Embracing Human Error: in Madden

[KBLover]

Quote: Originally Posted by The_Rick_14

If it's the latter, then I will always expect the feedback phrase "Perfect" to coincide with "Success". Taking the example of Howard and three pointers again, to avoid the issue of players playing above ratings, a low rated three point shooter should not be able to trigger a "Perfect" out of the mechanic. Perhaps it's just semantics at that point, but that's what I expect out of a skill mechanic.

Not semantics at all, imo. It's a logical and worthy point of distinction.

Would that prevent players going above ratings? If the user was "close enough" could Howard still make the shot? If so, couldn't I become good/experienced at the game supersede Howard's rating for shooting from beyond the arc by mastering the mechanic to where I'm getting "very close" consistently, allowing me to make 'too many' threes with Howard? Would it be the "Baseball Stars problem" all over again? Or would it be like with Madden where I'm fiddling with sliders to get the ratings to express?

I can understand what you're shooting for, no pun intended. I can see it working. I think for me, it's the "perfect = success" aspect. I still want to hit that "perfect line" and still be like...did he make it? It looks good, it looks on line, but... especially in a crucial situation. Likewise, if I'm "close" but not quite - I want the "was it good enough?" "will it rim in?" Instead of knowing as soon as I push the button to get the meter result if I made it or failed.

[KBLover]

Quote: Originally Posted by IlluminatusUIUC

Now, in reality, actual trained Soldiers sometimes fumble and drop their magazine during a reload. Would anyone want a mechanic where your character randomly just goofed in the middle of a gunfight? It adds nothing to the experience but frustration.

Why not?

It happens. A trained soldier likely also knows what to do in that situation, how to buy time in cover to recover his ammo and reload, etc. Just like in any war - weapons jam, there's duds, etc.

Company of Heroes does that. If I have an MG42 team, sometimes the "burst" isn't as long. Sometimes the suppression isn't as immediate (and sometimes it's drastic and highly effective). My skill is dealing with the situations as they arise, not making WW2 go off without a hitch. Likewise, if I'm playing as USA and my Rifle squad is suppressed quickly and pinned - my skill is in either alleviating the pressure (silencing the MG), withdrawing my troops or taking other action (maybe I'll let them occupy the MG while I flank).

My Tiger tank might not hit the enemy Sherman in the rear armor. The Sherman might have an accurate shot and cause engine damage, etc. My skill is in my moving the Tiger, putting it in the best position as I can given the situation. My skill is in dealing with the situation as it changes and emerges, not making the Tiger do things it can't.

Quote: Originally Posted by IlluminatusUIUC

What then is the point of playing with any poor QB, because you'll never be able to accomplish more than they have? If you want to take a cannon armed but inaccurate passer, then you are banking on your own stick skill to account for their weakness. And why is that wrong to ask? If someone wants to spend the cap to get Rodgers, they have a much easier time of it.

Because if your stick skill is good enough, there's no difference between the scrub and Rodgers. How is that representative of the NFL? If a bad QB could consistently play like a good one, he wouldn't be a bad QB. My skill is in gameplanning around the poor QB, or developing a strategy to mitigate his ineffectiveness. My skill being 'rewarded' doesn't have to be me turning him into Aaron Rodgers. It could be me winning in spite of him.

[Big FN Deal]

Quote: Originally Posted by IlluminatusUIUC

I strongly disagree with this statement. Players should be rewarded for developing their skills with more success on the field. I agree with Rick, the ratings should determine the difficulty of the mechanic but the success or failure should be on the user's shoulders.

My issue with this POV is that the User is not the in-game player, they are possessing/controlling them. Hanz made a point earlier that essentially the User is the "driver" in Madden, not the "car". That's the simplest way to put it, in a racing sim, when in an in-game race, you are stuck with whatever car you have, including its' pros and cons. All the User can do is drive it the best the can, same thing with User control in a sports sims.

Quote: Originally Posted by IlluminatusUIUC

I haven't played NBA2K so I can't speak to that, but let's say you had a first person shooter with pretensions of realism. Lets say Battlefield 3, for the sake of argument. Now, in reality, actual trained Soldiers sometimes fumble and drop their magazine during a reload. Would anyone want a mechanic where your character randomly just goofed in the middle of a gunfight? It adds nothing to the experience but frustration. Compare that to Gears of War. Reloading is actually a game mechanic where users can develop the skill to reload faster - at the risk of jamming their rifle if they fail. Now the onus is on the user, rather than putting you at the mercy of the game everytime you try to accomplish a simple action.

You kind of lost me here. To focus specifically on bad snaps or hand off exchanges, as Kaiser just stated, just make those things rating and environment based, maybe a higher chance in the rain, in pressure situations with less composed players and just once in a blue moon period, it would add an organic feel to the game.

