NFL free agency officially begins on Monday, and it’s expected to be among the most eventful signing periods ever given the number of quarterbacks available — one of whom is arguably the greatest in league history.
But while teams are lining up to sign Tom Brady, who will be 43 years old at the start of next season, it’s been a different story for Philip Rivers, who is sixth all-time in passing yards and four years younger than Brady. His former team, the Los Angeles Chargers, unceremoniously severed ties with him weeks ago. Just one team, the Indianapolis Colts, is reportedly interested — though even that interest seems to be waning, made less appealing by the prospect of drafting a quarterback in April. That could force Rivers into retirement, and maybe even the broadcast booth.
This is despite Rivers besting Brady over the past three seasons in passer rating (96.2 to 96) and yards per pass attempt (8.0 to 7.4).1 Last year, Rivers crushed Brady in yards per attempt, 7.8 (11th-best among qualifying quarterbacks) to 6.6 (27th).
So why the lack of love for the Charger? Perhaps it’s River’s propensity for throwing picks in a league where coaches have done their best to scheme them out of existence. Rivers tossed 12 more interceptions in 2019 than Brady did — the 10th year in a row that Brady beat Rivers in the stat.
It stands to reason that Rivers, on a 5-11 team that was often playing closer games, would throw more interceptions than Brady would on a 12-4 Patriots squad that rarely had to take chances.2 Of course, this could be the NFL’s version of the chicken-and-egg question: Rivers’s turnovers and Brady’s lack of them could have shaped their team’s respective records as much as those records were caused by them. But looking back on NFL play-by play data since 2010, we can figure out the expected probability of an interception for each pass based on factors such as the team’s win probability (desperate QBs throw more picks), the air yards of the throw (deeper passes are more likely to be picked) and the field position of the play (passes in the middle of the field are less likely to be picked).3 From there, we can assign every quarterback an expected total of interceptions and compare it against how many he actually threw.
In 2019, the results were clear: Even after accounting for the impact of a team’s in-game position on interception rates, Brady remained elite at avoiding picks, while Rivers was far more turnover-prone than expected.
Interceptions over expected for Philip Rivers and Tom Brady since 2010
Let’s stipulate that the sample size is small and influenced by many factors, including passes that go off of receivers’ hands, ones tipped by defenders at the line of scrimmage and those resulting from miscommunications between the receiver and quarterback. Defenders also drop would-be interceptions and occasionally make great catches.
But it hardly seems lucky that Brady is under his expected number every year. While his steadily declining yards per attempt suggests that he can no longer be counted on to save the day for a team, his excellence at avoiding picks means he probably won’t cost his team the game, either. It seems like Brady could still fit as a game manager for a solid team that is being hurt by its current quarterback’s play.
But is Rivers as predictively bad at interceptions now as Brady seems to be good at avoiding them? While Rivers threw over five more picks than the Chargers should have expected, given the game situations at the time of each pass, his performance is fairly random: good or at least average at avoiding picks one year, bad the next.
But that volatility might be the reason teams are ruling out Rivers and keeping Brady as a possibility. One of Brady’s rumored suitors, according to the team’s head coach himself, is the Tampa Bay Buccaneers. The Bucs’ current quarterback, Jameis Winston, is Brady’s polar opposite: Last year, he became the first quarterback ever to throw at least 30 touchdowns and 30 interceptions in the same season. His 30 picks were the most thrown in a single season this century and tied for third-most since the 1970 AFL-NFL merger.
After Winston’s 30th miscue was returned for a touchdown — for the seventh time on the season, an NFL single-season record — Arians seemed to channel all NFL coaches on the topic of interceptions: “You’re not going anywhere. … You’re going home if you lead the league in giveaways.”
Those picks may end up being too high a price to pay even for Winston’s sterling passing yardage efficiency (8.2 per attempt last year, tied for fifth-best in the league, which brought his career rate up to 7.7, 13th all-time). And while it may fall short of changing Arians’s mind about his current quarterback, our interception model suggests that Winston was “only” 13.1 interceptions over expected — or 2.6 times the number that got Rivers his walking papers out of Los Angeles.