This got me thinking about how we can use QBR to untangle actual vs. expected QB wins.
Short thread... https://twitter.com/Tucker_TnL/status/1353520094345916417
Short thread... https://twitter.com/Tucker_TnL/status/1353520094345916417
If QBR provides us with the % chance that a team wins with a given level of QB play, all we have to do is sum those percentages to get a relatively crude measure of "expected wins."
In doing so, we can see which QB win totals best/worst represent their play.
In doing so, we can see which QB win totals best/worst represent their play.
Postseason QB wins over expected, 2006-20:
Wilson, Mahomes, Roethlisberger, & Flacco among those whose win totals may be inflated.
Rodgers, Ryan, Rivers, & Romo among those who've gotten the short end of the stick.
(Excludes 2017 due to missing data)
Wilson, Mahomes, Roethlisberger, & Flacco among those whose win totals may be inflated.
Rodgers, Ryan, Rivers, & Romo among those who've gotten the short end of the stick.
(Excludes 2017 due to missing data)
Obviously, this doesn't account for "butterfly effect" of losing one game and thus not having the chance to play in another, but it suggests:
-Rodgers hasn't had much support
-Brady hasn't "gotten all the breaks"
-Wilson got plenty of support from run game/defense in early years
-Rodgers hasn't had much support
-Brady hasn't "gotten all the breaks"
-Wilson got plenty of support from run game/defense in early years
Thanks to @thomas_mock & his QBR scraper, which you can access here: https://github.com/jthomasmock/espnscrapeR :)