In any context, we should expect outliers in a measure to regress towards the mean. This is especially true in small samples. Given the number of games, drives, and plays in a NFL season, football teams are ripe for experiencing wild swings in performances when studying how efficiently they turn drives into touchdowns. Before the 2020 season starts next week, I wanted to look at some of the most extreme scoring splits from the 2019 season, which teams were responsible for those splits, and point out the corresponding areas of expected improvement or decline.
First I am going to look at how often teams played with the lead and while trailing. The following visualization should give some insight into which teams found themselves in the lead most frequently and which were often trailing:
When reflecting upon the 2019, I think we would consider the Chiefs, Ravens, Saints, and 49ers to be the most successful teams overall. Sure enough, they ran the first, second, third, and fifth most plays while leading the game, respectively. Interestingly, the Patriots were fourth despite a bit of a disappointing season. Given the roster turnover in New England this offseason, I would be weary of drawing any conclusions from the team's 2019 results. The Rams, Broncos, and Bucs have surprising showings here. The Bucs and Broncos also spent an above average percent of plays trailing, so I would not throw them into the same cluster as any of the teams ahead of them. The Rams are different. They profile as a team similar to Minnesota last year and trailed much less than Green Bay (whose record was buoyed by a 6-1 record in one score games). Interestingly the Rams were not especially unlucky in one score games (3-3, so 6-4 in games decided by more than a score). Given every team ahead of them won at least 10 games last year and the Rams division rivals in Seattle and Arizona fared much worse by this measure, I think the general public is sleeping on the Rams going into 2020. The Lions also should be the beneficiaries of positive regression this season. Detroit finished middle of the pack in percentage of plays leading despite Matthew Stafford starting only half of the season. Yet the Lions finished 3-12-1 by way of a pitiful record of 3-7-1 in one score games (the sixth worst record in such games over the last five years). If the Lions get a healthy season from Stafford and their winning percentage in one score games approaches 50, I would expect the Lions to compete for a playoff spot in the NFC and even challenge for the NFC North title.
On the less positive side of things, the Titans and Falcons were in the bottom third of the league in percentage of offensive plays run while leading. Granted, the Titans made a switch at quarterback after six games and subsequently looked a lot better. Still, if the Titans want to improve upon last year and stave off whatever regression Tannehill faces after being one of the most efficient quarterbacks on a per play basis, they need to get off to better starts this season. If the Falcons percentage of snaps with the lead is similar this season, I would expect the Falcons to be worse than 7-9 and for Dan Quinn to be without a head coaching job. Houston spent an inordinate time either trailing or tied given their record of 10-6 last year. They went 8-3 in one score games, so they actually lost more games than they won by multiple scores. There was a lot of reshuffling of the roster this year, including trading DeAndre Hopkins and getting back David Johnson, trading for Brandin Cooks, and signing Randall Cobb. They still have Deshaun Watson, so you cannot count them out in the playoff hunt, but there will have to be some improvements in their underlying play if they hope to contend for the division crown again.
That is it for my thoughts on these score differential splits. In my next post, I am going to continue to pick out regression candidates based on per drive stats and red zone splits.
No comments:
Post a Comment