Embracing Human Error: in Madden

[Ice_Cold345]

It's been brought up a little bit earlier, but I would like an aspect of a meter or something. You have this meter that goes red->orange->yellow->green->black.

Green is to show a "100%" and you get all the potential out of the player on the throw and his accuracy. By this I mean, let's say a 80 in SA means a 60% completion percentage for short passes. If you get in the green, then you get all the possible accuracy out of the player and throw a "perfect arc", but you could still mis-fire because the QB you are using only has 80 in SA.

If you hit the red-yellow area, you get a certain percentage of potential from your player. Red could be 0-30%, orange be 31-69 and yellow be 70-99. Black is the "overswing" portion of the meter. It allows your QB to throw the ball like a laser beam, but at the cost of maybe 25% of potential accuracy. The sizes of each bar could change based on a number of factors of a QB, like a clutch rating(comes into two minute warning plays), awareness, and the type of release. This could be adjusted as it's just a rough outline of an idea. A lob pass would still be in the game but it would only be affected by accuracy and throw power, similar to how it exists now. The meter would come into a play once the "Throw" button is held longer for it not to be seen as a lob pass and where you would release it, is where the ball would be released, so it's going to be a quick motion and you will need a good eye to know when to release it.

Each QB would need some sort of signature throwing motion, so when you play with Rivers and his different throwing motion should have a different meter than Drew Brees quick release. It would allow ratings to play a factor and allow the user to have that control of the player, but the player's ratings would take precedence over user control, so each player is different and how the play.

[IlluminatusUIUC]

Quote: Originally Posted by Ice_Cold345

It's been brought up a little bit earlier, but I would like an aspect of a meter or something. You have this meter that goes red->orange->yellow->green->black. Green is to show a "100%" and you get all the potential out of the player on the throw and his accuracy. By this I mean, let's say a 80 in SA means a 60% completion percentage for short passes. If you get in the green, then you get all the possible accuracy out of the player and throw a "perfect arc", but you could still mis-fire because the QB you are using only has 80 in SA. If you hit the red-yellow area, you get a certain percentage of potential from your player. Red could be 0-30%, orange be 31-69 and yellow be 70-99. Black is the "overswing" portion of the meter. It allows your QB to throw the ball like a laser beam, but at the cost of maybe 25% of potential accuracy. The sizes of each bar could change based on a number of factors of a QB, like a clutch rating(comes into two minute warning plays), awareness, and the type of release. This could be adjusted as it's just a rough outline of an idea. A lob pass would still be in the game but it would only be affected by accuracy and throw power, similar to how it exists now. The meter would come into a play once the "Throw" button is held longer for it not to be seen as a lob pass and where you would release it, is where the ball would be released, so it's going to be a quick motion and you will need a good eye to know when to release it. Each QB would need some sort of signature throwing motion, so when you play with Rivers and his different throwing motion should have a different meter than Drew Brees quick release. It would allow ratings to play a factor and allow the user to have that control of the player, but the player's ratings would take precedence over user control, so each player is different and how the play.

I agree totally. In a related discussion on the NCAA board, I was toying with a similar mechanic suggested there (http://www.operationsports.com/forum...ccuracy-3.html post #24)

[Kaiser Wilhelm]

Quote: Originally Posted by IlluminatusUIUC

Remember, in the hypothetical we are dealing with a perfect player who is hitting every input right on the head. How then are you deciding which throws Tebow hits and which ones he misses? By rolling an electronic pair of dice behind the scenes. The ratings are only affecting the probability. ie Randomly. That's a logically incompatible statement. The users are controlling the players. If Joe Perfect is making perfect inputs and the game is forcing his passes off the mark, that's hindering. If Mike Average is missing but the game is aiding him by compensating with a high ACC player, that's helping. Either the pass goes where you tell it to go or it doesn't.

There is not a single modern sports game that plays like you are suggesting. Regardless of how "perfect" somebody is in their timing in NBA, FIFA, or MLB THE SHOW, ratings and a probabilistic factor (its not random) play into the success. There is indeed a timing window, but even in THE SHOW if you hit the timing window perfectly, the ball can sail, bounce, or hang over the middle based on ratings, confidence in the pitch and energy. In FIFA, players can slip on the pitch during rain games. In NBA, the way the shooting mechanic seems to work is that even if you hit that "perfect" window, there is still a probability the player can miss based on ratings.

The ratings always act as some sort of hindrance as they should. Also, we should mention that in Madden the throwing mechanic doesn't have an accuracy window. The window is for the type of throw.

Sports games are not built around the same arcade style mechanics that fighting and shooter games are built on, the same as strategy games are not built on the same premise as sports games or arcade games.

[hanzsomehanz]

Three sprinters meet for a race.

Sprinter A perfect start.
Sprintet B perfect start.
Sprinter C perfect start.

Who wins the race?

A perfect start is just that, a perfect start out the blocks. There is still another 90% of the race to complete.

In the flow of mechanics...

One sprinter might pull up with a cramp, another might lag behind but finish strong and the third in the heat might blow away the pack with his acceleration in the first thirty meters but begin to lose gains from sixty meters forward.

