Advanced NFL prediction model says Vikings good, Packers awful
Welcome to the Friday edition of The Cooler, where sometimes the surprises just find you. Let’s get to it:
*A colleague e-mailed FiveThirtyEight’s 2018 NFL predictions, which the site presents as its “ELO” ratings in various sports. Beyond just eyeballing strengths and weaknesses, these ratings assign a value to teams and then put each one through 10,000 simulations to arrive at predictions.
The Vikings, not surprisingly, are given a positive outlook. They are expected to win 9.9 games (let’s just call it 10) and are given a 7 percent chance to win the Super Bowl. Both of those are tied for the third-best in the NFL with Pittsburgh. Last year’s Super Bowl champion Eagles and runner-up Patriots have the most favorable chances of getting to the Super Bowl, and rightfully so.
The big surprise, though, was that you had to scroll … and scroll … and scroll before you got to the Packers. The model says Green Bay will win 7.1 games and has just a 23 percent chance of even making the playoffs. That’s worse than the Lions and not much better than the Bears. Huh?
Well, don’t get too giddy. At first I wondered if 538 had uncovered or detected some glaring weakness that other models — ESPN has Green Bay firmly in its top 10 — ignored. Alas, it seems to be an inefficiency in the model itself that is causing this strange forecast.
See, a deeper dive into each team’s preseason ranking shows that it is arrived at thusly: “At the start of each season, every existing team carries its Elo rating over from the end of the previous season, except that it is reverted one-third of the way toward a mean.”
So basically, 538 hasn’t factored in the return of Aaron Rodgers. Green Bay’s roster has some holes, to be sure, but if Rodgers is healthy for 16 games the Packers should be climbing up those rankings pretty quickly as the season moves along.
*Your semi-regular update on the Twins’ pursuit of avoiding 90 losses: They need to go at least 10-13 in their final 23 games to keep from hitting that magic number for the sixth time in eight years.
If we can borrow from 538 again — this time with more complete data — the site thinks the Twins will make it and finish 75-87. There are some winnable games, but I’d call it more of a coin flip at this point that 90 is avoided.
*And finally, as I noted on Twitter last night, my role at the Star Tribune is changing a little in the upcoming months. I’ll be the editor of our Timberwolves coverage this season, an exciting assignment that I’m very much looking forward to. I’ll still be blogging regularly, but my work will appear a little less frequently in print as I devote more attention to Timberwolves duties.
If you have thoughts or ideas about the Wolves, don’t hesitate to send them my way. It should be an interesting and entertaining year.