Madden NFL 10 Blog: Player Ratings - A New Philosophy, A New Era

[RogueHominid]

Quote: Originally Posted by Layoneil

there are too many players in the NFL for an emphasis on signature ability/animation packages. it's too much work for too little benefit. what signatures will do is pigeonhole players into using only those teams that have signature players. it's also difficult to apply to franchise mode when you start drafting players. how would they come about getting signature abilities and animations? do they start off with them out of the gate? do they progress into them? how would they progress into them? does every drafted player with a certain stat progress into them? would it be random?

Not sure I agree here. It's actually less work than giving every player 20 individual ratings, as individual players may be characterized by one or two abilities, some by five, and some by none at all (there are tons of nondescript guys in the NFL).

The issue of the yield/benefit being too little is questionable. The benefit is an emphasis on on-field capabilities, not a set of abstract numbers that have to be plugged into an equation, balanced, and then used to trigger a result that is often times boggling.

As for pigeon-holing players into using teams with the highest number of players with the highest number of abilities, from what I understand it, people online only run with one of 4 or 5 teams anyway, so it's not like that scenario would get worse. And in fact, if real signature styles were carried out, some of those teams might be less popular.

And to your last point about the draft, you would have to begin from the notion that the current system is somewhere close to functional. It's not. The rookies from NCAA have created all kinds of problems, and however they're generated, it's not like the importation of draft classes produces the kind of balance you're looking for. Signature abilities would have to be distributed in some way, sure, but the "problem" is the same for numerical ratings, and has yet to be solved using that system.

So while I think there are valid objections against this system, I don't see the logic to any of the ones you raise.

Honestly, I think it's a matter of taste--people have been so long accustomed to numerical ratings that they're loath to see them go, even if they don't work properly or don't make the most sense given what the team wants to do.

[acts238shaun]

Ian, I was thinking about this exact thing the other day watching Larry Fitzgerald leave the entire Steeler secondary near the end of the Super Bowl. I was curious as to each players 40 time, namely because Fitz 'separated' from the Steelers pursuit like an alley cat fleeing a dog pound. Later on I played Madden 09 and timkered with a few of the players to see what the difference was in the ratings and how each rating affected the player. I gave Carnell Williams 99 speed and agility, but zero acceleration. I left Graham's speed alone and gave him 99 agility (to be equal) and acceleration. I went into practice mode (offense only at first) and ran a few sweeps and iso's with each one to see if there was any difference between the two backs, and there a slight difference. I inserted the defense and the difference between the two magnified. Graham was visibly slower than Williams, whether it was on a sweep or between the tackles. I left practice mode and changed Graham to 99 speed, thinking that his 99 speed and acceleration would be better than Williams's 99 and zero. Nope. They were exactly the same as far as in-game. Simming games for a season each as the starting back, Graham's ypc was great and Williams's wasn't (had a 2 yards avg.). I practiced with defensive players the same way. Some with limited speed and maxed acceleration and some at no acceleration and maxed speed. The defenders with max speed caught guys with higher acceleration just the same. Corners stayed with receivers, linebackers tracked down backs, and the line rushed exaclty the same. Excluding agility, the only movement based attribute variable that mattered during gameplay was the speed category. To sum it up, acceleration during gameplay means nothing, as the speed rating on Madden 09 seems to be both the initial speed and maximum speed at the same time. The player runs as fast at the start of a play as he does at the end, assuming he runs a straight line. The agility rating does significantly effect turning and cutting.

I think a rating category called "speed" is too generic. I think back to the old SNES Tecmo Bowl games and there were ratings for running speed, max. speed, agility, and acceleration (which meant in that game how fast they got from initial to max speed). I'm not saying that was the greatest way to do it, but it gave me an idea to break up the different applications of speed.

A guy may run a 4.2 40, but what is he in the 20 yard and 60 yard dash? If he is 2.5 seconds in the first 20, that means he ran the last 20 at 1.7 seconds, showing good acceleration. At that pace he would run it in around six seconds. The first twenty yards reflects take off and initail burst, the second twenty acceleration from initial burst to top speed and the third would be actual straight line speed. Of course how often do football players get to top speed or run 60 yards? But that is a completely different subject.

Basically, break up the speed rating itself into two diferent categories. Have one rating for beginning/initial speed and another for top speed. Let acceleration be the measure of how fast a player takes to get from beginning to top speed. I think by doing that you will see more of a difference between the Earnest Graham's and the Chris Johnson's. You will also see a difference in how a defender 'closes' on the ball carrier, how the pass rushers pick up speed and how someone like a Ray Lewis or Warren Sapp can catch a 'faster' ball carrier from behind or beat them to the sideline. In that instance acceleration would matter. Oh, and for what it's worth ditch the turbo button concept in Madden 10.

Anyway, I looked up the 40 times on the Steelers defensive backs and on Fitzgerald to see if the 40 time mattered on that play. Fitz ran around a 4.5-4.7 various times while much of the Steeler secondary ran 4.2-4.6 (Ike Taylor ran Deion-like sub 4.2 if I remember). I guess Larry Fitzgerald hit the turbo button on that play, lol.

[rhombic21]

Quote: Originally Posted by youALREADYknow

No matter what garbage they choose to show you in game menus, there will ALWAYS be numbers that define a player's performance in video games. These are computers we're talking about people. Will people please stop asking for numeric ratings to end. I've even seen people foolishly ask for a "star" rating system in place of the 1-100 scale. What makes someone think that a 5 point scale will be more accurate than a 100 point scale? Do you really think there are only 100 variations of speed between all football players, let alone 5? Just stop the rating system bashing.

