Unrealistic with speed drops in CFM

[DeuceDouglas]

Quote: Originally Posted by oneamongthefence

Maybe we can appeal to EA to add a regression slider? Set the age for each position to start regressing and how much. Or a dynamic regression system. Where there would be fast normal or slow. And guys could change based on injury, goals met etc. So a player starts with a development rating, When he hits a certain age it turns into a regression trait. That way a player is dynamic through his entire career.

Definitely would be on board with this. Have it so that at somewhere between 25 and, let's say, 30 their developmental trait transitions into a regression or longevity trait which determines how quickly and/or how much they regress. Unfortunately, that would only help alleviate half the issue with the other half being how much players regress on such a consistent and completely static basis.

I think the one article that shows how current NFL players would have to be rated in their late 20's to follow their regression system really exposes how bad it is. A guy like Steve Smith would be in the 100+'s for nearly every attribute at 28, where he'd already have been on the decline for two seasons as well. Brady would have to have 114 THP to match what he has now at 39. Overall there just needs to be far more variance and a less consistent punishment for certain players when it comes to aging. I'd also like all of this to have the option to be hidden so there's more of a challenge but that's a battle for another day.

[ggsimmonds]

Quote: Originally Posted by ajra21

actually, this is not accurate at all. sorry. it's been proven that when it comes to speed & acceleration, players peak in the very early twenties, declining around 24/25yo. sprinters overcome this decline through rigorous training that NFL player to not partake in. for the vast majority of NFL players, their latest sprint times will come during draft season. in general terms, players peak (physically) around 25-27yo in most sports. they maintain their performance by understanding and developing their awareness of how to play the sport they play. guys likes green and young were insane (speed) athletes at their positions. young is on record saying he he was no longer overly fast by the time he was starting for the niners. instead, he learnt angles, jukes etc that enabled him to run past opponents. only last week on KNBR radio he mentioned knowing he was struggling to run around guys during his second season as the starting QB. jerry rice got behind D because his route running & cuts were exceptional and he played in an O that was built on route deception. no DB knew where he was going to cut. equally, his downfield behind D threat was never his best skill because he wasn't overly fast. he was very good at quick movements & finding angles that prevented opponents from getting to him. he often beat CB deep with double moves and not speed. the regression in madden is relatively accurate to real life. improve the players' AWR, PRC & INJ high enough and they'll maintain their OVR. however, there should be more diversity on when different positions and players drop off. it's not perfect and because it's so different from what madden players are used to in past games, many believe it's not a fair reflection. our understanding of physical peaks should not be set by previous madden games. edit: moss' post-NFL career 40 time was dubious to say the least. it was hand timed with other factors that should make us raise an eyebrow.

I agree with everything you said except the most important part. Speed does not decrease realistically in Madden. My starting corner had 91 speed when he was 29. Now at 31 his speed is 84. That does not happen without a major injury.

[ForUntoOblivionSoar∞]

Quote: Originally Posted by ajra21

actually, this is not accurate at all. sorry. it's been proven that when it comes to speed & acceleration, players peak in the very early twenties, declining around 24/25yo. sprinters overcome this decline through rigorous training that NFL player to not partake in. for the vast majority of NFL players, their latest sprint times will come during draft season. in general terms, players peak (physically) around 25-27yo in most sports. they maintain their performance by understanding and developing their awareness of how to play the sport they play. guys likes green and young were insane (speed) athletes at their positions. young is on record saying he he was no longer overly fast by the time he was starting for the niners. instead, he learnt angles, jukes etc that enabled him to run past opponents. only last week on KNBR radio he mentioned knowing he was struggling to run around guys during his second season as the starting QB. jerry rice got behind D because his route running & cuts were exceptional and he played in an O that was built on route deception. no DB knew where he was going to cut. equally, his downfield behind D threat was never his best skill because he wasn't overly fast. he was very good at quick movements & finding angles that prevented opponents from getting to him. he often beat CB deep with double moves and not speed. the regression in madden is relatively accurate to real life. improve the players' AWR, PRC & INJ high enough and they'll maintain their OVR. however, there should be more diversity on when different positions and players drop off. it's not perfect and because it's so different from what madden players are used to in past games, many believe it's not a fair reflection. our understanding of physical peaks should not be set by previous madden games. edit: moss' post-NFL career 40 time was dubious to say the least. it was hand timed with other factors that should make us raise an eyebrow.

