I know that coaches salaries are tied to attendance. What happens when a team tanks? Does attendance take a commiserate hit (it should if it doesn't)?

What thing that may discourage super tanky behavior is to tie the GM's job to attendance in the same way. If a team choose to tank, then attendance should also tank, and the player controlling the team should be "fired". In that way, the player would be forced to take over another open team in the league, and if there were no open teams in the league, then the player would be out of that league.

However, that raises a question about where to draw the line in terms of tankiness. What about teams that are otherwise competitive, but still tank because they are missing key positions. Those teams will end up losing, but not by the margins we see here. So maybe if a team loses multiple games by 50+ or 60+ during a season the player is automatically "fired."

One possible solution I am considering consists of the following restrictions:

1) Restrict player position changes to positions that the player has similar scores. So you'd be able to move a TE to the OL if he has good blocking skills and only drops his future rating by, say, 5 or 10 points, but wouldn't be able to move a FB to QB and drop his rating by 30 points. (The TE to OL in this case wouldn't be able to stair step down, the metric it would be measured by would be the best-rated position for that player)

2) Restrict depth chart assignments to the same restrictions as #1

3) Add a test on game day where if your roster doesn't meet a certain standard the AI will take it over. The standard I'm thinking of is that your active roster can be upgraded with, say, 10 players available in free agency - i.e, your best QB is rating 50 but there is a 70-rated QB in free agency. If there are 10 players in free agency that are better than players in your roster, you're tanking.

#1 is actually going to come sooner rather than later. I'd love to get your feedback on #3.

Regarding #3, I'm assuming it would use the owner's weights, as their could be some considerable differences between those and the defaults. Is this correct?

Also, I think the difference in rating should have to be dramatic. Just because a player has a higher rating doesn't mean he's a better fit for a particular team, or scheme.

Also, by active roster you are including backups. What happens if a team has some scrubs on the bench simply for cap reasons? Or players who are not highly rated in their positions, but are good on special teams? Does it make a difference if the players are starting or backups?

I think there is a lot to consider with this before anything is implemented. I am already not a very big fan of the AI's decision making, and I would hate to have it do something to my team that I didn't want.

What if the team is cap tight? I have a team that another owner ran to the dirt im carrying a huge % of deadcap after releasing like 10 to 12 30 something rated players but can no longer make any moves until some of its gone. Then what rating are you going by If team A has low rated huge potential players coming up would he be penalized for having good future players because their ratings havent been built up yet. The i think there should be a bit of a grace period after taking over a team.Personally i think just moving to a lottery would be a better option Even though i hate them personally

It would have to be using the default weights, otherwise it would be too easy to work around it by mucking up your own weights



That's why I'm thinking it would have to be more than just a few players. I'll have to play with the point where it kicks in - probably will run the algorithm against teams in various leagues just to see what it would do.



By active roster I'm basically meaning that it ignores inactive players (so you can't get around it by having good players and making them inactive). It also would only consider the best player at each position.

[/quote]I think there is a lot to consider with this before anything is implemented. I am already not a very big fan of the AI's decision making, and I would hate to have it do something to my team that I didn't want.[/quote]

The last thing I want to happen is the AI take over and make decisions for an owner who is not trying to game the system, so I agree that it would be a delicate path. It's probably something that won't happen soon - there are some smaller ideas that I hope will influence away from tanking that are not nearly as extreme as this - and like I said I will probably take some time to test the algorithms against real teams and see what would happen to make sure it's not too aggressive.

I'd use future ratings to evaluate, not current, or possibly even some calculation that considers both. If the cap is tight, the AI would be under the same restrictions you would be, so if there's not enough cap room he wouldn't be able to make any moves - _however_ - if there is a high-rated free agent available on a game stage he'll sign for the minimum salary if it's the only offer on the table, and any time you would cut a player he'll leave at least the minimum salary available in your cap.

He could also leave a future cap hit if there was any bonus money left.

Why worry about this when there are so many other problems with the game?

I'm sorry this sounds like absolute garbage, and I would probably encourage any owner that this happened to to quit playing the game. I won't state all of my reasons because most have been stated above, but I will directly comment on the default weights versus custom weights issue.

I use a lot of custom weights for the following reasons:

1) The default weights **** pretty hard, especially after the QB under pressure code was implemented. Despite the QB pressure code issue, the default weights will overvalue trainable attributes while undervaluing the non-trainable ones. Why would I want a RB with 100 break tackle, 100 ball carry, 100 route and catch when his top speed is 10? The default weights are going to put him as one of the best backs in the league and he will never get more than 1.5 ypc if that. This is why lower overall players tend to perform as well if not better than their counterparts at key positions. I ran the ball for 3000 yards with two RBs with overalls in the mid-60s because they had decent AC and SP and a few other key skills.

2) I want to run a specific defensive scheme so I value only attributes needed for that position. What do I care about zone coverage if my CBs are always going to be man up? What do I care about M2M coverage if I play single deep coverage almost exclusively?

3) I tend to have cap money to spare and don't reach for FAs. I'm typically good at cap management. I don't want this to be looked at as a bad thing because I'm not filling my roster with all of the available players and wasting cap. If I'm winning, I'm winning with the roster I have. If I'm not winning, maybe there is a reason why I'm not signing additional players.

4) Not all losing, even if you know you're going to lose, is tanking. If you have a team that is eliminated from the playoffs and you have a veteran QB (or other position) that is close to retirement, why not start playing the young guys to get them game experience and scout them. This will probably lead to some losses, but tell you a lot about your team.

5) Finally, this doesn't address the root cause of super tanking, which is just filling your roster full of the worst players in the league. Part of this cause is how the game generates players. Why are there players in the FA pool that have stupid low non-trainable skills? So say you have a WR with 10 SP, 10 AC, 5 ST, and 15 INT, but has 20 or 30 on hands. Hey that sounds like me, but you don't see the NFL banging on my door. Why are these players even available? I would favor having a floor for non-trainable skills per position. That way what actually separated the elite players from the scrubs were their trainable skills.

/rant/

To be honest, I havent see a lot really to prove me that higher rated players are better. Still looks like to me that QB at 70 with high accuracy are as good as 85 QBs. Same with D-Line. A few 60s RBs do very good.

I agree with most of what you are saying here.

In general, I don't like the idea of the AI making personnel decisions for me.