Perfect mechanics are not the be all end all but nevertheless,

Someone has to win the race.

Someone has to make the shot.

Someone has to make the throw.

It is not all about perfect mechanics and by the same measure, it is also a poor shot that can go in a hoop or a poor throw that can fit thru the wheel. It depends what we are evaluating, form or accuracy?

Throwing a strike is not limited to force nor form.

Bowling simulates this law of averages very well in the hit or miss probability of a strike. A bowler with perfect form might still hit the gutter or use too much force or not enough.

Similarly, not every reciever has the same catch window and some have a wide enough window to display a propensity to bail out QBs on bad thrown (off target) balls. Recievers become like bowling pins that have minds of their own.

This is why we need true ball phsyics and simulated catch windows that display a receivers catch radius through the frame of his body and the placement of the ball and his effort to then "go get it."

As it is now on the base score of sliders, I do not see receivers that have to recover or adjust? They either absorb the ball in a way that is so unnatural the ball seems to be adjusting itself mid flight so it reaches its intended target. Other times I simply do not see the effort to reach or recover - the ball is simply deemed dead. It is not much of a variance and pales in comparison to *comp% averages and catch rate averages, both of which are 60%.

* CJ Jr finished with a 59.8 catch rate%, equal to the comp% of Stafford.

There is a clear correlation between hand success (catch windows) and balls that are outside of the reach of your #1 target.

In Football, timing extends beyond throw mechanics and users should be impacted by different reaction times when clicking through different player models, like a vehicle that has different reactions to your inputs versus another vehicle.

Sent from my SGH-I727R using Tapatalk 2

[Trick13]

Looking at QB play in Madden 13 - way too easy, even with sliders.

The "accuracy windows" have to be increased to account for over and under throws and I mean even for highly accurate QBs. I am talking about expanding them so that even Manning/Brady/Brees can - with full down left stick input - throw a pass in the dirt on a 5 yard hook route.

Tebow should be like 69, or less in all three accuracy ratings, and for the moment let's forget that his throwing motion is horribly slow (which if properly accounted for in animations should make "timing" a huge issue).

So within those ratings even if you "perfectly" lead a receiver in terms of "stick skills" his ratings should mean by their definition that his ability to hit the spot he is aiming for, even with you (the user) determining that "aiming point" is less than 70 percent (assuming All-Pro -base -50 pass accuracy slider).


That means that even if you perfectly time the throw, select the perfect "lead" on the pass, you should have, at minimum, a 30 percent chance that he doesn't hit the spot you aimed his throw at. Maybe it is lower, higher, out in front, or behind the targeted "spot" or "aiming point". Perhaps because you "aimed and timed" it so well that his miss is within the receiver's "catch radius" and the pass is still completed, but maybe the lack of accuracy that Tebow has means the receiver must break stride and loses YAC potential (he has to bend, dive, jump, slow down). And no matter how perfect your input is as the user, maybe he just flat out misses.

Accuracy for QBs in Madden is too high in my estimation, because the pass is tethered to the receiver way to closely - again expanding the "accuracy windows" to allow an intentional dirt ball or sail out of bounds - and allow for some "user error". I some times will catch a receiver flashing open and hit wrong icon button or what have you, but it is very rare to be able to over do the precision passing and that to me is poor game play design.

As the accuracy rating falls the ability for a QB to hit the user's "perfect aiming point" should fall and the "windows" for Tebow should be much bigger than say Drew Brees...

If full down on the left stick causes a one hopper for Brees on a hook route, then if Tebow user holds full down and Tebow's accuracy check means he actually throws to the very lowest end of his window - the ball should be at least 2-3 yards short and in the dirt...

[IlluminatusUIUC]

Quote: Originally Posted by Kaiser Wilhelm

There is not a single modern sports game that plays like you are suggesting.

I'm not sure what you mean there. I've never claimed any game works on the perfect input model, only that it should.

Quote:

Regardless of how "perfect" somebody is in their timing in NBA, FIFA, or MLB THE SHOW, ratings and a probabilistic factor (its not random) play into the success.

We are seemingly using two different definitions of the word "random." If I'm reading your post correctly, you are using it to mean "All outcomes are of equal probability" in which case no, no game uses that method. I agree with that much.

I'm using random to mean "without an objective" or "not the product of conscious choice." If you take a 6-sided die, and paint five sides red and one side blue, then the ultimate outcome of rolling it is still random even if the probabilities are not equal.

And that's what people are proposing to hold back Joe Perfect. Like this:

Quote:

the way the shooting mechanic seems to work is that even if you hit that "perfect" window, there is still a probability the player can miss based on ratings. The ratings always act as some sort of hindrance as they should.

and this:

Quote: Originally Posted by Trick13

So within those ratings even if you "perfectly" lead a receiver in terms of "stick skills" his ratings should mean by their definition that his ability to hit the spot he is aiming for, even with you (the user) determining that "aiming point" is less than 70 percent (assuming All-Pro -base -50 pass accuracy slider). That means that even if you perfectly time the throw, select the perfect "lead" on the pass, you should have, at minimum, a 30 percent chance that he doesn't hit the spot you aimed his throw at.

