By Bill Barnwell, Football Outsiders – special to BSMW Patriots Game Day

This Sunday, the NFL schedule brings Buffalo to Foxborough for the opening week of the NFL season. Leading Buffalo into battle is third-year quarterback J.P. Losman, who won the starting gig over journeymen Craig Nall and Kelly Holcomb in training camp. Losman’s lot is an interesting one: drafted by a previous regime in Buffalo, he only lasted several games last season before what amounted to a popular revolt cost him his job. He would eventually regain it, but the damage to Losman’s confidence was done. Wins at home against Houston and Kansas City were the only ones Losman could muster; he was 2-7, without winning a single game on the road.

While that sort of data is an obviously small sample and too tiny to draw conclusions about any larger group from, it got me thinking to the usual stories you will see each week from the press about how young quarterbacks will struggle on the road against opposition for a variety of reasons: they’re not used to the noise, they’ll be scared, they don’t get the last substitution, etc. There are also the claims about weather: that a young quarterback will struggle because of the wind in Chicago, the snow in New England, or the ice in Green Bay. I’m skeptical of such claims, and with the trusty pro-football-reference database in tow, I decided to analyze them. I chose to analyze quarterback performance as opposed to winning percentage or a record-based analysis, as I was worried that I would be rewarding or harming quarterbacks for games where their team performed poorly and lost, but the quarterback came in and played well during mop-up time

First, I made a list of every quarterback who’s played since 1996. Then, I selected those quarterbacks who were enjoying their first season in the NFL with over 160 pass attempts (I chose 160 to try and get guys who had played at least a half season of games and averaged 20 passes per game in that timeframe).. This left me with 73 quarterbacks, starting with Tony Banks in 1996 and ending with Alex Smith last year, who had played a total of 803 games. From there, I tracked the performance of each quarterback on both home games and road games during that initial season with 160 passes. The statistics of those quarterbacks on the road and at home are below:


As you can see, the performances are more similar than you might imagine. Quarterbacks throw slightly worse on the road than they do at home, but not anything too dramatic.

We have the average percentages of all quarterbacks (of all experience levels) available for the 1999 season, for example, to compare the numbers to:


As you can see, rookie quarterbacks over the last nine years are slightly less accurate than a selection of all quarterbacks over the course of a season, at home, whilst away, and when moving from one to the other. While they throw for fewer yards per attempt than their veteran brethren, they also lose less of their yardage when going on the road, although the difference is very slight. To be honest, I was expecting a more dramatic difference between the two groups.

What I did next was consider the weather aspect of the problem. While compiling accurate weather data for all games is a project for the future, that data isn’t currently available on a game-by-game basis. What I did instead was, well, hypothesize some bad weather games. I defined a game played in bad weather to be one played after Week 8 of the regular season on the road at Baltimore, Buffalo, Chicago, Cleveland, Cincinnati, Denver, Green Bay, Kansas City, New England, New York (both Jets and Giants), Philadelphia, Pittsburgh, or Washington — a total of 14 teams. This gave us 102 “bad weather” games to look at, out of a subset of 403 total road games, which makes sense; since we are selecting about half the teams of the NFL and half their season schedule, we should be getting about 25% of the games altogether to look at. The data for those games versus the aforementioned rookie road performance is below.


This really surprised me. These quarterbacks actually performed equally as well in road games over the second half of the season in what would be considered poor weather games as they did during any other road game! Their completion percentage and Yds/Att fluctuated very slightly, but not with any sort of definitive change.

With that in mind, I wondered whether all rookie quarterbacks performed better over the second half of the season than they did over the first. The idea made sense, after all: they’d adjust more to the game, they’d know the playbook better, and/or they’d have more rest. The numbers:


As you can see, the concept was confirmed, although again, the differences aren’t truly dramatic. Rookie quarterbacks do certainly play more in the second half of the season, and when they do, they perform slightly better than their counterparts in the first half of the season do. However, it’s worth noting that if all quarterbacks who played solely in the second half played in the first half as well, their numbers very well might go down: there’s a reason, after all, they’re not playing in the first half of the season.

The results of the study, unfortunately, neither confirm nor deny the idea that young quarterbacks struggle when placed on the road. The numbers do imply, though, that any struggles aren’t as dramatic as they are sometimes made out to be. After all, it’s not like Tom Brady couldn’t win on the road.