Those screenshots look like Lego characters playing Football.

Bring back Legends of Football.

Without intending to sounding too negative, this game is going to suck the most enormous donkey testicles that have ever been produced by ID/Evolution.

There isn't one post on that forum that leads me to believe that this game will be anything but a waste of time.

Besides, more money has been thrown at the madden franchise than any other football game in history and they still can't get the 3d aspect of the game right... how does this game have any chance?

Folks that play these types of sim games demand absolute accuracy. There is no way on earth this guy is going to be able to accomplish anything near satisfactory simulations.

At least madden has pretty graphics... from a simulation standpoint, madden is terrible. At this point in technology, you cannot have a football sim with the spreadsheet power of FOF and an accurate on the field simulation (again,that not even the front-runner madden gets right). Football is just too complex for our slow computers.

Someday? Yes. Now? No, especially by an indie developer with only a track record of missed deadlines.

I could not disagree with this statement more.

Everything said about Maximum Football, however, I do agree with.

B.S. Madden hasn't done it yet because Madden focuses on joystick jockeys, not armchair coaches.

Hmm... i may get this game.. just to see if a 8088 processor can handle those bad ass graphix!

To pile one, you are really wrong about this.

The reasons that nobody has released a game that does both well is economics. To develop such a game would cost more money than developing a game that does one well, and they would not be able to raise the price (since the only acceptable price for a new game is $50).

Football pro 97 did both well, and was released in 96. I don't think football has gotten much more complicated, or computers slower, in that time.

I just re-read some of what SD posted on page 1 from the Maximum Football forum. I loved this from the designer/programmer/whatever:

"I posted a new build for the testers on Sunday night. Lets see what happens with that."

Gotta love the confidence!

So you are of the opinion that it is possible to accurately model 22 individual A.I.s on the field so as to be indistinguishable from the real thing?

Somebody needs to include this in a memo to the Madden Team.

I may not have the most knowledge of football, but I know it when I see it. The current state of computer power isn't enough to model the game exactly as we watch it on Sundays.

I don't want a graphical representation of the game until it can perfectly match life (and someday technology and programming will get there), otherwise-- what is the point? We nitpick with Jim and Arlie to get the text portion of it right; imagine how crazy you would be watching the game unfold in 3d on your screen and you see the AI doing stupid shit.

Maybe Jim or Arlie could weigh-in on this and tell me I'm full of shit and that we can do this with today's machines, but I highly doubt it.

I've never seen so much talk about such a ham and egg project.

Yes. Or at least close enough for gaming purposes. Heck, most of the football games out there really only have one or two annoying AI gaps, and they don't focus on AI really much at all (again, they're building AIs that are good opponents for joystick jockeys, not AIs that are good football players). If this were the focus of a game (a 3D football SIM rather than a 3D football action game), there's no reason it couldn't be very solid.

I think the only thing missing from this thread is a post by Wrongway saying "this will be the greatest selling game of all time. And Wie sucks." Sadly, that can't happen.

You may be right with the "perfectly match life" portion of your statement, but the FPS series was/is/might always be the best of both worlds. It had some problems, but it did a lot right. The series was evolving like Madden does in terms of graphics, but Sierra pulled the plug. :(

Absolutely. The reason they don't doesn't have to do with computational power so much as it does economics - the amount of time and energy to tweak the AI algorithms to approach 95% of reality is simply not worth the payoff.

And even if they did, people would still bitch. Think about every time an AI player misread/forgot/blew the play call and was wildly out of position (just like what happens every Sunday in reality) - the gamer would bitch and moan about what a "stupid" AI it is.

Now, what may also be limiting the AI is the connection to the 3D engine. There's a range of possibilities that can happen when you connect the logic code of the game to the 3D engine, ranging from the logic has absolute control over the movement of the players at the expense of the animations to the animations have complete control at the expense of the logic. Madden runs closer to the first extreme - they realize it's more important to get the player in the right spot than to have the animation run smoothly. That's why you see foot sliding and awkward looking transitions. But I think they do provide a slight amount of concession to the animations at the expense of the logic to keep the animations from looking too awful.

As game console processors get more powerful, this will allow for more sophisticated blending systems to be integrated into the 3D engines which will allow the logic code to completely govern movement in the games while also providing very realistic animations and transitions.

It is a trade off.
AI takes lots of cpu time.
Graphics takes lots of cpu time.
Which do you think Madden chooses?

I'll be the lone dissenting vote here. If I was a solo developer, I'd be pretty damn proud if these were the graphics I was able to offer.

The graphics don't bother me at all, I'd be more concerned about how well the AI and career play turns out (assuming the game is actually released at some point).

They can do it in SIGAMES Football Manager. What makes you think they can't do it in American Football.

Wrong.

Graphics are generally done with the video card now. The Processor has very little to do with the graphics.

Wrong!

I don't get it.

This is correct - most (if not all) of the graphics processing is handled by the GPU, not the CPU.

If anything, the real trick is having the players make mistakes. An AI quarterback can look at all his receivers on a play and make the highest-percentage pass 100% of the time. To be a realistic model, though, he has to make bad choices occasionally, and his reaction time has to be slowed down to something less than instant.

Hehehe. You're funny.

Maybe for actually DRAWING the graphics, but there is still a TON of setup work in deciding what to draw. Trust me, the rendering folks still want access to PLENTY of CPU for their work.

I'll concede that Soccer must be a little more complicated than just running around on a big field trying to kick a little ball in a giant goal, but please, American Football is way more complicated than Soccer.

