Comparing a few Quarterbacks

So my friend Mark wrote this. I'm responding with this:

PASS ATT COMP YARD TD INT rank Young 4149 2667 33124 232 107 1 Manning 4890 3131 37586 275 139 2 Brady 3064 1896 21564 147 78 3 Marino 8358 4967 61361 420 252 4 Favre 8224 5021 57500 414 273 5 Fouts 5604 3297 43040 254 242 6 Elway 7250 4123 51475 300 226 7 RUSH ATT SACK YARD TD FUM rank Young 722 358 4239 43 68 1 Manning 269 170 701 13 45 7 Brady 239 182 435 3 59 6 Marino 301 270 87 9 110 4 Favre 526 424 1774 13 138 3 Fouts 224 476 13 5 Elway 774 516 3407 33 137 2 value ATT rank Young 7103 4149 1 Manning 4927 4333 4 Brady 1614 2548 7 Marino 6752 8358 2 Favre 2672 7612 6 Fouts 5017 5604 3 Elway 3155 7250 5 YARD TD INT ATT AY/A rank Young 33124 232 107 4149 7.382 1 Manning 37586 275 139 4890 6.970 2 Brady 21564 147 78 3064 6.372 4 Marino 61361 420 252 8358 6.487 3 Favre 57500 414 273 8224 6.001 7 Fouts 43040 254 242 5604 6.190 5 Elway 51475 300 226 7250 6.111 6 RW RL R% PW PL W% projR% projW% rank Wrank Young 78 25 0.757 8 6 0.718 0.685 0.667 2 4 Manning 92 52 0.639 5 6 0.614 4 8 Brady 70 24 0.745 12 1 0.783 1 2 Marino 131 85 0.606 8 10 0.583 0.617 0.594 6 12 Favre 139 85 0.621 11 9 0.610 0.620 0.607 5 10 Fouts 3 4 0.429 7 14 Elway 91 48 0.655 14 8 0.650 0.593 3 6 rank sum final rank Young 8 1 Manning 23 3 Brady 22 2 Marino 25 4 Favre 31 6 Fouts 33 7 Elway 26 5

There are a few things that I need to explain. The top chart shows QB passer ratings and the components of that rating.

The second chart shows a QB rushing rating that I built off of the passer rating and the components of that rating.

The third chart shows a QB value that was developed by Chase Stuart. The explanation for this value can be found here.

The fourth chart shows the adjusted yards per attempt for each QB.

The fifth chart shows the regular season winning percentage, playoff winning percentage, and weighted total winning percentage for each QB. I came up with this myself, and each playoff game is worth twice as much as each regular season game. For the older quarterbacks there was some missing info, so I had to project what the starting record for them would have been in these games. For Steve Young it doesn't include the 40 games he started from 1985-1991, 19 of those with Tampa Bay and 21 of them with San Francisco. For Dan Marino it doesn't include the 26 games he started in 1983, 1987, and 1993. For Brett Favre it doesn't include the 13 games he started in 1992. For John Elway it doesn't include the 92 games he started in 1983, 1984, 1987, 1988, 1989, 1992. For all of these missing games I projected how they most likely did based on their team's winning percentage and their history.

The sixth chart shows combines the results of all the other ones. In each chart I rank the quarterbacks. For the last chart I found the sum and the ranked each one, from lowest to highest point total (like in golf). Each category was worth one point, except the winning percentage chart, for which I made it worth double. In the final tally, the best QB of all the ones Mark mentioned in his post is Steve Young. (This wasn't surprising to me, as the same Chase Stuart who came up with the value rankings has him listed as the best QB of all time here.) Young is followed by Tom Brady, then Peyton Manning, Dan Marino, Dan Elway, Brett Favre, and Dan Fouts. (Although the Fouts info is really spotty, so who knows if he could have been ranked higher.)

Mark could certainly argue with the way I came up with my formulas, but based on the ones I used, all the talk of Brady being better than Manning is based on reality.

*PASS-passing stats
ATT-attempts
COMP-completions
YARD-yardage
TD-touchdowns
INT-interceptions
RUSH-rushing stats
FUM-total fumbles
AY/A-adjusted yards per attempt
RW-regular season wins
RL-regular season losses
R%-regular season winning percentage
PW-playoff wins
PL-playoff losses
P%-playoff winning percentage
W%-weighted winning percentage (each playoff game is worth two regular season games)
projR%-projected regular season winning percentage for those QBs with missing stats
projW%-projected wighted winning percentage for those QBs with missing stats
Wrank-weighted rank, with winning percentage rank being worth twice as much as other ranks