Quantifying Brock Purdy's Success: Player or Scheme?

Adjusting for surroundings, Brock Purdy has been average, which is plenty good enough for the 49ers to win it all

After Jimmy Garoppolo went down with a injury in Week 13, I looked through the numbers to show that he’s performed well throughout his career in all number of differing circumstances, and perhaps he wasn’t only a productive of the system. In fact, the there was strong evidence that he mattered a lot.

The subsequent evidence from the San Francisco 49ers’ playoff run has made my prior analysis look a little foolish. Not only has the loss of Garoppolo not sunk the 49ers title hopes, the team has looked better than ever with Brock Purdy under center.

The 49ers have won seven straight games, including a 41-23 drubbing of the Seattle Seahawks in the Wild Card round. The 7th round rookie is averaging 0.30 EPA per play over seven games, only slightly behind league leader and likely MVP Patrick Mahomes (0.31).

We all intuitively know that the 49ers offense gives lots of contextual advantages to its quarterback in scheme, receivers and offensive line, but intuition isn’t enough to answer the relevant questions of exactly how good Purdy has been relative to those advantages, and how much of a boost he gives the 49ers (if any at all) versus Garoppolo, who returned to the practice field this week for “light work”.

Luckily, we have a tool to measure in a concrete, definable way the value quarterbacks have brought to their teams outside of surrounding in the form of adjusted quarterback efficiency (AQE). This metric went through a few iterations during its development this season, but was in a good place by the end of the year.

Updating AQE for the first round of the playoffs, shows that Purdy has been a solid quarterback when you adjust for surroundings, but has also had the most favorable situation in the NFL.

THE ADJUSTED EFFICIENCY NUMBERS

Purdy’s per-play efficiency moves from second without adjustments to 13th with them. Nearly two-thirds of Purdy’s positive EPA per play is due to his receivers, scheme, blocking, and other factors according to AQE.

Garoppolo also benefited greatly from his situation with the 49ers, though not to the same degree. Despite Garoppolo having an unadjusted efficiency roughly 0.07 EPA per play lower than Purdy’s, post-adjustment he’s slightly higher (+0.03).

When we break down the per-play effects of the AQE adjustments, you see that Purdy has had better luck with receiver drops and strength-of-schedule.

Purdy’s receivers have only dropped three of his 135 catchable passes (drops + completions) according to charting data from FTNData. That 2.2% rate is the lowest in the NFL for any quarterback with at least 200 pass attempts this season, significantly lower than Garoppolo’s 6.3% (10th highest).

Purdy has only faced one top-10 pass defensive according to my schedule-adjusted numbers, while matching up bottom-15 defenses in the rest of the games. Purdy faces the 4th ranked pass defense in the Dallas Cowboys this divisional round matchup.

Both quarterbacks had tremendous advantages in yards-after-catch EPA generation over expected, either due to scheme or receiver ability.

BRINGING IN SAMPLE SIZE

Because many quarterbacks have had more than double the number of dropbacks as Purdy, it’s not really apples-to-apples comparison to rank order they per-play numbers. The more sample we have for a quarterback, the more confident we are that his play is sustainable, and not mostly noise.

Bayesian updating is my favorite way to adjust for sample size and outlier results, which generally narrows the range of “true” efficiency, but also differentiates between quarterbacks with similar per-play numbers, but vastly different sample sizes.

All quarterbacks move closer to the mean expectation for an NFL starter, and ones with only 200-ish dropbacks like Purdy move even further.

With the Bayesian adjustment on top of the contextual adjustments, Purdy drops from second in per-play efficiency to 15th, two spots lower than without the Bayesian adjustment.

It might seem insulting to say that Purdy has been roughly average when his results have been stunning, but a 7th round rookie being average is a huge success. And Purdy will continue to have the surrounding advantages that come with being the 49ers quarterback around him the rest of the offseason. If Purdy can continue to just be average and leverage the efficiency of the rest of the offense, the 49ers can get top-tier passing efficiency, and ride their rookie all the way to the title.