What you guys are saying is that even with perfect inputs, the game should intervene and make the player fail even when the user succeeded.

So no matter what the user does, the random number god will always act like a brake, keeping the player from being 'too good' and thus 'unrealistic.'

I think it's a major flaw to build a game on that mechanic. Randomize elements that are outside the player's control, sure. If Joe Perfect throws a pass but the wind unexpectedly gusts and pushes the pass off-course, that's perfectly fine. That's outside the player's control, so it should be outside the user's control too. But don't override the user's inputs to force them to fail.

[Trick13]

Quote: Originally Posted by IlluminatusUIUC

What you guys are saying is that even with perfect inputs, the game should intervene and make the player fail even when the user succeeded. So no matter what the user does, the random number god will always act like a brake, keeping the player from being 'too good' and thus 'unrealistic.' I think it's a major flaw to build a game on that mechanic. Randomize elements that are outside the player's control, sure. If Joe Perfect throws a pass but the wind unexpectedly gusts and pushes the pass off-course, that's perfectly fine. That's outside the player's control, so it should be outside the user's control too. But don't override the user's inputs to force them to fail.

Actually what I am saying is that Tebow's accuracy should work just like any move you attempt with a RB against a perfectly squared up Patrick Willis. If while usering AP you try to truck Willis, and your timing is perfect, the defender is in perfect tackle position and by his ratings prowess - you are most likely not breaking that tackle - maybe you get a clean "truck" once out five times in that scenario based on the ratings.

Same scenario with Felix Jones (a huge bust from my favorite team) and as his ratings are much lower his success rate at a clean truck animation - no matter how perfectly timed it is, should be more like 1 in 35...

Are you seriously telling me you believe that Felix Jones (a finesse guy, and not that good of one at that) should be able to truck Willis despite the ratings advantage Willis has and Jones' low truck rating - at a clip in the same neighborhood as AP, just cuz "the user" timed it right???

Get real man, it is the view you are expressing that has held Madden back from its true potential - you are the one who told us all you can beat the game on some respected sliders with 3 plays - actually I think you said 'beat the CPU by 50 with 1 running play' - are you happy with that? Because that is exactly what you are saying should continue.

I don't have the time to spend on Madden that I used to have, but when I was very good and the controls were pretty intuitive (PS2 layout) I used to take lesser teams with low rated players (even trading away stars for picks) to create a challenge. I fully expected my stick skills to be hampered by the lack of ratings of said lesser teams/players - that was the whole idea of taking the Houston Texans back then...

Side note - please no "meter" for passing - too much going on in a football game for meters or that happy horse hockey that Backbreaker had where both sticks were required for passing - QB evade needs to be on R Stick, I need to be able to move in the pocket, scrambling then becomes some ridiculous combo of trigger buttons held and it becomes a mess.

[Kaiser Wilhelm]

Quote: Originally Posted by IlluminatusUIUC

I'm not sure what you mean there. I've never claimed any game works on the perfect input model, only that it should.

No sports game is like that because it is a horrible model for a video game that's ultimate goal, outside of sales, is to emulate a sport. Ratings are there to make a character play or perform a certain way. Player input is just one part of the equation. That is something for arcades, not a multimillion dollar simulation.

Quote:

We are seemingly using two different definitions of the word "random." If I'm reading your post correctly, you are using it to mean "All outcomes are of equal probability" in which case no, no game uses that method. I agree with that much. I'm using random to mean "without an objective" or "not the product of conscious choice." If you take a 6-sided die, and paint five sides red and one side blue, then the ultimate outcome of rolling it is still random even if the probabilities are not equal.

The player does have input though. The player is starting that player, that was his/her input. They know the ratings. The characters are not just meat sacks for the gamer to inhabit and turn them into gods.

Quote:

What you guys are saying is that even with perfect inputs, the game should intervene and make the player fail even when the user succeeded. So no matter what the user does, the random number god will always act like a brake, keeping the player from being 'too good' and thus 'unrealistic.'

Well that is not exactly how I would put it but sure. Getting a "perfect input" is merely increasing the probability for success, but not guaranteeing it. For one, it makes the game feel more organic because a player can have an off day, even if the gamer is on his game. The game does not step in and say, "Nope you fail." However, it should be pointed out once more, that Madden does not even have "success windows" for anything other than catching and kicking. Everything else is handled by probabilities that are influenced by a number of factors, one of which is the ratings.

Quote:

I think it's a major flaw to build a game on that mechanic. Randomize elements that are outside the player's control, sure. If Joe Perfect throws a pass but the wind unexpectedly gusts and pushes the pass off-course, that's perfectly fine. That's outside the player's control, so it should be outside the user's control too. But don't override the user's inputs to force them to fail.

Its not forcing to fail though. Again, like I said above, all the user does by getting a "perfect" in the timing window is increase the likelihood for success. Each game runs on a different model, and some models actually work for the perfect = 100 but not all, and that is for the better in my opinion.