Predicting the 2018 NFC East winner & losers: consolidating the position rankings

Over the past two weeks, I’ve posted several articles about position groups, asking Hog’s Haven members to rank those position groups & head coaches in the NFC East.

It was nice to see commenters (mostly) put away the tendency to root for the home team and try to be (somewhat) objective. Burgundy-colored glasses certainly had an effect, but it was more akin to a halo effect, where — when it was close -- the advantage went to the home team. Overall, though, commenters were pretty fair in their assessments.

At a time of year when NFL blogs struggle to offer anything to readers, and when comments on most NFL-related articles number in the dozens and often don’t muster more than a single-digit total of comments, Hogs Haven members really came through on these ranking articles. We’ve seen between 1,000 and 3,000 votes on each of the various polls, and the 10 ranking articles had a total of 1,700 comments!

Given that I merely posed the questions, and the comments sections supplied the answers, this is a wonderful example of reader-driven content.

Now that the series is finished, to wrap it up and try to put a bow on it, I wanted to put the many rankings on a chart to consolidate them and see what they said about the expectations for the coming season among the Hog’s Haven members. Unfortunately, there’s nothing really straightforward about doing that.

I would have liked to have used the raw poll results to establish the rankings by position, but that would have been unlikely to reflect the reality of the situation or the opinions of the readers accurately.

Instead, I started with the idea that when you rank the teams from 1 - 4, there are 10 position points to give out (1+2+3+4=10). I decided that the basic idea should be to allocate the 10 points across each position ranking, allowing ties between teams wherever it seemed appropriate. The rule was that the 4 rankings had to add up to 10 points.

For each position ranking, I looked first at the poll results, but then leavened them by taking into account the comments section to establish general rankings. Finally I added a dash of something — common sense, balance, judgement, normalization... I’m not sure what to call it. Basically, I just tried to make sure that the final rankings weren’t unreasonably skewed by the raw voting results.

A few things were obvious when I looked at the results of the polls and the content of the comments sections. Firstly, it’s a Redskins website, so there’ll be a significant Redskins bias. Secondly, reading the comments section made it clear that the Eagles fans had showed up in bigger numbers than the Giants fans, and the Cowboys groupies almost hadn’t showed up at all.

The other thing that I found from this exercise is that recency bias plays a role.

Using the recent past to project the future

The NFL is intrinsically designed to be a parity-driven league; the draft, revenue sharing, the salary cap, compensatory draft picks, even the schedule; everything about the NFL is designed so that every team from every market has a legit opportunity to compete year-in and year-out.

Fact is, the ‘competitive balance’ in the league gives each team hope of finishing at the top of the standings regardless of its record the previous season. Sometimes for no other reason than that other teams in the division are even worse. In the NFL, this is called parity.

Yet you wouldn’t notice that if you were to look at the majority of season outlooks published at this time of the year, as most of them seem to do little more than copy last year’s standings and add a little “surprise” here or there.

Case in point: SI.com’s Prediction for every NFL team’s 2018 season record published last month. It’s a 5,800+ word missive in which author Jonathan Jones boldly declares that “next season’s playoffs will probably look a lot like last season’s,” and has all eight division champions from 2017 repeating in 2018.

Do you know how often all eight division champs have repeated since the NFL moved to an eight-division format in 2002? Never. Not once.

And I’m not trying to pick on SI.com particularly, theirs was just the first 2018 projection that popped up on my Google search. Plenty of other projections follow a similar pattern where they stick pretty close to last year’s standings.

But why is that?

Recency bias.

Recency bias is the tendency to think that trends and patterns we observe in the recent past will continue in the future. Because it’s easier, our minds are hardwired to use our recent experience as the baseline for what will happen in the future. In many situations, this bias works just fine, especially if you’re making short-term predictions. Even for highly changeable events like the weather or the stock market, making short-term predictions according to events in the recent past works fine much of the time: If it’s 75° outside today, odds are it will also be around 75° tomorrow, give or take a little. But it’s unlikely temperatures will jump or drop by 30° tomorrow. Not impossible, but highly improbable.

But predicting the future in the long-term according to what has recently occurred is no more accurate than flipping a coin.

Take the NFL: Since 1990, when the NFL moved to a 12-team playoff format, an average of 5.7 new teams made it into the playoffs every year. That means only half of each year’s NFL playoff participants make it back to the playoffs the following year. In fact, the Falcons are the only NFC team to make the playoffs the last two seasons. Yet every preseason team ranking has last year’s top teams at the top. Why? Recency bias.

