Mass Roster Attribute Revamp

[oklandr8rs]

Backstory: I downloaded FOF8 a few weeks ago and ran through several seasons with default rosters to get the hang of what to expect, as far as how in-depth and the randomness of how games/seasons will go. I wasn't happy with the final season stats and I wanted to create a better roster, with updated contracts/players. I took this time to try to develop a method for quickly trying to change end-of-season stats. After about 50 hours to testing and hundred of creating rosters, running a couple seasons, comparing stats, making tweaks, and starting the process over, I have come up with a roster I am happy to play with. I understand that it isn't perfect by any means, but the game's engine seems to be severely outdated with today's game.

Method: I started by trying to ensure than the majority of quality players are within the game and somewhat the correct position, focusing on OL. I then decided that I wanted to rate each player within 5 talent categories, Superstar, Good, Replaceable Starter, Backup, Roster Filler, and did so by creating a spreadsheet matching players with their respective PFF scores. Each position has 5 possible attribute ratings, filtered by talent categories. I had a house rule that if a player was named on an All-Pro list within the last 2 seasons they were automatically awarded Superstar, although most of the Superstar level players were already considered Superstars based on PFF scores.
- The reason I decided to only have 5 tiers was because even if I individually updated every single player, the game will still mess with the ratings enough to not make it worth it. Therefore, I am separating them enough to let the game's random aspect still have it's effect while hopefully keeping most of MY tiers together.
- All QBs are set to prefer long passing and minimum scrambling, with the exception of scrambling QBs.
- Punt/Kick Returning attributes are automatically middle-tiered unless they were All-Pro for Returning/Special Teams duties.
- Dropped passed are out of control regardless and don't seem to be affected by the rating, therefore I used it to help differentiate tiers of receivers for Contract salary purposes.
- Kickers and Punters are slightly different, but I don't want to go into it too deeply.

After the rosters are created I simulated 10 seasons to gather information. I included my latest copy of results.

Goals:
- League Pass Attempts: 18,500 +/- 300 (18,200 - 18,800)
- League Comp %: 64% +/- 1% (63% - 65%)
- League Pass Yards/Attempt: 7.2 +/- 0.2 (7.0 - 7.4)
- Passers with 4,000+ Yards: 10 +/- 2 (8 - 12)
- League Sack %: 6.2% +/- .5% (5.7% - 6.7%)
- League Rush Attempts: 13,500 +/- 200 (13,300 - 13,700)
- League Rush Yards/Attempt: 4.3 +/- 0.1 (4.2 - 4.4)
- Rushers with 200+ Rush Attempts: 15 +/- 5 (10 - 20)
- Rushers with 1,000+ Yards: 11 +/- 4 (7 - 15)
- Receivers with 100+ Receptions: 8 +/- 3 (5 - 11)
- Receivers with 1,000 Yards: 23 +/- 6 (17 - 29)
- League Kick Returns: 950 +/- 30 (920 - 980)
- League Kick Return Yard Average: 22 +/- 2 (20 - 24)
- League Punt Returns: 880 +/-50 (830 - 930)
- League Punt Return Yard Average: 7.9 +/- 0.6 (7.3 - 8.5)
- League Field Goal %: 82.9% +/- 2% (80.9% - 84.9%)
- Kickers with Seven 50+ Yard Attempts: 4 +/- 2 (2 - 6)
- League Average Score: 23.8 +/- 1 (22.8 - 24.8)
- Players with 10+ Sacks: 15 +/- 4 (11 - 19)
- Players with 6+ Interceptions: 4 +/-2 (2 - 6)

Total Success:
- Receivers with 1,000 Yards: 23 +/- 6 (17 - 29) 21.6
- Kickers with Seven 50+ Yard Attempts: 4 +/- 2 (2 - 6) 4.2
- Players with 10+ Sacks: 15 +/- 4 (11 - 19) 14.3
- Players with 6+ Interceptions: 4 +/-2 (2 - 6) 4.4

