formula in place. The Harris Poll accounted for one-third of the points toward the final score, the Coaches Poll accounted for another one-third of the points, and a composite of six computer rankings filled out the final one-third of the average.

That is what we will be working with when setting the baseline for how to compare the College Football Playoff selection committee.

Margin of Victory in the Late BCS Formula versus the CFP Committee

In the BCS era, there was a conscious effort to deemphasize the impact of margin of victory on the decision making process. While there was no way to eliminate the implicit biases of individual pollsters in the Harris and Coaches Polls, the computer formulas used by the BCS were explicitly chosen for their methodologies that sought to dampen how much margin of victory skewed their rankings.

Twice in the final five years of the BCS, the top-ranked team that earned one of the championship game spots featured an average margin of victory that ranked outside the top 10 nationally. That has yet to happen during the College Football Playoff era, as the top seed with the lowest margin of victory ranking to date was Clemson in 2015 when they were ranked 10th with an average margin of 16.8 points.

Yet ultimately the BCS only went so far in reducing the influence of margin of victory. Not once in the last five years did a team with an average margin of victory ranking lower than 25th make the top six of the final BCS rankings. That has already happened three times in the first four years of the College Football Playoff era.

In some ways the College Football Playoff selection committee has been more willing to drift beyond traditional thresholds of excellence to see the value in teams that dealt with much narrower margins to get to a top-tier record. In other ways, though, especially when it comes to their top two seedings, the selectors have been less willing to look outside the box than the BCS did in the long run.

What we can see through the first four-plus years of the College Football Playoff, however, is a willingness to go beyond teams that merely win and win big to incorporate more teams that might not win by as big a margin but that bring other value to the table.

In either case, what you do not find are teams with low margins of victory getting into position for a shot at the national title. With the 43rd-best margin of victory in 2015, Big Ten champion Michigan State was the College Football Playoff semifinalist with the lowest margin of victory yet at just 8.1 points per game.

A year earlier, defending BCS national champion Florida State reached the inaugural College Football Playoff bracket with an identical margin. But that 8.1-point margin placed the Seminoles 37th in the country in the 2014 season.

 Dataset: Margin of Victory Data (BCS Era from 2009-2013, CFP Era from 2014-2018)

Final Takeaways

As was articulated earlier, it is dangerous to try to read causality into these figures. Whether the selection committee even looks at margin of victory in their final deliberations is unclear, given their lack of transparency in the methodology behind their selection process.

The College Football Playoff selection committee, at the very least, subconsciously considers margin of victory just in looking at teams’ records. Winning teams are inevitably more likely to have higher margins of victory by nature of losing fewer games that might drag down the average.

But it is hard to say that they have as much emphasis on considering the margin as the BCS seemed obsessed with. So much of the tinkering with the formula in the early years was meant to deal with reducing what seemed like an outsized impact that margin of victory averages had on the selection process.

Whether it actually had much impact, though, requires further study. By the time the structures were implemented that increased the human component to a two-thirds supermajority of the formula in the BCS rankings, margin of victory had been tamped down in the computer component even as emphasizing human votes increased its influence surreptitiously in that component of the rankings.

As it is, though, teams have no direct incentive to run up the score. At the same time, it doesn’t hurt a team to pour on the points. Especially when trying to differentiate themselves from fellow contenders with common opponents, getting a bigger win than a fellow challenger can prove critical in the anecdotal moment. But empirically, over time, there seems to be little indication that having a high margin of victory and a high ranking correlates to anything bigger than having a high winning percentage.