Ratings scale of EA?

[Franchise408]

I see it more like a grading scale.

Yea, a grading scale goes 0% - 100%, but how many people in a class will actually get 0%?

The worst of the worst will probably be around 50%-60%.

50-59% = F
60-69% = D
70-79% = C
80-89% = B
90-99% = A

That's how I view the Madden rating scale.

[strawberryshortcake]

Quote: Originally Posted by ggsimmonds

I would prefer simply stretching the ratings out. Use the 0-100 scale, but normalized for NFL caliber players. I think setting the game to be able to create a scenario where the average Joe could be recreated with realistic results is a waste of time. Setting the floor to 0 does not require you to use the mathematical concept of 0; it can be your arbitrarily defined minimum, e.g. a speed of 0 results in a 40 time of 6 seconds. Set your scale and allow the distributions to fall where they should. The animations are just as important though. Even if the scale remains the same it would be good if they adjusted the animation thresholds/tiers. I would love to see some graphs and charts wit the current ratings and mechanics in place. Edit: Using a non arbitrary 0 would create problems across the board, not just with speed and strength. What would 0 awareness be, a comatose state? Would zero accuracy mean that the destination of a pass is entirely random? A jumping rating of 0 would mean that the player is incapable of jumping?

Then scratch off 0. Start at 1.

Zero would be the player is simply there, not doing anything. Even if entering zero in the formula returns an "error," designate "error" as simply having the player there as a moveable object on the field.

Accuracy:
100 = "small circle" pinpoint accuracy. Throwing a dart and hitting the center of the dart board.
1 = "large circle" terrible accuracy. Could very well be random. Throwing a dart and hitting wherever on the dart board or even missing the board completely.
0 = error code, designate error code as lowest common denominator which is "1." Or error code, designate all error codes as person not moving.

Jumping:
100 = highest standing vertical jump, either with or without knee bend.
1 = the person "jumps" 2 inches off the ground.
0 = error code, designate error code as lowest common denominator which is "1." Or error code, designate all error codes as person not moving.

Quote: Originally Posted by deaduck

Which is exactly the point I keep trying to make. Someone can produce all the theoretical formulas in the world saying "It's simple" but it not. Any number times zero is still just zero, any number divided by zero remains it full value. You can't set a ratings scale with zero as the bottom in a mathematical simulation game and say it's a non attempt at the activity. Even if we forget, the scaling issues of a zero output versus a top output. Why would you want to insert a player onto the field of play who is unable to move, or think, or see? It's ludicrous to expect it to balance game play or accurately test formulas/calculations.

If entering zero into the formula comes back with an error code, set "error" to equal lowest common denominator of 1. Or simply set "error" to equal object not moving.

Quote: Originally Posted by deaduck

Your formula doesn't make sense. Zero can't be plugged into it because zero can't be a factor in a division of a activity. I think you keep avoiding the question of setting a zero scale for any of these measurable activities because,when you do, the math will show a rough guideline similar to how EA get's to their own ranking. Your initial contention that how EA stops their lowest ranking at a activity within the game at a 40/50 ranking is still a valid complaint but only in the frame of that logic. You can't introduce a zero output or even a realistic "average" scale at the activities and not expect the upper scale to condense into further elite numbers. I go back to the punters blocking example. Yes, the scale allowing them to play O line is silly but you can't say they are incapable of blocking within the scope of the game. The majority of them having played football at the highest levels gives them training and skill set that doesn't translate to a ZERO blocking. That would be standing still and not slowing the momentum of anyone traveling into/ through them. Beyond faulty math/ physics, it's just goofy. A physical object has to impede in some fraction a opposite force.

Blocking example.

Zero = If enter zero returns an "error" code, designate error code as lowest common denominator of 1. Person is simply "there" as an object on the screen. The "1" rated object can still provide some sort of friction/impediment to incoming forces because this "1" rated object is still present.

