Don’t Take Power Rankings So Seriously

In the weeks that led up to the start of the NFL season, most of those rolling out power rankings didn’t think much about the Broncos, and two weeks into the season, some still aren’t ranking them that high.

But the one thing we must remember when we debate power rankings is that we need to ask everyone who puts them together these four words:

What. Is. Your. Criteria.

While a few who put together preseason rankings look at specific things that are based on past trends and statistics, the majority are doing them based on whims. And even those that put them together based on past trends and statistics might be doing them only because people want to talk about those things. (In other words, they’d rather not roll them out, but happen to contract with sites who insist upon those things because those sites want clickbait.)

I’ll talk specifically about two rankings in which each site specifies its criteria: Football Outsiders and FiveThirtyEight.

What Football Outsiders looked at is how the teams fared last year, the factors that went into how those teams fared and whether or not it’s realistic that the teams will perform at that level for the next season. Factors they try to weigh range from players lost in free agency to players who were injured last season but back this season to what can be expected at particular positions.

How that applies to the Broncos goes like this: Football Outsiders will note that when a team puts up a DVOA that suggests an “historically dominant” unit, that it’s not likely the unit will reach that level the following season. Please note that does NOT mean the unit will become a bad unit; it only means there’s a difference between being a great unit and being an “historically dominant” unit.

The perfect example is the 2013 Seahawks, whose defense had a DVOA of -25.9 (keep in mind a negative number is a good thing for defensive DVOA). The 2014 Seahawks DVOA was -16.8. Both years, the defense ranked first in defensive DVOA, but only in 2013 was it “historically dominant.” Thus, Football Outsiders assumed that the Broncos’ defense, which was -25.8 DVOA in 2015, would go back down to around, say, -15, which is still a great defense, but not the “historically dominant” defense it was last year. And because other teams that had top-five defenses in the past did not make the playoffs, because their offenses ranked low in DVOA (which the Broncos did last year), Football Outsiders will be skeptical about the Broncos’ chances.

What Football Outsiders could not have predicted, though, was the improvements on offense. Football Outsiders looked at it this way: Trevor Siemian was a quarterback with no NFL starts, the offensive line struggled last year and returned three players who started last year but mostly struggled when they played (Matt Paradis, Michael Schofield, Ty Sambrailo) while its notable addition (Donald Stephenson) wasn’t considered starting material. And although Max Garcia had some game experience last year and had solid outings, he was now tasked to replace Denver’s best offensive lineman last year, Evan Mathis.

As we know through the first two games, Paradis has been one of the better centers in the NFL, Schofield has improved, Stephenson played well against Carolina (but had his game against Indy cut short because of an injury) and Garcia has settled well into a starting role.

Last year, the O-line ranked 17th in FO’s rankings. This year, through two games, it’s ranked second. Football Outsiders could not have predicted the offensive line would show this much improvement. Hence, those who put their preseason stuff together made their best guess based on what they knew about the personnel in place from what they had already seen, not what they hadn’t yet seen.

FiveThirtyEight looks at things differently. Its ELO ratings before the season pick up where they left off from last season, but more compressed because of teams regressing toward the mean. Teams then gain or lose ground based on game outcomes and how unexpected outcomes were based on pregame ratings. In other words, the Denver Broncos got ranked higher there because they benefitted from a high ELO last year, even when accounting for any regression toward the mean.

So for those of you who grumbled about FO while praising FiveThirtyEight, neither one was that reliable of an indicator of where things would actually go for any team. Thus, don’t read much into either.

Now that we are a couple of games into the season, ELO still has the Broncos ranked high thanks to a 2-0 start and the defense still playing at a high level. Meanwhile, Football Outsiders has the Broncos ranked ninth after two games, with the defense ranked 11th (a ranking you might argue with) and the offense ranked 15th (which Aaron Schatz, who wrote about the DVOA ratings, said was somewhat unexpected). But there’s a caveat with both: It’s still early in the season and the rankings, as either one stands, don’t indicate for sure where teams will be by season’s end or who makes the playoffs.

In other words, one shouldn’t read too much into either ranking after two weeks because it’s still early, but you at least know that both sites will explain what they examine when they put the rankings together. For others who do power rankings, some will tell you what criteria they consider, but others don’t bother and may not have any at all.

