No, the preseason AP Poll doesn’t alter the Playoff rankings

The Associated Press Top 25 Poll no longer plays an official role in determining college football’s national champion. That’s been the case since the AP asked the BCS after 2004 to stop using it to formulate BCS rankings ... and the BCS is long gone anyway.

But the AP is still the default ranking used by just about all media outlets, including SB Nation, until the Playoff rankings are released in midseason, and the AP is the closest thing to an official post-bowls top 25.

Considering the AP Poll’s long record — every week back to 1936, and every preseason back to 1950 — it’s a rare historical authority in a sport that often lacks consistent data. We can use it to figure out all sorts of cool stuff, like Oklahoma as the greatest team of the AP Poll era or the most overrated team from each of the last 68 years.

One concern a lot of people have with any preseason poll, though:

No games have happened yet. The same arguments (“Waste of time!” “They oughta ban preseason rankings!” “No rankings until December!” “Oh, well, if the very biased media has it all figured out, we should just cancel all the games!”) thus flare up every 12 months, on the dot.

So? Anyone who doesn’t want to look at them can just ignore them, right?

That’s not easy, since each team’s AP ranking will be right next to its name on the website schedule and TV scoreboard. You’d think these people would be used to this, though, since it’s been happening every year for decades now.

So some people are convinced that preseason rankings end up impacting your rankings later on.

It makes some sense. Confirmation bias alters human behavior constantly, so what if the Playoff committee is subconsciously steered by the polls?

We could look at AP Polls back to 1936 and try to figure out whether a team that starts out highly ranked will stay loftier in the later rankings only because it started out highly ranked. We’d have to find that these overrated teams aren’t overrated for any reason other than that they were overrated in August, though.

Luckily, we don’t have to parse that, because the AP no longer votes on the rankings that matter anyway.

In the Playoff era, we can just compare the committee’s rankings to the AP’s.

With four piddly years of data, let’s compare each year’s initial Playoff rankings with the AP Poll that came out two days prior each time. Each subsequent committee ranking will be a reset and even further away from the AP’s preseason poll, so the first committee ranking will be the one likeliest to suffer from AP contamination, if such a thing exists.

Alabama was AP preseason No. 2, and the committee had the Tide three spots below where the AP did in late October. Evidently, that preseason ranking didn’t impress the committee all that much.

The committee also dropped (so to speak) a preseason AP No. 5 Ohio State and preseason AP No. 13 LSU, and its most prominent boosts went to Ole Miss, preseason AP No. 18, and TCU, preseason AP unranked.

On this night, when the first-ever Playoff rankings dropped, the reaction was surprise. No one expected the committee to differ from the AP quite this much; the BCS rarely did. Looking back, it’s clear 2014’s preseason rankings didn’t move the needle for the committee.

The AP didn’t drop its preseason No. 1, Ohio State, until a week after the committee barged in and dumped the Buckeyes to No. 3.

The committee also dropped preseason No. 2 TCU and preseason No. 4 Baylor, showing not only that it does not care about preseason top-four teams, but also that it had no plans to apologize to the Big 12 for 2014’s Playoff.

The boosts into the top spots went to preseason AP No. 12 Clemson and No. 14 LSU. If the committee were swayed by preseason rankings, just leaving Ohio State up there would’ve been the move.

The biggest jump, Texas A&M, might have you yelling about SEC BIAS. But, first of all, it doesn’t show any preseason rankings bias (the subject of this review), since the Aggies weren’t ranked then. Secondly, take a look at the biggest jumps in 2016.

The committee’s most newsworthy jumpers were the Aggies, who again weren’t preseason-ranked by the AP.

But! The biggest jumps were by Penn State, Colorado, and Virginia Tech, all unranked in the preseason AP Poll, plus preseason No. 21 Oklahoma State. The AP had not gassed these teams up whatsoever, and yet the committee elevated them immediately. In cases like these, you can see the AP being slow to inch a team into the rankings, while the committee, which starts from scratch each week, can see a team with fresh eyes.

Among the teams the committee dumped was AP preseason No. 4 Florida State, which was getting a lot of credit from the AP for having two close losses and having wins over since-collapsed Ole Miss and Miami teams. The AP doesn’t have the time to reevaluate each team’s wins and losses in full, while the committee reviews not only your schedule, but your schedule’s schedule.

I think we see eye to eye by now. If the preseason AP Poll has so much sinister sway, then why did the committee start with preseason AP No. 15 Georgia ahead of preseason AP No. 1 Alabama?

From the four years we have, I think it’s pretty clear that the Playoff committee does not give a shit about preseason AP rankings.

Maybe a fuller study over a decade will reveal how newspaper rankings from mid-August alter the thinking of committee members in November, but it seems to be empty paranoia.

Understanding how the committee makes its rankings is key. It tosses out its top 25 each week; AP voters don’t have time to do that. AP ballots are due Sunday morning, often shortly after writers are done covering their own teams, so its voters often just review scores and slide teams up and down. The committee takes hours to argue its way through the entire board anew.

Therefore, the preseason AP Poll probably does not skew the season’s results in any meaningful way.

But is the preseason AP Poll still a thing people will scream at each other about like it’s the most important thing in the world?

Oh, hell yeah. So in that sense, it matters.