Does it bother anyone that...
 

...if I submit a gameplan for a game with 100% runs, the opposing coach will not adjust and put 9 or 10 men in the box to stop the run? They will just consistently get pounded for 4 yards a carry. To me it's just unrealistic, I mean in the NFL if a team came out and ran the ball every single time, the opposing coach would adjust and stop it very quickly.

Football is a game of adjustments. The chess match is one of the greatest aspects of the sport. But has it been lost in FOF?

It may be that the AI can adjust, but not to values so extreme that they would never occur IRL.

Is it realistic for the opponent to know what gameplan you're going to run when he gameplans for the matchup? Because that's what it sounds like you want to happen.

I'd say it's more realistic that if you run 100% of the time, the computer adjusts during the game and your run attack become less effective. ...and based on what I've read, I believe that 's what happens now.

Well what I mean is, if you go to an NFL game and through three quarters Team A has ran the ball 53 times and thrown it 0 times, wouldn't Team B's defense pretty much stuff the run for no gain or minus yards every time?

I haven't found that the cpu adjusts to 100% run gameplans during games, either. I did it for an entire season and went 10-6. That wouldn't happen IRL.

Try it for a few more seasons, I'll be shocked if you make the playoffs again.

It seems the Ravens are trying your gameplan in real life though.

It depends. If you've run it 53 times and thrown it 0 times, but you keep putting two wide receivers split wide, what are you going to do, not cover them? Then maybe you'd throw an easy TD to one. Sure, they'll cheat more towards the run, no question.

You're correct, you wouldn't go 10-6 in real life with a gameplan like this, but would you see a gameplan with 100% runs in real life either?

At the end of the game, everyone knows that Team A is almost certain to run the ball, but Team A sometimes is still successful at yet. It may not be 100% certainty but at least 75% in the last 2 minutes.

Todd

If you decide to play the game in an unrealistic fashion, you are going to get unrealistic results.

You say "that wouldn't happen IRL" to going 10-6 while running the ball 100% of the time. You are right. It also goes the other way. The AI isn't Godlike. It isn't going to be able to stand up to every extreme gameplan you throw at it.

TroyF

Here is what I got with 100% run

Yeah I thought there was code to prevent this from happening. When you call a play too often, don't you see a "the defense looked awfully familiar with that play" tag?

Edit: Well here's your problem:



If you hadn't had those 59 pass attempts, clearly you woulda gone 10-6!

FTR, the word from Jim on this is that the AI can adjust pretty well to fairly extreme (maybe 85-90%) plans, but it's a problem to code for the AI to correct for punks who try to do anything and everything within or not within the realm of possiblity to beat the engine.

~rpi-fan

Here is where I'm going with this. Stay with me now! Say I am in a multiplayer league. There is a guy in the league who is in first place and all he has is 3 gameplans. Gameplan A, which is 100% runs, Gameplan B, which is 50-50 run/pass, and Gameplan C, which is 100% pass. How do you gameplan against a guy like that? If you try and guess with him, there is only a 33% chance that you will guess right. This means that the other 67% of the time, you will guess wrong and he will probably win. And as we all know, a .667 winning % is not bad.

Now, this guy is doing what it takes to win. It may not be realistic, but if a league is set up, the goal is to win the league. I'm one of those competitive types where, as long as it is within the rules of the league, do what you can do to win. That's the goal. And even if you set up a rule for the league that says "try to stay realistic", you just know there will be people in the league who will bend the rules just enough so that they will win. And in most cases, that is why the decent hearted people in the league will lose. And what kind of a league is that?? Rewarding people for bending rules?? This is why I just wish it would be built in to the game. In-game adjustments, they're a big part of football but they seem to have been forgotten. The ole chess match. I love that. As far as Easy Mac's league, I don't know how you finished 0-16, I hope that is the norm; and it would make me feel better. But it could be because you have Dallas' subpar o-line and backs. The league I went 10-6 in, I had Ahman Green and a very good o-line. But even with Ahman and a great o-line, nobody would go 10-6 running the ball every down of their season.

I don't know, maybe I'm just a perfectionist. Maybe I'll never be satisfied. I just was wishing for in-game adjustments because I know there will be people who submit unrealistic-enough gameplans to rip off the multi-player leagues.

SC - Easy, gameplan your D against the 50/50. Just because someone runs every play or passes every play doesn't mean they will always get 1st downs or tds.

I remember playing a Tecmo Super Bowl season once with the Bears and having Neal Anderson run about 98% of the time. I think I led the league in rushing by about 2000 yards. The sad part is that my friend was in the league with me and during head-to-head matchups wouldn't ever pick the right run. (Plus Anderson was at one of those Hulk stages where his ratings were all close to 100) He ended getting injured at the end of the season and I had to start counting on some scrub like Wendall Davis.

This is what happens when I set the Rams to 100% pass:

Here is my redo of the running 100% with a good running team, the Steelers (Bettis is great in the game):

Hey now, some scrub like Wendell Davis was a stud WR at LSU! And if it weren't for a freak knee injury, he would have been mentioned in the same breath as Tim Brown and Sterling Sharpe. Seriously he was one of the greats pre-injury.

Chubby, see I don't know if that would work. I would think that a team with 100% runs or pases would beat a 50-50 defense most of the time. But that example I just came up with in 5 seconds. Imagine what someone hellbent on winning could come up with when they have 5 weeks to think up something.

I'm just worried that in multiplayer leagues people will abuse the fact that there are no in-game adjustments.

and now a tewam at exactly 50, that is about average in both passing and running, the Bills:

So it seems that playing 50/50 on offense gives you a higher chanceo f winning that doing straight 100%.

