Originally posted by brodiebluebanaszak:
Originally posted by 5_Golden_Rings:
Do you know what passer rating is? Do you know how that formula works? If you did, you'd know it is very limited (much like a certain QB).
Here it is:
100*(a + b + c + d)/6
where
a = 5*(COMP/ATT - .3)
b = .25*(YARDS/ATT - 3)
c = 20*(TD/ATT)
d = 2.375 - (25*INT/ATT)
(negative numbers are set to zero, and any of a,b,c, or d that is greater than 2.375 is set to 2.375)
You can see how this not only doesn't distinguish between QB contribution and WR contribution, but it also effectively counts completion percentage multiple times, since it's weighted so heavily. Only TD% is weighted more, but then again, with the YPA term in there, it further bolsters the passer rating of someone who throws short passes (because their high success rate).
What you should do iis create your own metric. Then, apply it to all the qbs, and see how they rate. Isn't that simple? It's more informative then guessing which quarterback is better than another based on how the public will respond. That's not scientific. Arriving at a metric and applying it against objectively actually collected statistics is scientific. That's what you should be doing. And I totally support your efforts to find metrics which help evaluate the performance of a quarterback over the course of an NFL season. Well done.
I think you can get a lot of information form this metric I just made:
(Completed Air Yards + (45*Passing TD) - (20*Interceptions))/(Passes Attempted)
*
This takes into account most of the important factors, and it normalizes for attempts, while also subtracting most of the contributions of WRs, TEs, and RBs by using completed air yards instead of total yards. Now, if you want to take it a step further, you can add sacks and yards lost to that with this: (Completed Air Yards - Sack Yards Lost + (45*Passing TD) - (20*Interceptions)/(Passes Attempted + Times Sacked).
*If the QB passed for two digits of attempts, change the 45 and 20 to 4.5 and 2.0 (the coefficient on the TD and INT is to make the TDs and INTs the same number of digits as attempts. Why 45 for TDs and 20 for INTs? Because while INTs hurt you, scoring points is what wins games. If you're not throwing INTs, you're not attacking the defense. But if you're throwing too many, you hurt your team. A ratio greater than 2:1 seems ideal. These coefficients could be adjusted, of course).
.
.
Just for s**ts and giggles, here is how Jimmy and a few other QBs compare with this metric. These are just randomly selected QBs off the top of my head:
Vs. some of the best QBs.
Jimmy: (1844 + (45*20) - (20*12))/(441) = 5.68
Stafford: (2160 + (45*41) - (20*17))/(601) = 6.10
Brady: (2762 + (45*43) - (20*12))/(719) = 6.19
Rodgers: (1947 + (45*37) - (20*4))/(531) = 6.65
Mahomes: (2140 + (45*37) - (20*13))/(658) = 5.39
Herbert: (2627 + (45*38) - (20*15))/(672) = 6.01
Josh Allen: (2664 + (45*36) - (20*15))/(646) =6.17
Prescott: (2425 + (45*37) - (20*10))/(596) = 6.53
Wilson: (1730 + (45*25) - (20*6))/(400) = 6.83 (<== this is a problem... no way this asshat is the best QB)
Some controls with lesser QBs:
Murray: (1986 + (45*24) - (20*10))/(481) =5.96
Ryan: (2244 + (45*20) - (20*12))/(560) = 5.12
Tannehill: (1958 + (45*21) - (20*14))/(531) = 4.94
Heinickie: (1655 + (45*20) - (20*15))/(494) = 4.56
Mac Jones: (2027 + (45*22) - (20*13))/(521) = 5.29
And of course:
Lance: (315 + (4.5*7) - (2.0*2))/(71) = 4.8
*
Most of this is unsurprising. What is quite surprising is that Mahomes ranks so low, given his stellar TD to INT ratio. This tells me there's probably something wrong with this metric. But then again, he did struggle for a good portion of the year. However, Russell Wilson is not the best QB. So I'll work on this some more.