"1" rated player will simply flop backwards when a #99 rated player collides. All NFL players are at least rated #50, showcasing blocking abilities.

A #40 rated player will block worst than a #50.
A #30 rated player will block even worst than a #40.
A #20 rated player will block even worst than a #30.
A #10 rated player will block even worst than a #20.
A #5 rated player will block even worst than a #10.

As we trend downwards, it'll be ridiculously easy for the #99 - #80 rated players to get pass a #40 - #5 rated player. Touch a #5 rated player and he says "ow/ouch." Doesn't mean a #5 rated player is any current NFL player, just that this #5 rated person is an object on the field that provides a small inkling of impediment to a defensive tackler/rusher.

[Shards]

one of the main issues in madden is that we rate based on last year. If you want to have an accurate franchise no player should be rated above 95 in anything unless they are possibly the greatest at that attribute. I'm not concerned with overall because that is a strange calculation that is mostly based on awareness.

Here is a basic example using speed since that has been the main attribute. Just because a player ran the fastest 40 time ever does not mean he should be a 99. In theory no one should ever be a 99 because someone could eventually be faster. Of course I'm considering 40 yard dash the measurement of speed even though we all know running without pads is not a measurement of speed and acceleration is definitely a part.

If we consider Bo Jackson's supposed 4.12 40 yard dash and Terrance Cody's 5.71 the known extremes. The next step would be to pick the possible range. I think 4 seconds as the fastest and 6 seconds as the slowest is fair. Are there slower people out there? Yes. Will they even be athletic enough to play in the NFL? No. Using that 2 second range makes the math simple.

6 - "40 time" X 50 = Speed Rating

That would make Bo a 94 and Cody a 15 if you round up. This leaves room and spreads out the ratings. Using the full 0-100 would allow players to be rated properly so guys like Barry Sanders could have mid 90's agility and accel but a low top speed while still being a dangerous weapon.

[deaduck]

Quote: Originally Posted by Shards

If we consider Bo Jackson's supposed 4.12 40 yard dash and Terrance Cody's 5.71 the known extremes. The next step would be to pick the possible range. I think 4 seconds as the fastest and 6 seconds as the slowest is fair. Are there slower people out there? Yes. Will they even be athletic enough to play in the NFL? No. Using that 2 second range makes the math simple. 6 - "40 time" X 50 = Speed Rating

Shards....here is the problem with your formula.

The actual worst 40 time in the electronic scores era is Isaiah Thompson's flat 6.0 score.

According to your formula, he has a zero speed rating. Meaning anyone who ran under a 6, has a negative rating, and in a simulation, a BACKWARDS motion when engaged in a formula. That can' t work.

By the same token, Strawberry want's to enter a hypothetical bottom ONE rating which he will not identify because as soon as you choose a quantity for one and set a highest observable speed at 99 ( say at 4.24 a tied record in the electronic era) then the data becomes unworkable. The average range of NFL 40's is 4.55 for specialty positions ( WR/CB) to 5.36 average for guards. The entire league's 40 is within a average .81 range. That's a very tight window.

[ggsimmonds]