Two things:

(a) Look, saying players DON'T DECLINE IN SPEED AT ALL in their late 20s is worlds different than saying they don't decline at the rate Madden says they do. Madden makes them decline at a constant rate, which is unrealistic in the first place, and I argue it is too steep a decline early on. Consider the following:




The slowest WR in Madden has 81 speed. Thus we can call that the origin. The fastest WR is 97 speed. Thus our average of the two extremes is 89. Use that as the baseline.

Now normalize it with 81 as the origin. Then 89 transforms to 8. 8 is approximately the median speed of Wrs in Madden, with 0 the minimum and 16 the maximum.

So, after a player hits 27, they lose 2 points of speed, which is TWENTY FIVE PERCENT, with respect to the coordinates from which Madden issues speed ratings. Is it realistic for a player to lose 25% of their speed from 26 to 27? No sir it isn't.

What they ought to do instead is make the reduction of ratings a slow cubic curve, with the rate of change of reduction increasing with age. I.e., lose 1 point per year for the first three of years, then 2 points per year for the next two, then 3 points per year every year after.







(b) Where's your source? You're making a scientific claim regarding the bodies of elite athletes with access to the most advance medicine, so where's the source?

I'm sure you're in the ball park but I want to see the data. As for Steve Young, regardless of what he claimed, he was STILL fast as crap in 1995. You can look at the film and see he was very fast for a quarterback. Even in 1998 games (of which I own), he still runs pretty fast.


Anyway, Madden would fix the problem if they made the decline function a polynomial rather than a line. Start off with 1 point per year, and then after about 3 years raise it to 2 points per year, then after 2 years raise it to 3 points per year, and continue in a similar fashion until an asymptote is reached, and make the decline constant from there.





EDIT- regarding Moss: you don't need a hand timer to watch him burn people in 2007, 2008, 2009. If he had 99 speed in 1998, then by 2009 he'd have 87 speed, which is below Madden's midpoint between it's two extremes, which is ludicrous. Even in his last season in 2012 he was played with off coverage with a safety over the top- which by Madden standards would mean he would have had 81 speed. Completely inaccurate.

[timhere1970]

Quote: Originally Posted by ForUntoOblivionSoar∞

Two things: (a) Look, saying players DON'T DECLINE IN SPEED AT ALL in their late 20s is worlds different than saying they don't decline at the rate Madden says they do. Madden makes them decline at a constant rate, which is unrealistic in the first place, and I argue it is too steep a decline early on. Consider the following: The slowest WR in Madden has 81 speed. Thus we can call that the origin. The fastest WR is 97 speed. Thus our average of the two extremes is 89. Use that as the baseline. Now normalize it with 81 as the origin. Then 89 transforms to 8. 8 is approximately the median speed of Wrs in Madden, with 0 the minimum and 16 the maximum. So, after a player hits 27, they lose 2 points of speed, which is TWENTY FIVE PERCENT, with respect to the coordinates from which Madden issues speed ratings. Is it realistic for a player to lose 25% of their speed from 26 to 27? No sir it isn't. What they ought to do instead is make the reduction of ratings a slow cubic curve, with the rate of change of reduction increasing with age. I.e., lose 1 point per year for the first three of years, then 2 points per year for the next two, then 3 points per year every year after. (b) Where's your source? You're making a scientific claim regarding the bodies of elite athletes with access to the most advance medicine, so where's the source? I'm sure you're in the ball park but I want to see the data. As for Steve Young, regardless of what he claimed, he was STILL fast as crap in 1995. You can look at the film and see he was very fast for a quarterback. Even in 1998 games (of which I own), he still runs pretty fast. Anyway, Madden would fix the problem if they made the decline function a polynomial rather than a line. Start off with 1 point per year, and then after about 3 years raise it to 2 points per year, then after 2 years raise it to 3 points per year, and continue in a similar fashion until an asymptote is reached, and make the decline constant from there. EDIT- regarding Moss: you don't need a hand timer to watch him burn people in 2007, 2008, 2009.

Also, we have to remember those numbers mean a certain 40 time. So, if the slowest wr in madden is an 81, that is he runs a 40 in 4.77 seconds. That is the slowest guy playing the position in the nfl at the start of your madden franchise. Fast forward 10 years and there are guys running a 5.00 or higher with overalls still in the 80's starting. If they are going to reduce physical skills to less than high school averages they need a way to get non nfl caliber players out of the game. This would be like the show having players with speed in the 50's leading the league in stolen bases because he reads pitchers well.