So what does all this mean for the [NFC East]? It means that the six NFC playoff teams from 2017 (Eagles, Vikings, Rams, Saints, Panthers, Falcons), frequently mentioned as “locks” to repeat as playoff teams this year, are anything but “locks”. There is a strong likelihood that at least two, maybe three, and possibly even four of those teams will not repeat as playoff teams in 2018.

The power of recency bias is intense in the business of sports prediction. For example, last year, in the wake of the Giants’ 11-5 season and the Eagles’ 7-9 finish, Hog’s Haven voters largely put the Giants at the top of the voting and the Eagles near the bottom.

This year, following the 2017 Super Bowl win for Philly and the Giants’ 3-13 meltdown, things were turned around.

The results of the voting, in light of all the factors that influenced them, were not surprising.

When I consolidated the voting, the comments and some personal judgement, and blended it all together, here’s what I came up with:

Unsurprisingly, the Eagles came out with the best average ranking following their Superbowl win. The Eagles were rated either a clear first place or tied for first in 9 of the 11 rankings. Further, they were ranked last in only a single position group — running back.

The Eagles, when all the position group rankings were averaged, had the best ranking at 1.8, indicating that readers are expecting them to, once again, win the division title.

The Redskins, with a composite score of 2.3, were ranked second by the readers. The real surprise is that they weren’t ranked 1st by Hog’s Haven readers, but — as I mentioned above — the Eagles fans showed up in force to read, vote in, and comment on the articles in this series.

The Redskins were in the top rank for only two position groups (OL and CB) and in both cases Washington was in a tie with at least one division rival. The real benefit for the Redskins was that they weren’t ranked #4 in any position group by the readers. The ‘Skins were typically ranked 2nd or 3rd.

The Giants & Cowboys were routinely found in the 4th position in the individual position rankings in the series of articles on Hogs Haven, with the Giants being ranked as the worst in a position group 5 times, and the Cowboys bringing up the rear 6 times. The Giants could only muster 2 first place finishes (ties for WR & S), while the Cowboys had two uncontested first place rankings (RB & ILB) and tied with the Eagles and ‘Skins for the best ranking at the OL position.

In the end, the unscientific analysis indicates that voters/commenters see a huge talent gap, with the Eagles clearly having the best roster, the Redskins second, and the Cowboys last. As is often the case, fans seem to be projecting the results from last season onto the season ahead, with a bit of a Redskins positive bias to push them up the rankings, and some Cowboys negative bias to push them down.

Once again, given that these rankings were derived by asking opinions from Hogs Haven members, it’s no surprise that the Redskins came out ahead of both the Cowboys and the Giants. What is perhaps surprising is how much respect was given to the Eagles, though the evidence suggests that their fans ‘traveled’ to Hogs Haven and heavily influenced the voting, and participated in the comments. Giants and Cowboys fans had the same opportunity, of course, but — and this is especially true of Dallas fans — they simply weren’t here voting and commenting. I suspect that the Eagles fans, with their team having won their first-ever Lombardi Trophy, are staying more ‘plugged in’ through the offseason in an attempt to continue the celebration, and with anticipation of another successful season coming up.

The overall message for the Redskins faithful from this series may be that the readers seem to think that the ‘Skins have enough to get to the playoffs, but that it will be as a wildcard team, not as division champs. Instead, they see the Eagles — returning the head coach and most of the starters — as the clear favorites to win the NFC East again in 2018.

I’ll remind you that, last year, the same exercise ranked our division rivals in this order:

• Giants
• Cowboys
• Eagles

which is the exact opposite of their actual finishing positions in the 2017 division contest. The team we all seemed to expect to be the weakest in the division — the Eagles — ended up winning the Lombardi Trophy. The team that was projected to do the best — the Giants — started the season 0-4 even before the injuries really started to bite, and they never recovered.

Final thoughts

In the parity-driven NFL, last season doesn’t matter. Outside of the Patriots, there are very few teams that have shown a year-on-year consistency that would allow you to confidently project them as a playoff team year after year.

As always, the NFL is a league of great parity; the NFC East remains (in my opinion) a division without a single dominant team; every season begins with the knowledge that anything is possible. And that’s why last year doesn’t matter for NFL teams, and why you’re allowed to be optimistic about the Redskins’ playoff chances this season. 2018 should be another great season for the division. Personally, I expect it to be a great season for the Redskins as well.

Here are the links to the 2018 “Ranking the NFC East” articles:

Quarterbacks & Running Backs

Wide Receivers

Tight Ends

Offensive Line

Safeties

Cornerbacks

Linebackers

Edge Rushers

Defensive Line

Head Coach