Absolutely.

And exactly how knowledgable are you in the field of computer programming and artificial intelligence?

I don't think so. American football is only 4 to 10 seconds long per play. Soccer is continous flow.
When a player has a ball in soccer he must make a ton of decisions.

In football, a lineman either blocks or well blocks. Granted he must decided who to block. But, in general he either man blocks or zone blocks. Now there are positions that require more thought such as the QB, but in general on a play a person is given a command. He is supposed to perform the action the playbook tells him too.

Either game can be modeled accurately. I believe both would take years of tweaking the engine and input from experts in the field. Years meaning, it took SI Games about 8 years or more to accurately depict 2d gameplay. I don't believe you can say one is easier or harder then the other.

All I want is a 2d model, simple x and o for a football game. That will prove the engine.

I forget which version of Quake it was (maybe III) but John Carmack (a man that knows a little about this stuff) spoke about AI and how extremely difficult it is to code. He also mentioned the amount of CPU that needs to be dedicated to each "bot" and the number wasn't small (even for that time), and that is just for simple FPS games.

Keep in mind that this is all way beyond simple if/then logic.

Yes, I know you are a coder. I've written my share of c++ apps in my day, so, while I'm not making a living at it, I at least can look at this with some knowledge of how this stuff is structured.

I think the issue is that a FPS like Quake has to run at so many frames per second to look good (regardless of whether the GPU is doing the rendering work). If you have multiple logic/movement decisions going on, they all have to be able to occur within that frame rate timespan. Perhaps if you took a less graphics intensive appraoch like FM, you can get by with a lower framerate and have more time to run logic intensive routines between each frame.

I spend more time reading, designing and coding AI than I do any one other thing, so I know a bit about this too. And if the AI bots for Quake 3 were the pinnicle of his AI expertise, I hope he sticks to just the math and graphics engine programming (Carmack is a genius, and I do look up to him, but he's not going to be god at everything. For instance, he's just not that good of a game designer and he's even acknowledged that. And AI is more about design than it is technical.).

Besides, an AI bot for Q3 is not an AI bot for football. They are two very different things.

I have a pretty good idea as to what would be needed to create a (2D top down) simulation of football and the AI required (and yes, well beyond simple if/then logic). It's very do-able.

Out of the major sports, I can think of few that would be easier (note: this does still not mean easy) to have a fluid 2D display for than American football, except for of course baseball. Basketball/soccer/ice hockey/rugby etc. are all fluid sports that are, by and large, continous. With American football it's a down at a time, and the offense and defense are both coming out with set plays. Of course there's still a heck of a lot that goes on down on the field, but after the play is over, the whole thing resets and everyone is back in position.

Largely I agree. To have a game with the graphics of a Madden game with the financial engine of FOF has nothing to do with CPU resources, video cards or anything else -- it has to do with the fact that you are talking about double the development time and resources. You're essential designing two games.

Sierra's FPS was a hybrid -- but I'd say that rather than double development time, they spend the same amount of development time as say Madden did, but re-allocated resources from graphics to the spreadsheet side. As a result, you got a game that wasn't as good as Madden graphically and a game that wasn't as good as FOF spreadsheet wise. I think Sierra also wisely built onto the game each year rather than starting from scratch. Unfortunately, with '99 they decided to do just that and showed that you can't do high-end graphics and high-end spreadsheet within the traditional development cycle. Given another six months and another $15/game, they could have pulled it off.

You can't get a football game with Madden-esque graphics and a FOF-esque engine without adding another $15 in price at least, and the market simply won't bare that. Joystick jockeys won't pay it.

Yeah, especially because of the set plays. Which is not to argue that other sports dont have them, but in football, you'd already have a basis for where the AI is planning to run/pass the ball, so that you could work the options off of that. And in many cases, 2-3 members of the defense wouldnt even be involved in the first level of the play, so you could probably mimic their AI without much intensive use.

You ain't kidding. I call the original Quake AI a "PacMan" AI, because that's pretty much all the monsters did. It boiled down to:

Do I see the player? If so, attack!
Have I seen the player in the last 5 seconds? If so, run after him!
Otherwise, go back to stomping around in the room I was placed in.

That's pretty much it.

you mean this game STILL hasn't been released?? I thought we were assuming it would be out weeks ago!!

According to the web site, its release will coincide nicely with last year's Superbowl. Fear not.

Best. Thread. Ever.

oh i don't fear, because i have no intention of buying it. i fall in the same camp as SD really...if the game took this long to come out, i can't imagine the support for it is going to be where it needs to be for a first-gen game.

:D

Give the dude some time god damnit, Matrix is still working on the box and manual.

Ham and Eggs you say? Tell me more.

They should offer a pre-release discount like OOTP.

I finally went to the website just now and as soon as I saw the first screen shot I went "Oooooohhh...... THIS game, I remember looking at these screenshots 3 years ago and thinking that this had potential"...

FUNEX?
IFX.
FUNEM?
IFM.
IFMNX.

.

Calling someone a ham and egger back in the day meant you were a street tough.

This is great news. I wonder if it comes with the 2002 rosters?

Taking a step into fantasy land for a moment and assuming the game is released mid-February...considering the game's been in development for 5-6 years (or whatever it is now), how smart is it to release it AFTER the football season is over?

After 5-6 years, i would think just getting the thing out is something of a moral victory, whether football season is over or not.

I'm thinking of releasing a hockey sim....the day baseball season starts.

I am a marketing god.