Quote: Originally Posted by strawberryshortcake

Then scratch off 0. Start at 1. Zero would be the player is simply there, not doing anything. Even if entering zero in the formula returns an "error," designate "error" as simply having the player there as a moveable object on the field. Accuracy: 100 = "small circle" pinpoint accuracy. Throwing a dart and hitting the center of the dart board. 1 = "large circle" terrible accuracy. Could very well be random. Throwing a dart and hitting wherever on the dart board or even missing the board completely. 0 = error code, designate error code as lowest common denominator which is "1." Or error code, designate all error codes as person not moving. Jumping: 100 = highest standing vertical jump, either with or without knee bend. 1 = the person "jumps" 2 inches off the ground. 0 = error code, designate error code as lowest common denominator which is "1." Or error code, designate all error codes as person not moving. If entering zero into the formula comes back with an error code, set "error" to equal lowest common denominator of 1. Or simply set "error" to equal object not moving. Blocking example. Zero = If enter zero returns an "error" code, designate error code as lowest common denominator of 1. Person is simply "there" as an object on the screen. The "1" rated object can still provide some sort of friction/impediment to incoming forces because this "1" rated object is still present. "1" rated player will simply flop backwards when a #99 rated player collides. All NFL players are at least rated #50, showcasing blocking abilities. A #40 rated player will block worst than a #50. A #30 rated player will block even worst than a #40. A #20 rated player will block even worst than a #30. A #10 rated player will block even worst than a #20. A #5 rated player will block even worst than a #10. As we trend downwards, it'll be ridiculously easy for the #99 - #80 rated players to get pass a #40 - #5 rated player. Touch a #5 rated player and he says "ow/ouch." Doesn't mean a #5 rated player is any current NFL player, just that this #5 rated person is an object on the field that provides a small inkling of impediment to a defensive tackler/rusher.

Why?

Why should EA implement players that are that low skilled? You want them to waste time making animations for guys jumping two inches?



It is a monumental waster of time.

I am all for stretching the current ratings but when it comes down to it, it is nothing more than an aesthetic change. What you are suggesting is much more than that.

I guess I am in favor of changing the scale but you wish to change the range.

Quote: Originally Posted by deaduck

According to your formula, he has a zero speed rating. Meaning anyone who ran under a 6, has a negative rating, and in a simulation, a BACKWARDS motion when engaged in a formula. That can' t work.

That is incorrect. A value of zero does not need to be absolute. Madden can simply say a 6 second 40 is what we have established as the cutoff. So they assign that as a zero. A 6.10 40 would still be a zero. It would not create any backward motions.

Right now Madden's cutoff is 40 or 50. If the change the scale and set the floor to a arbitrary 0 nothing would really change. The gameplay would be identical to what we see now. Guys that are rated in the low 80s would drop in their ratings, but it would not change the on field action.

[deaduck]

Quote: Originally Posted by ggsimmonds

That is incorrect. A value of zero does not need to be absolute. Madden can simply say a 6 second 40 is what we have established as the cutoff. So they assign that as a zero. A 6.10 40 would still be a zero. It would not create any backward motions. Right now Madden's cutoff is 40 or 50. If the change the scale and set the floor to a arbitrary 0 nothing would really change. The gameplay would be identical to what we see now. Guys that are rated in the low 80s would drop in their ratings, but it would not change the on field action.

You're missing a large part of the debate where post 6.0 speeds were spread across the scale but a hypothetical bottom 40 yard speed was never set...so until you set it, it has to be assumed as zero.

Secondly, if you do set 6.0 as the bottom of the scale...accept 4.24 as the top, you have a 1.76 second variance in all speeds. Then every .17 seconds represents a point differential in speed on a 0 to 99 scale. That means every cat who runs between 4.41 to 4.58 gets an identical 97 speed rating. That could easily effect play depending on the formulas in game. I don't how EAs scale runs but I am pretty sure the variance in all WR/CB speeds is greater than 3 points on a 100 scale.

[khaliib]