I agree with this post completely.

Even with 2K's system, there's some sort of underlying numerical ratings that is happening, it's just not shown directly to the user.

The problem with both Madden and NCAA is essentially the fact that they have 20+ ratings, but that a lot of those ratings don't seem to actually do anything to effect play on the field. The numerical system is not the problem. The problem is that the ratings are unbalanced, in terms of which ones matter.

[BezO]

Quote: Originally Posted by thudias

After an injury, lets say a QB. If you sign one off the street will his overall be at it's lowest point based on not knowing the offense? ---- Any player should not be signed off the street and be able to come in and start for a team. (unfamiliar with the playbook, conditioning etc. etc.) ----------------------------------- Answer: Players off the street will struggle coming on to your team if they don't know the playbook. There is a chance they'll make mistakes during the game and they won't play as well as they would if they knew the play. We have the same core player ratings. What we changed was how the overall rating was calculated. Several grades (like real NFL scouts use) now make up the overall rating. They are Athletic (physical skills like speed), Intangibles (mental skills like awareness), Durability (how healthy a player is), Learning (how quickly a player learns and retains information), Potential (how good the player can get if everything works out in his career) and Size (does the player fit the classic mold at the position). We also use a different system to determine a player's value to the team, which in conjunction with their overall grade, determines how valuable they are in free agency or trades. Value is determined by overall, upside (remaining potential), contract value (is he overpaid, paid just right or underpaid) and playbook knowledge (how much does he know of your playbook). From HC developer's Q and A thread. There is actually one that goes into more detail of player value but I couldn't find it.

Sweet! Yeah, Madden needs ALL of this.

Thanks

[Layoneil]

Quote: Originally Posted by The Autumn Wind

Not sure I agree here. It's actually less work than giving every player 20 individual ratings, as individual players may be characterized by one or two abilities, some by five, and some by none at all (there are tons of nondescript guys in the NFL).

it seems like you just don't get it. these are video games processed by computers, like the other guy said, there will always be numbers. what you're asking for won't remove the numbers, it will simply add a level of complexity on top of the numbers.

and they'll end up hiding the numbers from the players so that we won't be able to edit and analyze them. you can't make a video game without formulas that govern how one thing interacts with another. you can't have formulas without numbers.

what you're asking for is for the game to have "weapons" matter and nothing else. any player without a "weapon" is a clone that's completely interchangeable.

[youALREADYknow]

Quote: Originally Posted by Layoneil

it seems like you just don't get it. these are video games processed by computers, like the other guy said, there will always be numbers. what you're asking for won't remove the numbers, it will simply add a level of complexity on top of the numbers. and they'll end up hiding the numbers from the players so that we won't be able to edit and analyze them. you can't make a video game without formulas that govern how one thing interacts with another. you can't have formulas without numbers. what you're asking for is for the game to have "weapons" matter and nothing else. any player without a "weapon" is a clone that's completely interchangeable.

Exactly... and a "weapon" is still a number. 1 for on, 0 for off. The way that number is converted into an "ability" in the game like speed or power would only allow for a small fraction of the options compared to the possibilities that exist in the 1-100 scale.

It's amazing that nearly everyone wants real world physics calculations in Madden but some of this same group wants kindergarten level mathematical probabilities in an area as important as ratings.

[frankrizzo380]

Quote: Originally Posted by bgizle

I definitely agree that Henderson's Route Running needs some work, but I don't think he has the worst hands in the league or anything. He is the Saints 'big play' receiver, because of the speed and agility. Devery Henderson -2008 Targeted: 57 Catches: 32 Drops: 2 The top of my head, when it comes to WR's, I think straight to Terrell Owens, who is usually the perennial leader on Dropped passes. But since he's such a physical specimen, he's a beast.

thank u for pointing this out sir, for obvious reasons!, i have seen dev henderson drop a few passes, not more than t.o., 1998 nfc div game against green bay he drop the first 4 out of his five passes, dev will catch the deep ball, now the short ones he will have trouble with, but henderson dose not get targeted as many times as t.o. thats y i believe hes rated as he is in madden

[RogueHominid]

Quote: Originally Posted by rhombic21

I agree with this post completely. Even with 2K's system, there's some sort of underlying numerical ratings that is happening, it's just not shown directly to the user. The problem with both Madden and NCAA is essentially the fact that they have 20+ ratings, but that a lot of those ratings don't seem to actually do anything to effect play on the field. The numerical system is not the problem. The problem is that the ratings are unbalanced, in terms of which ones matter.

Yes, there are hidden components in other engines. However, I don't think that the basic equivalency being drawn here means that both ways of doing things are equal. With signature abilities, there are animation packages unique to each that manifest themselves during gameplay. These animation packages individuate players.

If you just go on a 1-100 scale and ditch the animation packages that link to specific abilities, then you lose a degree of individuation.

I don't get why people get so uppity about a suggestion like this. It's not a 2K vs. Madden thing; it's a what's going to give you the best opportunity to make the distinguishing characteristics of each football player visible thing, and that's something the design team cares about.

I do "get it," and I'm not looking for "kindergarten" football. From the early information on Madden '10, it's clear that there's a commitment to continuing to work from the 1-100 scale, and that's fine by me. I do think that it would help, though, to augment the work on ratings with some work on signature animation packages that help players stand out, even if it's as simple as some animations for power backs that finesse backs don't get, some for possession receivers that deep threat guys don't get, etc.