Simply put: If you disagree with how somebody determines a power ranking, ask them what factors they considered to find out if they are staying consistent with criteria or they are just making it up as they go. If they tell you what factors are used, that doesn’t mean you have to agree with the ranking, but you at least know where they are coming from. If they can’t explain their criteria, now you have a case in which you can call them out for making assumptions and nothing more.

With all this said, there are a few things to keep in mind about when you can start judging where teams really will fall.

Week 1: As Schatz puts it, Week 1 of the NFL is National Jump to Conclusions Week. The truth it, the first week of the season doesn’t tell you anything about how the rest of the season will go, so relax.

Week 2: You can make a few judgments about trends, but again, it’s still early.

Week 3: You can safely judge which teams are likely to miss the playoffs but not much more than that.

Week 4: By this point, you have a good idea where most teams are heading and what trends you saw through the first two weeks actually stand a chance of holding up the rest of the season.

So check back after Week 4 to see how DVOA and ELO are taking shape and you’ll get a better idea of the NFL landscape.

I will close this with a few observations from me about teams who are 2-0 and teams who are 0-2 and whether or not you can safely judge them at this point.

Denver, 2-0: The Broncos defense still has the makings of a top five unit but I’m not expecting the dominance it demonstrated last season. That doesn’t mean the offense needs to bail the defense out; it only means the 2016 defense isn’t going to be put into the conversation of the best single-season defenses of all time. The offense, meanwhile, needs to finish more drives.

Houston, 2-0: The Texans are similar to the Broncos. They have the makings of a top-five defensive unit but the offense needs to finish more drives.

New England, 2-0: The offense has been in fine shape but the defense is not looking as good as it was last season. The Patriots are still the favorite to win the AFC East but the defense needs to get better if the Patriots have Super Bowl aspirations.

Pittsburgh, 2-0: All things point to the Steelers returning to the playoffs but it remains to be seen how they fare against the rest of the division.

Baltimore, 2-0: Neither team the Ravens played was a good team and now they draw an 0-2 team in Week 3, the Jaguars. We’ll probably know more after Week 4 when the Ravens face the Raiders, who have the best offense in the NFL through the first two weeks.

New York Giants, 2-0: The Giants have a leg up in the NFC East thanks to their win over the Cowboys. Their run defense looks good, but it’s too early to tell if they are the division favorites.

Philadelphia, 2-0: The Eagles beat two weak teams, so it’s still too early to judge them. They get the Steelers in Week 3 and we’ll have a better idea about how good the Eagles are.

Minnesota, 2-0: Here’s another team that’s going to be relying on its defense to compensate for an offense that hasn’t quite gotten on track. The Vikings’ win over the Packers was a good thing toward its quest to win the NFC North, though.

Buffalo, 0-2: The Bills draw Arizona in Week 3 and New England in Week 4. If the Bills don’t win one of those, they will miss the playoffs again.

Jacksonville, 0-2: Two chances coming up for my surprise playoff qualifier pick to prove my faith is justified: Week 3 vs. Baltimore and Week 4 vs. Indy. Lose both and all I can do is say I blew it.

Indianapolis, 0-2: The Colts have made it to the playoffs before despite an 0-2 start. But now they have to find a way to win either against San Diego in Week 3 or Jacksonville in Week 4 if they are going to do it again.

Miami, 0-2: If the Dolphins lose to Cleveland this week, stick a fork in them, they’re done.

Cleveland, 0-2: Yeah, they’ll probably finish with a top-five pick in the 2017 draft, but they could still beat a fellow bad team or two to avoid a winless season.

Chicago, 0-2: The Bears haven’t looked good the first two weeks but they get winnable games in Week 3 (Dallas) and 4 (Detroit). Go 0-4, though, and we’re talking about a top-five draft pick contender instead.

Washington, 0-2: If the Crimson Potatoes lose to the Giants in Week 3, we can go back to laughing at Dan Snyder’s ineptitude.

New Orleans, 0-2: As long as Drew Brees and Sean Payton are around, the pundits will never stop believing the Saints are a playoff contender. But if the Saints lose their next two games against Atlanta and San Diego, those pundits better start changing their tune.

Published by

Bob Morris

I'm a sports writer in real life, though I've always focused on smaller communities, but that hasn't stopped me from learning more about some of the ins and outs of the NFL. You can follow me on Twitter @BobMorrisSports if you can put up with updates on the high school sports teams I cover.