An AREA of AI and CPU that hasn't really been discussed...

[KBLover]

I don't know about the whole morale thing.

Yeah, a 2-7 team might (MIGHT) be feeling all bad or in a severe slump for whatever reason (though there's usually some reason - not just a random roll of the dice like in sports games), however, that doesn't mean they can't play well THAT day.

I mean, the 1-12 Panthers won. The 3-10 Lions beat the Bucs in OT. Things like that should happen.

I mean, do players play "depressed" on the field because they are having a bad season? The way the Lions played all year makes me think it's not some hard-and-fast rule. It might be why the Bengals collapsed - but not all bad teams pack it in and give up.

Quote: Originally Posted by Lisa_Bonami

I do agree with you about the parody in the NFL, however, I still feel like there should be some sort of relationship between your actual Franchise games you play and the games that are simmed. There should be some sort of continuity in regards to players having good/poor seasons above or below their actual ratings when you actually play the games out.

What you're talking about is a different issue and it's one I've had with Madden since like, forever, and that's the sim engine being totally out of whack with the "live" game engine. Part of it, though, is sliders, which are not a factor in simming. I think they SHOULD be. That would help with the continuity.

I agree - it's totally screwed up. I'm like the guy who said that an 87 rated team should not likely be 2-7, unless they had a LOT of bad breaks - or everyone else is like 95, which would be a problem with the ratings being/growing too high (which probably is an issue too)

[KBLover]

Quote: Originally Posted by phant030

Add a Consistency rating, the value can be the actual percentage of the time the play plays to his full ratings potential..b/c all of the 99 this week, 98 the next can stop b/c u can alter his consistency higher so he always plays at the top of his game [ratings]. (Ex. Peyton Manning 98 OVR w/ 94 CON, Carson Palmer 86 OVR 70 CON)

I like the idea of consistency more than chemistry.

Actually, I'd like it if consistency could somehow make streaks/slumps both. Not just either slump or just full ability. So 100 consistency would make players stick very close to their abilities, but 0 consistency would make for a high chance of being way below ability OR way ABOVE ability. This could account for a career year, an inexplicably bad season, or a peak-and-valley (or feast-or-famine) season that worked out around a solid year - you just never know in that ONE game if he's going to be aces or scrub, but, over time, it tends to even out.

That kind of consistency, I'd love in games. This would make some risk-reward issues to team building.

[Burner_22]

Any Given Sunday fellas.

[BlackBetty15]

Quote: Originally Posted by Lisa_Bonami

Folks, forgive me if this indeed was discussed at some point somewhere, I just havent seen it and its been on my mind for a few versions now. One thing I have noticed, and would love some others to give me their observations too, is that the CPU's Stats and Franchise wins/losses and stats dont really match up similar to the results you observe when you play them on the field. Let me explain, In my franchise, season 3, the New England Patriots are struggling at 2-7 in the division and have some of the lowest offensive and defensive Team Stats in the League. So I expected to see this play out when my USER controlled Dolphins (stacked and modified team who were 7-2) played them....This was not the case. And I've noticed this over the years of Madden to be true as well. It seems, no matter, how good or bad an AI team is performing in the Franchise, it doesnt have any carry over to your games on the field. It seems like Madden's gameplay engine simply performs to only the user ratings and doesnt factor in any slumping, momentum, or rhythm a team is demonstrating in the SIMMED games against other CPU teams. It's like all smoke and mirrors and leads to a pretty dull and predictable performances by the AI when you play them. Back to my example above, New England had a team rating of 87, despite the fact that they were having a lousy season and were 2-7. They trounced me easily, which conformed to the notion that the AI simply factored in the team's and player's true ratings, while simply ignoring any type of slump, or struggles they were having all season against other teams. Now I just isolated this 1 example to explain my point. But I've been noticing this, both ways, throughout many of my franchise games, and over the many iterations of Madden thru the years. It's like you can throw out all the stats, and performances the cpu has achieved during the season against other teams, no need to scout hot or cold players, because it seems like it's all franchise window dressing/eye candy to make the mode feel deeper than it really is. So that hot Overall 72 rated AI rookie that has been tearing up the league during the season and making a name for himself, simply plays like a 72 rated player when it comes time to play you the USER, on the field. I was just hoping to see more of a realistic tie in and continuity with the Franchise AI results and trends carry out thru my USER vs AI franchise games to get a better sensation of realism....Im not really seeing that relationship.....Its almost as if all your USER Franchise games could simply utilize the PLAY NOW feature and obtain the same results. Am I making sense...and what are your observations? thanks Merry Christmas

I hear a on this and to some extent your right. Also figure that on any given sunday, anything can happen. I mean the sun shines on a dogs butt every now and then and it just so happened to be that way when you played them.

[Lisa_Bonami]

Im in total agreement of the "Any given Sunday" in football, however, I think what I am also trying to point out with this is actually two issues:

What I have been noticing, The first problem is the Any Given Sunday format doesn't really apply in the franchise games you play and teams all too predictably perform at their TEAM RATINGS (when you actually play them out)

And the 2nd problem, which I have mentioned at the top of the post, is that there really is no coherence or relationship between a CPU's SIMMED games and then when you play them. For example, a solid rated team of let's say the Pittsburgh Steelers, with an OVR Rating of an 87, may be having a horrible year at 3-10, or something, but then when you play them, they simply play at their OVR 87 rating, indicating no algorithm relationship between the Franchise Stats and their original team attribute ratings.


I think what I'd liked to see improved in future versions of Madden, is a better carry over and relationship between the stats, performances, slumps and streaks that get simmed out versus what actually gets played out. As well as balancing it also with a bit of that 'On Any Given Sunday' unpredictability.

To be fair here, I've noticed this issue also with other EA Sports games as well. The NHL series seems to be just as bad if not worse at having teams perform differently when you play them versus what their season long franchise stats, wins and losses indicate.

[LaW97]

i notice this and it sux

i also whoop teams that are rated higher than mine but lose games or have close games with teams rated under mine

jus weird

[KBLover]

Quote: Originally Posted by Lisa_Bonami

What I have been noticing, The first problem is the Any Given Sunday format doesn't really apply in the franchise games you play and teams all too predictably perform at their TEAM RATINGS (when you actually play them out)

Unless we see the probabilities, we have no way of saying this definitively.

I know I've been beaten or at least "played harder" by crappy CPU teams and then if I replay it, it's a walk-over. So what made them play harder the first time?

Granted, that was in M10 so maybe they changed it in M11, but I can't say it's too predictable.

Heck, do the team ratings actually reflect the strength of the team? Just like player OVR, it might not capture the right things or everything.

[Hova57]

I think each team should have an Morale at the start of the season as key players get hurt, or losses happen it should decrease the morale of the team. A player playing well for four games his morale maybe high even though his team is losing. Morale for a team should play a big part in how they operate offenses and defenses. A team with low morale and can't pass ala panthers but run well there offense run through their run game.

Also teams should have tendancies something they are trying with gameplan, but not really. If your playing the Colts you know your playing a no huddle spread team. or If you playing the Steelers you'll get a single back set with two te and suddenly they play no huddle with a banana set. its because the Steelers are a 55% no huddle team where as Colts are a 95% no huddle team.

one last thing i'm not completely sure of this, but they need to bring back cpu play recognition .