[desertrats2]

I've made a little bit of research compiling passing data over a 40-season span from real NFL and 8.2/8.3 versions.
I'll analyze Att/Game, % Completed, Yards/Game, Yards/Catch, Yards/Attempt, TD %, Int %, TD-Int ratio and QB Rating.


METHODOLOGY

Compiled passing data from:

- Real NFL (1980-2019)
- 8.2 (40 seasons in a row)
- 8.3 (40 seasons in a row)

Then I divided real NFL seasons in 8 groups of 5 seasons each (1980-84, 1985-89 and so on), trying to reflect changes in the approach to the game over time.

Didn't do the same with FOF8 versions as background is static and standard deviations are 4 to 5 times less significant than in real NFL across history.

So we have 120 seasons to have a look over.


RESULTS


1. Differences between 8.2 and 8.3 versions (80 total seasons)


- Attempts/Game: 8.2: 35,8; 8.3: 36,3 (+0,5)

QBs throw more in 8.3. About 0.5 attempts more per game. The best 19 seasons att/gm-wise were run on 8.3. Only 3 of the last 20 were run on 8.3.


- % Completions: 8.2: 62,87%; 8.3: 60,87% (-2%)

QBs are consistently less accurate on 8.3. The best 8.3 season ranks 28th over 80 total and has a lower completion percentage than the average on version 8.2. (62,56%). The worst 30 seasons were run on 8.3.


- Yards/Game: 8.2: 241,7; 8.3: 231,4 (-10,3)

Although this stat does not weigh in QB Rating, it is clear that QBs throw less yards in 8.3. The best 8.3 season ranks 29th now, and again, the worst 30 seasons are 8.3 related.


- Yards/Catch: 8.2: 10,75; 8.3: 10,47 (-0,28)

Not one of the biggest differences, but again, the best 8.3 season ranks only 26th, and the worst 30 belong to 8.3 aswell.


- Yards/Attempt: 8.2: 6,76; 8.3: 6,37 (-0,39)

It is almost a 6% drop, which is a lot. Only two 8.2 seasons are worst than the best season run on 8.3, and just by a mere 0,02.


- TD %: 8.2: 3,81%; 8.3: 3,39% (-0,42%)

Quite a big difference here again. The secuence repeats over and over, and this time the best ranked 8.2 season comes in 32nd place.


- Int %: 8.2: 2,30 %; 8.3: 2,30 % (equal)

No real changes. Made a projection here, and have QBs in 8.2 thrown as many attempts as in 8.3, the difference over 40 years would have been +10 interceptions in 8.3.


- TD-Int ratio: 8.2: 1,66; 8.3; 1,47 (-0,21)

Pretty clear having checked TDs and Ints before this.


- QB Rating: 8.2: 85,75; 8.3: 81,08 (-4,67)

Pretty clear also, as all categories but interceptions have dropped from 8.2. The best ranked season is only 30th with a 84,7 rating, trailing 8.2 average by 1 QBR point and 8.2 best season by 5,1 (89,8 QBR).



2. Which stats contribute the most to the QBR drop?

I made some projections through ceteris paribus assumptions in order to determine which stats weigh the most in QBs posting worst QB Ratings, ranked from highest to lowest:

- % Completions: +1,67 QBR

Surprinsingly, YpA is not the highest contributor to the QBR drop. Taking % Completion percentage to the same level of 8.2 version would result in a 1,67 rise of QBR.


- YpA: +1,60 QBR


- TDs: +1,40 QBR


- YpC: +0,70 QBR


- Ints: +0,01 QBR


If we change both YpA and Comp % QBR would increase 3,27 to 84,35 QB Rating, still trailing 8.2 version by 1,4.


3. How do 8.2/8.3 versions compare to real NFL?

As I said before, real NFL stats have been changing dramatically over time, and 1980 stats cannot compare to 2010 in anything but maybe YpA and TD %. Because of this, I divided real NFL seasons in 8 groups of 5 seasons each to render better comparisons. Therefore we have 10 groups: 8 real NFL related, and two more FOF related (one 8.2, one 8.3).


- Att/Game:

8.3 36,3
8.2 35,8
2015-19 35,0
2010-14 34,6
1995-99 33,4
2000-04 32,7
2005-09 32,6
1985-89 32,0
1980-84 31,4
1990-94 31,4


- % Completions:

2015-19 63,28%
8.2 62,87%
2010-14 61,13%
8.3 60,87%
2005-09 60,47%
2000-04 59,07%
1990-94 57,39%
1995-99 57,16%
1980-84 56,06%
1985-89 55,03%


- Yards/Game:

2015-19 252,1
2010-14 246,2
8.2 241,7
8.3 231,4
1995-99 225,3
1985-89 224,5
2005-09 224,4
1980-84 222,3
2000-04 221,8
1990-94 214,8


- Yards/Catch:

1985-89 12,7
1980-84 12,6
1990-94 11,9
1995-99 11,8
2010-14 11,7
2000-04 11,5
2005-09 11,4
2015-19 11,4
8.2 10,7
8.3 10,5


- Yards/Attempt:

2015-19 7,20
2010-14 7,12
1980-84 7,08
1985-89 7,01
2005-09 6,88
1990-94 6,84
2000-04 6,79
8.2 6,76
1995-99 6,75
8.3 6,37


- TD %:

2015-19 4,48%
2010-14 4,37%
1980-84 4,30%
1985-89 4,14%
2000-04 4,05%
2005-09 4,04%
1995-99 4,00%
1990-94 3,84%
8.2 3,81%
8.3 3,39%


- Int %:

1980-84 4,34%
1985-89 4,00%
1990-94 3,45%
1995-99 3,23%
2000-04 3,23%
2005-09 3,06%
2010-14 2,75%
2015-19 2,35%
8.3 2,30%
8.2 2,30%


-TD-Int Ratio:

2015-19 1,90
8.2 1,66
2010-14 1,59
8.3 1,47
2005-09 1,32
2000-04 1,26
1995-99 1,24
1990-94 1,11
1985-89 1,03
1980-84 0,99


-QB Rating:

2015-19 89,95
2010-14 85,79
8.2 85,75
2005-09 81,88
8.3 81,08
2000-04 79,65
1995-99 77,73
1990-94 76,82
1980-84 74,52
1985-89 74,27


Conclusions:

8.2 version can be compared to 2010-14 in QBR, but with a way more conservative approach (more completion %, more throws and safest interception % vs less YpA, YpC and less TD %, resulting in a similar TD to Int ratio).

There are no stats too off from a given real NFL 5 year span but maybe Att/Game and YpC, and I can see the compiled stats as an evolution (towards carefulness) in the game like any other across NFL history.

8.3 version, on the other hand, presents more flaws. QBR is similar to 2005-09 but twists have gone the wrong way in my opinion. TD%, YpC and YpA were already at the limit in 8.2. and went low too far from any 5 year real NFL span. It looks like as a conservative twist also, but with no reduced Int % and lower completion percentage the whole picture seems incoherent and unreal.