Quote: Originally Posted by PGaither84

This was an area I focused on at the Madden 11 community day years ago. I performed the "punters and kickers as linemen" test for them live in their office and that caused a stir in the building. They said a memo went around and a temporary fix was made that afternoon to try and address some of the problems. I am too lazy to put in my old copy of madden 11, take out my hard drive and play he core game without updates, but as I recall, the game launched with code that allowed players with a large (40 point?) ratings difference to suffer a "Super win/lose" animation something like 95% of the time in the internal dice rolls. I know that change, and a change in the blocking logic were both direct changes made to the game because of my input, yet in later title updates were removed. To be fair, maybe they weren't "removed" but were "broken" by other updates. What ever the case, it was (for a short time) part of the game. How much did that matter? Not much at all. Starters and back ups never really see something like 40 point differential in comparable ratings to even see this matter. A bigger problem is in the core game itself. Unlike ESPN 2k5 or All-Pro Football 2k8, Madden doesn't have a system where different animation play based on he players internal ratings numbers. Oh, and yes, before anyone says there aren't ratings in APF2k8, that is not true. You just can't see them. Anyway, the point is that in those games, the motion capture team recorded multiple a variety of NFL players, including some stars to represent realistic movement (compared to Madden hiring actors), and they also got a variety of examples between elite players and average players. That is reflected in the game. A player with, let us say, 70 pass blocking doesn't block any worse than an elite 99 rated lineman. He doesn't have worse technique or anything. he just loses the dice rolls more often. Of course the line interactions have gotten a lot better over the years. If you don't think so, go put on a copy of Madden 09 and Madden 25 and you can see and feel the differences. (I don't have M15). Those interactions are still far from what we want in the end, but have improved.

I've seen the "Ratings" discussion grow and come to the forefront, which is good, but there are a couple of "KEY" aspects of this post that has been lost in many of theories/ideas/type of scales used scales etc..., and those are;

1) Animations available to be linked to the ratings in the scale.
2) Win/loss triggering of those animation/rating "interactions" during gameplay.

The fact is that the greater number of "animations" a game has at its disposal, the more fluid/realistic the bigger the sandbox is for the developers to create/expand what plays out.

A major problem is, how do you animate "bad technique"?

2nd, is that even the best players suffer from losing a battle/interaction even when they used proper technique.
So technique is not the end all, be all for deciding a winner/loser in a video game.

The way I've always coached, is that the loser of a battle will be the one who allowed other gameplay variables to affect him/her at the wrong moment during play.
Fundamentals/technique only help to minimize that "wrong moment" happening at the right time for your opponent.

How does any football video game replicate this with animations and triggering rating thresholds?

The current perception that most seem to have is that because player "A" is rated ##, player "B", who is rated lower, should not be able to have moments of winning against player "A" due to their ratings (your punter example).

When if you really think about it, the "Ratings" are more so just placeholders for "animation" triggering and/or to call up "die roll" formulas associated for those interactions.

Next, EA has to consider giving itself the room/ability to "Add/expand" animations each release.

**Maybe that's why the entire scale is not being used currently (just guessing).

Something is already in place that allows them to add/expand "Animations" on without the need to tear down or rebuild every time a new animation is injected.

I don't know, but the bottom line is that there are a lot of "moving parts" for games to do what they do during gameplay.

As with any game that utilizes some type of rating system, the best method of use is that the "Animations" drive/dictate the rating system that's being used.

Again, if the animations are not there to support the fluidness of animations transitioning, it doesn't matter what ratings system/method is used.
The game will look/play in a manner no gamer will like on these NG Gaming Systems.

I've finally realized that it's not the current scale that's the issue with gameplay fluidness and player differentiation.
It's that a "greater" amount of animations is needed for the current scale to have some type of meaning (animation triggering wise).

Your post PG just triggered some out loud thinking!!!

My battle cry going forward is "more animations" in/at every phase of the game!!!
I screaming, "go overboard" with it!!!

The truth is, is that EA can never inject too many animations and the gaming community would fault/argue them for doing so (well some would just because it's EA lol), nor would it hurt the game.


Anyways...,
Bring on M16, I got a sweet tooth that needs to be satisfied!!!

[deaduck]

Quote: Originally Posted by khaliib

I don't know, but the bottom line is that there are a lot of "moving parts" for games to do what they do during gameplay.

I concur completely. I keep repeating that I have no idea how EA arrives at this stuff. I just refute that "fixing" it is as easy as most people suggest. Playing devil's advocate towards theories suggested shouldn't be construed as me saying I have the answers or even agree with the question.

I happen to enjoy Madden and don't mind it not being a unidentifiable amount closer to real/perfect. In fact, I am more inclined to want more fictional aspects and choices ( ie. relocation cities and logos).

That's just me.