Peyton Manning had more seasons with the highest success rate on passes of any QB in NFL history* (1994-2023)

What is a successful pass? It’s not just a completion because the defense will give you a four yard completion on 3rd and seven every time in just about every situation. The folks over at Pro-Football-Reference.com have defined a successful pass as one that gains 40% of the needed yardage to convert on first down, 60% of the needed yardage on second down, and converts (100% or more of the needed yardage) on third or fourth down. The value is a percentage that is found by dividing successful passes by dropbacks. Dropbacks are defined as passing attempts + sacks taken. Scrambles are not taken into account since no pass is thrown on those (duh!).

So what does this mean. That means that if a QB throws on first and 10, the receiver has to gain four or more yards for that pass to be considered “successful”. So let’s say that our hypothetical QB threw for four yards on first down - he threw a successful pass. It’s now second and six. Sixty percent of six is roughly four yards (0.6 x 6 = 3.6 which rounds up to 4) so now if he throws on second down his receiver has to gain four more yards again to count as a successful pass. It’s now third and 2 (if four yards were gained on second down). To throw a successful pass on third down the QB has to convert (his receiver needs to gain two or more yards).

This metric rewards QBs who are able to “throw their receivers open”, in other words, QBs who get their receivers lots of YAC through ball placement. It’s also interesting because a wildly successful QB could literally be throwing for an average of less than four yards per completion. But wait, if you look at completed air yards per pass attempt, most QBs in 2023 were below 4.0 yards. Only eight Starting Qbs had greater than 4.0 CAP/PA during 2023.

PFR has play by play data in their database going back to 1994. So we are able to look back that far to see things in this stat. The best season that any starting QB has had during this timespan was 57.8%. That was Tom Brady in 2007. This means that on roughly three of five passes Brady was successful in 2007. Brady had a completion rate of 68.9% that season, which led the league. That means that only 11.1% of his completed throws were unsuccessful. 68.9% Completion Percentage is tied for the 40th best single season completion percentage in NFL history.

Drew Brees 2018 was the best ever at 74.4%. Brees success rate that season was 57.3%, which is one of the best ever, but because his completion % was so high, 17.1% of his completions were “unsuccessful”. Compare that to Akili Smith in 2000; he had the worst success rate by any starting QB ever at 27.4%. His completion percentage that season was 44.2%. So his unsuccessful rate was 16.8% on completed passes. That was actually LOWER than Drew Brees unsuccessful rate in 2018. Context matters. No one other than Akili’s mom, is going to claim that he was ever a better QB than Drew Brees.

As an aside, how someone thought he was going to be a successful NFL QB still blows my mind, but as a fan of the team that spent the 26th pick on Paxton, I can’t cast too many stones.

The average success rate among starting QBs has been going up as overall QB play has gotten more and more efficient in the NFL during this timespan (see my tweet below).

The yellow line above is the average among starting QBs (apologies to those who are color blind). That bottomed out in 1997 at 41.0 and has been steadily climbing since. Interestingly enough it has been going down over the last four seasons from its peak at 48.4 in 2020.

No QB has ever appeared at both first and last on the success rate rankings among starting QBs. The closest that I see if Josh Allen who was second to last in 2018 as a rookie and then ranked first in 2020. Allen improved his success rate from 37.1% all the way up to 56.5% which is almost a 20% absolute improvement. That’s literally unprecedented (since the data only goes back to 1994).

I should note that there have been a few seasons where the worst starting QB was comparable to the average starting QB in 1997 (2016 and 2019). I should note that league completion percentage in 1997 was 56.2%. Twenty years later (2017) it was 62.1%.

Usually starting QBs who have a success rate below 40% don’t last as starting QBs in the NFL. Look at the guys who finished dead last in starter success rate over the years in the table below.

There are a few exceptions though. Russell Wilson in 2022, and Jeff Garcia in 2004. Both QBs had fairly long (or really long) runs as starting QBs in the NFL. Case Keenum, who was undrafted, only had two (maybe three) seasons as a team’s primary starting QB so he is much more akin to someone like Chad Hutchinson (undrafted), Bruce Gradkowski (also a 6th round pick) than someone like Tim Couch (first overall pick) or Blake Bortles (3rd overall pick). Bortles had the longest run as a starter among those guys. Oddly enough Keenum has started 66 games in his career while Bortles has started 73.

Only thirteen NFL starting QBs have led the league in success rate during this time frame and all of them made the Pro-Bowl that season (before it became the joke that it is now) while twelve of thirty made first team All-Pro during the season in question.

So let’s look at the arc for each of the 13 QBs starting with the most recent, Brock Purdy.

While Purdy only started five games in 2022, he had a success rate of 50.3 which would have tied him with Jared Goff for 6th if he had enough throws to qualify. He improved to 54.6% to lead the league in 2023, but I would not expect him to be able to improve much, if any, over that 54.6 value, given that there have been years when no QB topped 54.6.

Patrick Mahomes is the next of the magnificent thirteen, but we’ll shift to looking at the whole group rather than just focusing on each one individually, partly because Kermit the QB gets enough press already.

This also made me wonder what the heck happened in 2003 to cause all of these passers to have such a huge drop in success rate year over year? I can drop a link to the numbers as a google sheet for those who want to see them.

Interestingly enough, if you look at all starting QBs from 2002 and 2003, almost to a man, their success rate dropped year over year (or stayed the same).

Only Peyton Manning, Jake Plummer, Aaron Brooks and David Carr improved significantly. (I bet you never thought you would hear those four names in the same sentence.) While a number of Pro Bowl QBs had a huge drop-off year to year such as Rich Gannon, Brian Griese, Drew Bledsoe and Drew Brees, there was a general downward trend across the league. (In case you forgot, Griese did make the Pro Bowl one year.)

The NFL completion rate in 2002 was 59.2% which dropped to 58.8% in 2003. If I recall correctly this may have been the year when they were allowing (not calling) pass interference like the did in the previous year. This was over-corrected later after the Patriot defense shut down the Colts offense in the playoffs by ostensibly tackling Peyton’s receivers at the LOS.

There have been other dips and spikes in league completion percentage. While 2002 to 2003 is large, it is not the largest.

The biggest dip year over year since the merger was from 1984 to 1985. Completion percentage went down by 1.6% in that span. The biggest year over year jump happened between 2017 and 2018. The league-wide completion percentage went up 2.8% from year to year then.

The correlation between NFL COMP% and SUCC% is 0.911. So there a really strong correlation in the world of sports. For those who not into statistical analysis, that means that when 91% of the time when one stat moves in a direction the other moves in the same direction with roughly the same magnitude.

So what does this mean for the 2024 Broncos? Sean Payton needs a QB that has a top 5 success rate for his offense to work. So I believe he is looking at who is available, both in the draft and otherwise, and figuring out who can best execute his offense like Drew Brees used to. Unfortunately, PFR does not have success rate data for college QBs. Bo Nix led all BCS QBs with a completion percentage of 77.4% last season. That was the best ever (tied with Mac Jones) completion percentage in major college football history.

Of course college completion percentage is not a great predictor for NFL success. Some of the guys with a top 20 season barely sniffed an NFL roster or failed miserably in the NFL.

It would be interesting to see how college passing success rate correlates to NFL success, but that’s a study for another day once someone has crunched the college passing play data.