Mild Success:
- League Sack %: 6.2% +/- .5% (5.7% - 6.7%) 8.3
- League Rush Yards/Attempt: 4.3 +/- 0.1 (4.2 - 4.4) 4.42
- Rushers with 200+ Rush Attempts: 15 +/- 5 (10 - 20) 18.9
- Rushers with 1,000+ Yards: 11 +/- 4 (7 - 15) 14.7
- Receivers with 100+ Receptions: 8 +/- 3 (5 - 11) 11.6
- League Kick Returns: 950 +/- 30 (920 - 980) 979.8
- League Kick Return Yard Average: 22 +/- 2 (20 - 24) 24.1

Needs Improvement:
- League Comp %: 64% +/- 1% (63% - 65%) 62.3
- League Pass Yards/Attempt: 7.2 +/- 0.2 (7.0 - 7.4) 6.6
- League Sack %: 6.2% +/- .5% (5.7% - 6.7%) 5.3
- League Punt Return Yard Average: 7.9 +/- 0.6 (7.3 - 8.5) 8.9
- League Field Goal %: 82.9% +/- 2% (80.9% - 84.9%) 85.4
- League Average Score: 23.8 +/- 1 (22.8 - 24.8) 21.4

Not Sure If Fixable:
- League Pass Attempts: 18,500 +/- 300 (18,200 - 18,800) 19187.4
- League Rush Attempts: 13,500 +/- 200 (13,300 - 13,700) 13212.8
- League Punt Returns: 880 +/-50 (830 - 930) 1288.8

1. The game seems to allow for an average of very roughly 71 offensive plays per team per game. In the NFL, in 2020, the league averaged 64 and it was the most ever. I am hoping that maybe I can find a way to shift to more rush yards and it will eat some time down. That said, if the game doesn't naturally eat more clock after rushes, then this won't do much.
2. I am having problems with forcing QBs to either complete passes at a higher percentage or throw deeper in order to cut down on pass attempts some. I have no way of knowing if the fault lays with QBs or receivers since the game doesn't track YAC based on team.
3. The default coaches simply are not aggressive enough to lower the punting attempts by much, I don't think. Hopefully finding a way to up completion rate helps out.
4. I think I can fix all of the stats that I have labeled under "Mild Success".
5. The "Needs Improvement" section, I think I can fix them, but they come with a risk of messing other things up.
6. The game seems to create random undrafted players that jump into the league with higher ratings that some of my "Good" tier. Not much I can do about that. It happens even with a maxed out 2500 player roster.
7. I haven't messed with the off-season contracts too much, but they don't seem to be TOO broken, outside for the fact that pretty much every team will use their Franchise Tag.
8. I haven't tried these in multiplayer yet, but I imagine they would be better there than with the CPU in charge.

File: shorturl.at/gvCRS

Please try a season or two out and let me know what you think or if I left off anyone.

[finkellll]

I feel like the biggest thing you are missing are the playbooks and game plans.

Yea, attributes matter, but do do coaches, gameplans, cohesion, affinities, injury settings, and probably some other factors I'm forgetting.

I can already tell you that none of this matters in multi-player. Most leagues are based on how well an owner can gameplan, assuming competitive rosters.

Getting the right bars on the players so league wide production is correct stops working as soon as game plans are changed from purely AI controlled.

Bars and production are correlated, but not guaranteed. Too many different modifiers for a strong correlation.

[oklandr8rs]

Originally Posted by finkellll

I feel like the biggest thing you are missing are the playbooks and game plans.

Yea, attributes matter, but do do coaches, gameplans, cohesion, affinities, injury settings, and probably some other factors I'm forgetting.

I can already tell you that none of this matters in multi-player. Most leagues are based on how well an owner can gameplan, assuming competitive rosters.

Getting the right bars on the players so league wide production is correct stops working as soon as game plans are changed from purely AI controlled.

Bars and production are correlated, but not guaranteed. Too many different modifiers for a strong correlation.

Is there a way to force CPU teams to use playbooks and game plans, or up the AI to put players in better positions for success? If not, then attributes are the only user adjustable aspect, right?

I will likely never play multi-player, and an sure there are others like me, so I figured I’d release what I was using.