Ok you guys are starting to make me feel a little bit better. 100% runs and passes pretty much only gave y'all a 7-24-1 record. I guess my 10-6 was in fact a freak thing. I'm gonna try it again when I get the chance.

So maybe in-game adjustments are there. Any takers?

SC,

A ton of problems with your worry.

Lets say a person does use three game planes. 100% run, 100% pass, 50/50.

You CANNOT assume that he'll be able to do all three with the same type of success. If he has a good O-Line, no QB, below average wide outs and a great RB, how well do you think the 100% pass option is going to work?

Before giving that explanation, I also need to know how solid my defense is. Do I have a great D-Line? Can I get pressure without blizting? Are my LB's good run stuffers? My CB's? How well do my guys diagnose plays?

You have one season where your team ran the ball 100% and went 10-6. Another person here just posted an 0-16 effort. We still do not know how the strategy would hold up long term against the computer AI. Against a human AI, my guess is you'd be murdered.

TroyF

Another thing. Are there game logs that we could look at in multiplayer leagues? Like if I'm about to play the Rams and I notice that they throw it every single time on 1st down, and I could scheme my defense to look for the pass on 1st down.

That's the thing I like about football simulations. The game of adjustments and scheming.

Yeah, chances are your commish would download Fido's soon to be released sweet util that turns logs into html. Then they could just post it to the website.

Easy Mac, you and friends have officially re-sold me on FOF2K4. And as it happens, that 10-6 all run season was flawed from the start: Turns out that I had the "Coaches re-do gameplan after every game" option turned on. What an idiot I am. Now that I've done some re-simulating here in the last few minutes, I'm getting 0-16 records coming out of my ears. This is great.

Let the chess matches begin!

The is one other crazy thing that just might work to stop extreme gameplans. The commish could just not allow extreme game plans to be submitted. The league could come up with some standard rules (so no more then 75% run in this situation or less than 15% in this situation)

It'll take leg work, sure, but a good quality multiplayer league takes leg work.

I was hoping that FOF2k4 would sort of do the "leg work" for you as far as extreme gameplans are concerned. And it seems they have. It looks like extreme gameplans might not work the way I was worried they would work.

Isn't Wendell Davis the dude who blew out both his knees on the same play on the Veterans Stadium turf? I remeber going to the vet in 1983 for picture day and it was so crowded that you really couldn't see in front of you, I was on the field a good 20 steps before I realized I was on the field, I thought we were still on the concrete stepping, It freaked me out that people actually played on that.

I remember walking on the turf at Veterans Stadium too. I couldn't beleive people played on it either.

Didn't the Raiders start off with something like 21 straight runs a game or two ago?

In-game adjustments are a part of football now?

Wow. You certainly haven't been watching enough Texas Longhorns football games.

Extreme game plans are the kind of things that a human opponet will absolutely crush. It might work in one extreme (say running) for a week or two - but if you try and swing to the complete opposite extreme (say, passing) it is really doubtful that you will have the personnel to do both in an on-line league.

Besides - if you keep running the same plays in FOF5, you get the old, "the defense seemed very familiar with that play" line.

hehe... I just simmed a season with 100% running and I have 2 backs over 1500 yards.

Oh, I went 1 - 15.

Funniest thing to me is that every player was shown to be content the entire season (1 or 2 disgruntled at most). Explains that performance has zero to do with attitude, but rather simply the fact that they are playing. Interesting, yet expected.

Went 4-12 the second 100% Rushing season and only 2 guys over 1000, the top RB coming in 2nd in the league with 1421.

I don't see 100% gameplans being used for anything but self abuse.

zing.

I think what SCJenkins81 is asking for is something like this:

Your team's defense should stick precisely to your game plan in the 1st quarter. But as the game goes along, the game plan should slowly change to match the proportion of run/pass actually executed by the opposing offense. This would be a weighted thing, with the weight given to the original game plan decreasing as the game progresses.

For example (keeping this really simple) let's say that the weight of the original game plan should be 100% on the first snap, decreasing to 60% by the end of the game. Your defensive game plan is initially 50/50 run/pass, but you play a wacky offense that runs the ball 100% of the time.

In the 1st quarter, your defense sticks to the original plan of 50/50 run/pass because the "weight" of the game plan is 100%.

By the end of the game, the original game plan's "weight" is only 60%, with the other 40% mirroring the opponent's actual play calling so far in the game. So, late in this example game, your defense would call plays as:

Original Game Plan: 50/50 run/pass, weight=60%
Opponent's actual calls: 100/0 run/pass, weight=40%

So late in the game your defense would call a run defense 70% of the time. (50% run defense weighted 60%, plus 100% run defense weighted 40%).

Of course it would have to be more complicated that this in the real game, because you can't simply look at the opponent's offense in terms of just run/pass. The opponent might have run the ball a lot because he found himself in a lot of short yardage situations. So you wouldn't want your game plan to adapt to this "pattern" when the opponent is in a long yardage situation. So likely the algorithm would have to track the run/pass breakdown in all the various combinations of down and yards-to-go, much like the game plans are organized. And only apply this "adaptive" weighting if a statistically significant number of plays have been called at that particular combination of down and yards-to-go.

It would be neat if you could set the maximum "weight" you want this adaptation to receive by late in the game. In my example I used 40%, but you could set it to anything you want, including zero which would indicate that you want to stick entirely to your original defensive plan and not be "adaptive" at all.

In a nutshell, it would be as if your AI coach is trying to reverse-engineer the opponent's offensive game plan by observing the opponent's playcalling, and then weights this information into his own defensive gameplan as the game progresses.