I have posted before that the reason the Madden community would have an issue with this is because they have been exposed so long to this ridiculous blanket win/loss stuff in so many areas of the game. There is very little process or variation in Madden, things are either done right or wrong, nuance is extremely lacking. When someone posts "botched snaps" the first thing the Madden community pictures is the ball hitting the ground or flying over the QBs head and the play being a failure. However, in real games I have seen QBs like Romo handle bad snaps, the point being, the errors we are referring to wanting is not asking for random automatic failure. It's a request for Tiburon to represent the varying degrees of performance in Madden that exist in real life, from the ideal to the potentially disastrous and everything in between.

Quote: Originally Posted by IlluminatusUIUC

Heh, play more with the Bills. Stevie makes me pull my hair out sometimes. What then is the point of playing with any poor QB, because you'll never be able to accomplish more than they have? If you want to take a cannon armed but inaccurate passer, then you are banking on your own stick skill to account for their weakness. And why is that wrong to ask? If someone wants to spend the cap to get Rodgers, they have a much easier time of it.

Not being funny but is this a serious question? In case it is, the point is the challenges the User faces and having to figure out what to do. My starting QB went down now I am stuck with a back up that has poor ratings, do I run the ball more, make a trade, search free agency, scrap the year to get a good QB in the draft, etc? You know, the same type challenges and choices real NFL teams are faced with when they have players that can only accomplish so much.

[The_Rick_14]

Quote: Originally Posted by KBLover

Not semantics at all, imo. It's a logical and worthy point of distinction. Would that prevent players going above ratings? If the user was "close enough" could Howard still make the shot? If so, couldn't I become good/experienced at the game supersede Howard's rating for shooting from beyond the arc by mastering the mechanic to where I'm getting "very close" consistently, allowing me to make 'too many' threes with Howard? Would it be the "Baseball Stars problem" all over again? Or would it be like with Madden where I'm fiddling with sliders to get the ratings to express?

"Bad", "Poor", "Close" are all scenarios where I expect a negative result to be possible. As I think about it more, "Close" or "Good" input should coincide with the player rating so if Howard is a 10% 3 point shooter, then with "Good" input he should have a 10% chance and with that low of a rating, getting a "Perfect" should be extremely difficult. One could even argue the design for a player that low should be to make a "Perfect" impossible. The only way for a system I'm describing to work is for the difficulty of the mechanic to be on point.

Quote: Originally Posted by KBLover

I can understand what you're shooting for, no pun intended. I can see it working. I think for me, it's the "perfect = success" aspect. I still want to hit that "perfect line" and still be like...did he make it? It looks good, it looks on line, but... especially in a crucial situation. Likewise, if I'm "close" but not quite - I want the "was it good enough?" "will it rim in?" Instead of knowing as soon as I push the button to get the meter result if I made it or failed.

Here we will just have to agree to disagree. I'm with you that if I'm "close" then I expect to be on edge of did he get it or didn't he? But just like professional baseball players know when they've hit a homerun, when input is "Perfect" you know immediately.

A properly tuned mechanic difficulty should result in an extremely large percentage of inputs falling into the "Close" and "Good" category even for highly skilled users.

[KBLover]

Quote: Originally Posted by The_Rick_14

But just like professional baseball players know when they've hit a homerun, when input is "Perfect" you know immediately.

No, they don't know. Sometimes they do the "trot" and the ball comes up short.

They know when they've put a good swing on the ball. They can feel the difference in contact, etc. But they don't always know it's going to be a HR. That's why they are supposed to run it out anyway. Especially guys who aren't superior power hitters where it's NEVER a guarantee where they have hit one out or not, even with excellent contact.

But let's say that was true: Does a QB know if his 50 yard bomb is perfectly on target? Does a kicker always know if he drills the kick down the middle? Does the center know that he's going to do a bad snap before he does it?

A lot of these things are effected by minor things being off. A pitcher might not put enough spin on a breaking pitch - he doesn't know that after he's snapped the ball out of his hand. As far as he knows, he might have gotten late "bite" on the ball and not threw a hanger. He doesn't know exactly how much movement he's going to get. And anyone who's thrown a knuckleball knows you don't know much of anything with that pitch that depends on randomness for it's movement.

[IlluminatusUIUC]

Quote: Originally Posted by KBLover

Why not? It happens. A trained soldier likely also knows what to do in that situation, how to buy time in cover to recover his ammo and reload, etc. Just like in any war - weapons jam, there's duds, etc.

Again, it's bad game design to randomly handicap a video game player through no fault of their own. It just adds frustration to the experience.

Quote:

Because if your stick skill is good enough, there's no difference between the scrub and Rodgers. How is that representative of the NFL? If a bad QB could consistently play like a good one, he wouldn't be a bad QB. My skill is in gameplanning around the poor QB, or developing a strategy to mitigate his ineffectiveness. My skill being 'rewarded' doesn't have to be me turning him into Aaron Rodgers. It could be me winning in spite of him.

