2018 NFL Preseason Predictions Review — How Did We Do?

2018 NFL Preseason Predictions Review — How Did We Do?

February 6, 2019 – by Seth Trachtman

Super Bowl LIII is in the books, and the New England Patriots are your world champs for the sixth time in 18 seasons. Prior to Week 1, we posted our 2018 NFL preseason predictions. Now that the season is officially over, it’s a good time to review how our predictions finished.

NFL Predictions 2018: The Overall Picture

Entering the season, we projected a Super Bowl matchup of the Patriots (27.6% odds) vs. the Rams (14.1% odds).

Our preseason projections had New England with the highest odds to win the Super Bowl (14.9%), more than 5% better than the second best Pittsburgh Steelers (9.7%).

As for the rest of our summary predictions:

2018 AFC Predictions Recap

#1 Seed Pick: New England Patriots (29.5%), with Pittsburgh Steelers (17.7%) the only other team over 10%.

Most Improved: Cleveland Browns (from 0 to 6.5 wins) and Houston Texans (from 4 to 8.5 wins)

Biggest Decline: Buffalo Bills (from 9 to 5.5 wins) and Pittsburgh Steelers (from 13 to 9.9 wins)

Biggest Loser: Buffalo Bills (5.5 projected wins)

2018 NFC Predictions Recap

#1 Seed Pick: Technically the Vikings, but there are six teams between 15% and 9% … Minnesota Vikings (14.8%), Los Angeles Rams (14.1%), New Orleans Saints (13.6%), Philadelphia Eagles (12.4%), Green Bay Packers (9.4%) and Atlanta Falcons (9.2%)

Most Improved: New York Giants (from 3 to 6.9 wins)

Biggest Decline: Philadelphia Eagles (from 13 to 9.5 wins), Minnesota Vikings (from 13 to 9.9 wins) and Carolina Panthers (from 11 to 8.2 wins)

Biggest Loser: Tampa Bay Buccaneers (6.3 projected wins)

Grading 2018 Our NFL Predictions Results

#1 Seed Pick: We didn’t correctly project either of the #1 seeds (Kansas City in the AFC and New Orleans in the NFC). We gave the Chiefs a 4.5% chance to get the #1 seed, sixth best in the AFC. The Saints trailed Minnesota by just 1.2% in their odds for the top seed, making them third most likely in the NFC according to our projections.

Most Improved: The AFC projections were on point, with the Browns and Texans both seeing 7-win increases vs. 2017. The Giants weren’t quite as good as expected, with only a 2-win increase vs. 2017. The biggest gainer in the NFC turned out to be the Chicago Bears, going from five wins to 12.

Biggest Decline: We projected some of the biggest decliners in the AFC with Buffalo (3-loss increase) and Pittsburgh (3-loss increase). Jacksonville had the biggest decline in the conference, losing five more games in 2018. All three of our biggest projected decliners in the NFC (Eagles, Vikings, Panthers) did plummet, though Arizona was the biggest decliner of all, going form 8-8 to 3-13.

Biggest Loser: Our projection of 5.5 wins for the Bills was only a half-win below the actual result, but the Jets, Raiders, and Jaguars managed to finish with worse records. The Bucs finished with five wins, even worse than our projection of 6.3 wins, but Arizona and San Francisco were even worse.

Winning Percentages: Projected vs. Results

Below is a comparison of our projected winning percentages vs. results.

Team That Performed Better Than Expected

Team	Proj Win Pct	Actual Win Pct	Difference
Chicago	.444	.750	+.306
Kansas City	.488	.750	+.263
New Orleans	.600	.813	+.213
LA Rams	.606	.813	+.207
Indianapolis	.425	.625	+.200

The combination of new head coach Matt Nagy and the acquisition of Khalil Mack spurred the Bears to a surprising NFC North title with a 12-4 season, 4.9 wins better than our preseason projection. The Chiefs were also far better than expected due to #MVPat, as Patrick Mahomes was projected to be a slight downgrade from Alex Smith, but instead led the Chiefs to 12 wins (4.2 more than projected). With the return of Andrew Luck and an excellent rookie class, the Colts were 3.2 wins better than our projection.

Teams That Performed Worse Than Expected

Team	Proj Win Pct	Actual Win Pct	Difference
Arizona	.438	.188	-.250
Jacksonville	.550	.313	-.237
Oakland	.463	.250	-.213
San Francisco	.456	.250	-.206
Green Bay	.575	.406	-.169

The Cardinals had an absolutely disastrous season, four wins worse than our projection, then fired head coach Steve Wilks after only one season. Jacksonville’s offense regressed severely this year, and was a huge disappointment after advancing to the AFC Championship in 2017. Green Bay won 3.2 fewer games than we projected, and fired head coach Mike McCarthy during the season.

The Final Verdict

Like every season, there were plenty of hits and misses, but our projections this year performed slightly better than expectations.

Here were the most likely results, given our predicted probabilities:

52% chance to correctly predict 3 or 4 of the 8 division winners (we got 3 correct)62% chance to correctly predict 3 or 4 of the 6 AFC playoff teams (we got 4 correct)59% chance to correctly predict 3 or 4 of the 6 NFC playoff teams (we got 3 correct)48% chance the Super Bowl winner would be among our 5 most likely teams (it was our most likely team)

Typically, we’d expect only 2 or 3 of those 4 outcomes to happen, but instead all four did. That either means luck was on our side or that our preseason probabilities should have been a little more confident. Given how much a single coin flip (see: Patriots over Chiefs) or missed call (see: Rams over Saints) can have, we’re still inclined to believe that our projections are likely doing a solid job, and that luck, as always, plays a big role in whether their final accuracy for a given NFL season ends up being somewhat better or worse than expected.

Look for our 2019 preseason predictions in early September, around the time we release our NFL Pick’em Picks and NFL Survivor Picks for the new season.

If you liked this post, please share it. Thank you! Twitter Facebook