Post

Conversation

Just in time for the NFL season, here’s something we’ve been cooking up for a long time: The best quarterbacks of all-time, as rated by our new model, QBERT. Every QB since 1950 on a level playing field with an insane amount of detail.
A scatter plot chart displaying quarterback performance. The x-axis shows QB plays, ranging from 0 to 14,000, and the y-axis shows QBERT ratings, ranging from 40 to 100. Green dots represent individual quarterbacks, with labels for Steve Young, Joe Montana, Peyton Manning, Drew Brees, Tom Brady, Brett Favre, Roger Staubach, Fran Tarkenton, Drew Bledsoe, Jack Kemp, Zach Wilson, and Josh Rosen. A curved line trends upward, and horizontal lines mark league average at 80 and replacement level at 60. The chart includes a watermark from Silver Bulletin.
David Watson 🥑
QBERT is both a rating system and a projection system. So the article also contains our projected ratings for every QB on a roster in 2025. The top 3: Lamar Jackson 104.5 Josh Allen 100.9 Patrick Mahomes 94.4 #4/5/6 are a bit spicier. Any guesses?
You’ve seen agents write code. Now meet agents that run your app, click buttons, and file bugs. TestDriver brings autonomous QA to your IDE. Like hiring a tester who works 24/7.
0:31
Book a Demo
From testdriver.ai
The irony of this system ranking Steve Young as the top QB is that Steve Young himself is probably the only player who can understand the methodology of this metric.
This is an interesting way to display the data, but isn’t it common sense (Lindy) that the QB longevity will track their performance: i.e., those QBs that are best will (generally) have the longest careers? I’m curious about the stories behind those QBs above 80 QBERT and fewer
Where do you suppose Roger Staubauch would have ended up if he had gone straight into the NFL instead of completing his term of service for the Navy?
I’ll take Dan Marino, Dan Fouts and Ben Rothlisberger. Where are they on this chart? All just threw an easy ball to catch. My definition of a great QB. Durable.
Nate, couldn’t you have normalized to 100? It’s bad enough to have the NFL’s crazy QB rating, but I thought you were a more user-friendly stat-geek.
Not very literate in the stats world but it would be interesting to have a graph of standard deviations above the mean across seasons played for each player, if that makes sense.
Can’t even imagine being a niners fan in the 80s/90s, according to the chart you get 10 years from the 2nd best QB of all time then immediately transition to the best QB of all time and both guys are throwing to the undisputed GOAT WR in his prime.
Long time listener, first time caller. So, I love this. Tried to comment on substack; alas, don’t subscribe. Curious: why no adjusted variable weighting for offensive teammate value/ production? Warren Moon was throwing to DHill, EGivins, HJeffries. Steve Young was throwing to
This is one of those analyses that true statisticians get incredibly annoyed with because no shit better QBs play longer. Bad QBs will get benched much earlier in their career. Good QBs will play much long provided that they don’t get injured. Outliers will exist.
Good stuff. About the only improvement that springs right to mind are 5 or 10 year intervals, to try and control for getting thrown in at 22 vs sitting and learning for 3 years
This is just selection bias in the greatest form. If we measure PLAYS INVOLVED IN, clearly teams will select the more successful QB to continue to call plays!
Super, super cool! Very detailed model, and ProductionScore actually produces a lot of similar conclusions. Namely, how much a QB contributes to their team's success. Can't wait to see more!
Does this model allow to see where QB’s were at at specific play counts? Where was Brady at on play count 8,000 for example. I’d assume everyone starts to drop after a certain volume
Your model has Dak number six overall in QBERT above Rodgers, Brady, and Brees…therefore it’s trash.
Imagine if Joe Montana played under the rules today and was protected like QBs today are Now imagine if Brady played when Montana played
So, what does QBERT actually stand for? None of the footnotes are readable for me, because they're stuck in the premium section.
Amazing! This was fun to dive into. Grew up in Denver and SF in the 90s and so Montana, Young, and Elway are my GOATs.
The two are certainly correlated. However, I think there may be a strong argument that performance predicts longevity rather than longevity predicting performance.
Might be worth it to double check which QBs you have listed as active vs retired. I noticed Jake Browning and Zach Wilson (current QB2s for the Bengals and Dolphins) listed as retired and Derek Carr as active.
Meet the AI-powered wearable over 1.2 million users rely on Hume Band continuously scans your biomarkers, catches health issues early, and guides you to smarter choices for a longer life. Click now to get 40% OFF + free shipping for a limited time only!
I hate that Brady gets so many accolades just because he played for so long (on good teams). His career stats are going to be padded because of longevity. But he was a cry-baby who consistently got penalties called on opposing defenses, meaning they didn't go after him as hard.