Imagine a hypothetically perfect user. One every single pass play, he knows exactly who to throw to and when to do it. His inputs are exactly correct to the pixel every time.

Now give that player Tim Tebow as his QB. Timmy's career completion % is 48%, the highest he's recorded in a complete game is 66%. What constitutes his "maximum" performance? 66% accuracy? 48%? So what happens in your suggested scenario? Joe Perfectuser picks the right wideout, makes the perfect input and the game overrides him to make the ball careen off into the stands? How many times should that happen? 34%? 52%? A game that overrides your inputs to make you fail is frustrating enough, but if it's happening 1/3rd of the time? Why even bother playing?

Quote: Originally Posted by Big FN Deal

My issue with this POV is that the User is not the in-game player, they are possessing/controlling them. Hanz made a point earlier that essentially the User is the "driver" in Madden, not the "car". That's the simplest way to put it, in a racing sim, when in an in-game race, you are stuck with whatever car you have, including its' pros and cons. All the User can do is drive it the best the can, same thing with User control in a sports sims.

Cars are mechanical objects with measurable physical limitations. Football players are human beings and several of their ratings are simply educated guesses. Michael Vick, for example, has all the physical tools to be one of the best QBs of all time. His flaws have always his decision making. But the user is stepping in to make those decisions for him, so the user's choices should be rewarded or punished.

Quote:

You kind of lost me here.

The BF3/Gears of War examples show my design philosophy on this point. If it's something simple and rote (reloading/snaps) then either make it work every time ala BF3 or put the success or failure in the user's hands ala Gears.

People who play a video game expect that their skill will be the deciding factor. The more random elements that you add, the further the user gets from the game and the less engaging it becomes.

Quote:

Not being funny but is this a serious question? In case it is, the point is the challenges the User faces and having to figure out what to do. My starting QB went down now I am stuck with a back up that has poor ratings, do I run the ball more, make a trade, search free agency, scrap the year to get a good QB in the draft, etc? You know, the same type challenges and choices real NFL teams are faced with when they have players that can only accomplish so much.

But the user is never given the option to try to overcome this limitations with his own ability?

What you seem to be proposing is that the AI handicap users who are using a bad player. My proposal is that the AI should assist users who are using a good player. If the user controls Peyton Manning, it should be extremely easy to make an accurate throw. The margin of error should be very high. If the user is playing as Tebow, the margin for error should be extremely small, but the user should be capable of hitting every throw if they hit the mark perfectly.

[DeuceDouglas]

Quote: Originally Posted by IlluminatusUIUC

Now give that player Tim Tebow as his QB. Timmy's career completion % is 48%, the highest he's recorded in a complete game is 66%. What constitutes his "maximum" performance? 66% accuracy? 48%? So what happens in your suggested scenario? Joe Perfectuser picks the right wideout, makes the perfect input and the game overrides him to make the ball careen off into the stands? How many times should that happen? 34%? 52%? A game that overrides your inputs to make you fail is frustrating enough, but if it's happening 1/3rd of the time? Why even bother playing?

I would say yes. Have you seen Tebow play? It doesn't matter how perfect the player playing the game is, the player he is playing with is FAR from perfect. The ball doesn't have to be careening into the stands but it should be consistently high, low, in front of, behind of receivers because he sucks. The ability for YAC should be severely limited because of the off target throws. That's what ratings are for. It's no different than playing a baseball game and throwing a 'perfect' pitch and watching it get hit 450 feet. It happens.

What would stop Johnny Perfect from taking a kicker and turning him into Dan Marino?

[IlluminatusUIUC]

Quote: Originally Posted by DeuceDouglas

I would say yes. Have you seen Tebow play? It doesn't matter how perfect the player playing the game is, the player he is playing with is FAR from perfect. The ball doesn't have to be careening into the stands but it should be consistently high, low, in front of, behind of receivers because he sucks. The ability for YAC should be severely limited because of the off target throws. That's what ratings are for.

So the game is pre-determining that 34% (at minimum) of his passes will be wild throws no matter what you do?

Pre-determined outcomes are really what people want?

Quote:

It's no different than playing a baseball game and throwing a 'perfect' pitch and watching it get hit 450 feet. It happens.

Well obviously the batter has a role in that, just as the receiver has a role in whether the catch is made. If Joe isn't switching to his wideout, then he's putting his faith in the AI.

Quote:

What would stop Johnny Perfect from taking a kicker and turning him into Dan Marino?

The kicker wouldn't be able to throw as hard or as far as Dan Marino because he's not as physically strong. Just as a Joe Perfect couldn't make Michael Turner run as fast as CJ Spiller.