Why Tom Brady Is NOT Better Than Joe Montana
On NFL Live on Monday, the panelists discussed whether Tom Brady is a better quarterback than Joe Montana. The reason for the argument is that, with his victory on Sunday, Tom Brady passed Joe Montana for all-time playoff wins, and as such he's passed Montana's greatest accomplishment. The San Francisco section of my Twitter feed exploded with rage.
The horror! How dare somebody even mention that Boston heathen in the same breath as the Son of Walsh himself. EAST COAST BIAS!
You know, the screaming hordes on Twitter may actually have a point this time. I decided to check it out, numberFire style.
A Peer's Comparison
Most of the statistical arguments made in favor of Tom Brady are often washed-away in one simple phrase: "Montana played in a different era, man. There was none of that can't-hit-the-QB garbage back in the '80s. Montana's better because of what he had to deal with."
Fine, then. Let's ignore for a second that trying to argue that point is like saying Flava Flav is a better artist than Jay-Z because producing equipment in the '80s was worse. I suppose it's fair that you only compare a player to his peers.
So let's do exactly that. To use the words of our Dear Chief Analyst Keith Goldner, "There is no Holy Grail of statistics that can determine who's better". But rather than use the raw data, as most do, the most effective way to determine a player's value to his team is to devise how much better a player was than the league average. Pro-Football-Reference.com's adjusted-net-yards-per-attempt (ANY/A+) statistic is an easy way to do just that.
For ANY/A+, you first determine a player's net passing yards per attempt. For example, in Tom Brady's Super Bowl-winning 2004 season, that number is 7.06. That number is then adjusted for the relative strength of the passing game in a particular season. Then, the stat geeks determine what the average ANY/A would be for all QBs with at least 14 pass attempts per game. They then determine how many standard deviations an individual player's season is away from the mean, multiply that number by 15, then add 100 to it to make it look pretty.
According to Pro-Football-Reference, an average season is exactly 100 in any given year. A typical league-leading season is in the high 120's or low 130's. The best seasons are in the 140's.
If there's one thing to be said about both Brady and Montana, it's that they were never below-average. Instead, both players were above the 100-point mark for every season of each one's career. To break these numbers down a bit further:
Seasons at 110 ANY/A+: Montana 12 (92%), Brady 7 (64%)
Seasons at 120 ANY/A+: Montana 4 (31%), Brady 5 (45%)
Seasons at 130 ANY/A+: Montana 2 (15%), Brady 2 (18%)
All of this is to say that if Joe Montana's Ke$ha, with some legitimate hits that you couldn't get away from but nothing too groundbreaking, then Tom Brady's Carly Rae Jepsen, with some super-mega-ultra songs (see, now Call Me Maybe's stuck in your head) but the low points are much lower.
A lot of this disparity may be due to the type of system that each played in. Under Bill Walsh's West Coast style, Montana completed more short, accurate passes, which caused him to be more consistent. Brady, meanwhile, has been categorized by a need to go down the field early and often, especially in recent years with Randy Moss, Wes Welker, and Rob Gronkowski in the fold.
ANY/A+ isn't the only stat that has this type of distribution. Adjusted interception percentage (INT%+), which is calculated the same way with respect to the league average, sees Joe Montana with fewer seasons below 110 (one vs. Brady's four), but Brady's the one with three seasons (2007, 2010, 2012) that are better or equal to Montana's best (1981). And it's the same with TD%+, where Montana has more seasons above 115 (five to four), but Brady has more seasons above 120 (four to three). Adjusted sack percentage? Well, you know where I'm going with this. Montana has more seasons under 100 sack%+ (two to one) but also more seasons above 120 sack%+ (again, two to one).
Our Net Expected Points figure, which we usually use for determining a player's efficiency, only goes back to the start of the 2000 season. In the time since, Brady has put forth three of the top 10 most efficient NFL seasons, in 2007, 2011, and 2012. But while Montana's efficiency cannot be clearly quantified, his 1984, 1987, and 1989 seasons would all likely be in the discussion as well.
The Playoff Race
But Zach, it's all about the playoffs! Everyone knows that.
In the past, this was the end-all argument for Joe Montana against anybody. But here, it can just as easily apply to Tom Brady. He does have more playoff wins, a better playoff winning percentage, and more game-winning drives, after all.
But as with the regular-season statistics, this one's as close as can be. Check out the breakdown of each's playoff numbers. And as a note, the statistics about playoff comebacks and game-winning drives also comes from pro-football-reference.com.
|Playoff Win %||.696||.739|
|Super Bowl Record||4-0||3-2|
|Playoff Game-Winning Drives||5||6|
|Playoff Passer Rating 110+||6||6|
|Playoff Passer Rating 130+||3||3|
The sample size is too small for legitimate ANY/A+ numbers, or else I'm sure those would be the same too. Neither one has a distinct advantage when it comes to the playoffs. In fact, I'm prepared to call this one a draw. No, Donovan McNabb, you can't tie this time; you still lose.
The NFL Live guys don't have it quite right this time. Neither do the angry 49ers fan hordes on Twitter. Because ultimately, saying either Montana or Brady is better is pointless. At this point, they are essentially clones.
From being later round draft choices that didn't start much at a big-name college to coming in unheralded to leading a team to greatness in each's first (essentially) full season to being two of the most clutch QBs in NFL history, it's hard to tell them apart. While Montana was more solid throughout the entirety of his career, Brady's peaks were higher than Montana could have ever dreamed.
In short: nobody wins. Satisfied? No? Then have fun continuing your bar arguments to which there is no clear answer... yet.