[ForUntoOblivionSoar∞]

Quote: Originally Posted by timhere1970

Also, we have to remember those numbers mean a certain 40 time. So, if the slowest wr in madden is an 81, that is he runs a 40 in 4.77 seconds. That is the slowest guy playing the position in the nfl at the start of your madden franchise. Fast forward 10 years and there are guys running a 5.00 or higher with overalls still in the 80's starting. If they are going to reduce physical skills to less than high school averages they need a way to get non nfl caliber players out of the game. This would be like the show having players with speed in the 50's leading the league in stolen bases because he reads pitchers well.

What they need is a nice elongated S curve for speed declining.

The rate of decline should start out slow, then increase with time, and then approach an asymptote where the rate of decline slows again.

For example, the decline for a WR might look like this from 27 years old on:


27 -1
28 -1
29 -1
30 -2
31 -2
32 -2
33 -3
34 -4
35 -3
36 -2
37 -1

And continue at -1. What do I base this one? Production. WRs have a nice a slow decline in production until about age 34, at which time they fall off a cliff.

After that point, the WRs who last aren't doing it because of speed anyway, so there is no reason to drop their speed below that of a high school back up .



https://www.profootballfocus.com/age-of-decline-wr/



Now yes, skill and awareness compensate for loss in physical skills, but the loss in physical skill is very minor in the mid to late 20s. As time goes on this loss of physical ability increases. It is NOT a steady sloped line. It's a polynomial that starts off with a gradual slope.

[Alex the Great]

Ya'll should read these four articles from Deep Dive Gaming if you haven't already.

From Good to Great: The Cost of Developing Elite Players in Madden 17

How Players Regress in Madden 17

Madden 17 Hates Old People

Why you should only develop players under 26

Really goes into the progression and regression system and how vanilla it is. I think they finally have the right idea going, but they need to make it more dynamic to allow for much different career arcs. As people have noted, a career arc like Brees or Brady or Steve Smith, Sr. would be impossible in the current Madden regression system.

For one, adding in varying and dynamic peak start and ends would be a huge upgrade over the current system. This hard stop at 26 where everything gets much more expensive no matter the development trait is too easily gamed. The same with regression starting at either 28 or 29 for everyone. Every player should have different peak starts and ends that are not upgradeable, but can change due to performance. Ideally, in my mind, peak start would generally only change by +1 or -1, with rare cases of +2 or -2. Then peak end could change a lot, based on injuries, performance and just randomness. I'm thinking injuries could cut it drastically, while consistent good performance could steadily keep pushing the peak end back, or winning league awards later in their career could cause an increase. Poor performance would cause a chance for the peak end to bump down a year or in rare cases two.

Positions all would have different average peak starts and ends as well, because the career arcs of quarterbacks and running backs are hugely different. As ForUntoOblivionSoar linked to, Pro Football Focus has a great series on the career arcs of the skill positions.

The more cues you take from sports simulations like OOTP and Football Manager, the better. They have the deepest progression systems and nuance of any games out there. Just adapt those ideas to football and everyone will have what they want.

[timhere1970]

A lot of good ideas. They need to put some randomness in it. The biggest problem with playing franchise is the predictability of it. I shouldn't be able to sign a 27 year old wr to a four year contract and know exactly what his speed is going to be in the last year.

Most of us think the speed drops are to steep but the consistency of them is my biggest complaint. A lot of talk about people who stay fast at older ages but a lot of career archs are much shorter. Michael irvin for example or Bo jackson. Most hAlf backs are only dominant for a couple of years. Just look at the league leader charts for the last ten years and see how many different names are on tbere. I would love to see in the first week of pre season on my things to do list, something pop up saying Johnny dumb dumb lost .4 off his 40 time over the offseason and take you to the depth chart. Would love to see the beast wr I just signed suffer an injury that makes him a shell of what he was and still owe him 70 mil on his contract. Would love to see someone retire at 27 with something left in the tank.

[SpyPirates]

A lot of good points in this thread, but I'll echo some of the best points and add my own:

I tend to think that guys in their mid-to-late 20s do actually lose speed more than we realize. If they're really good, they use their speed early in their career to make a name for themselves with their elite physical tools, and later in their careers they rely more on their intangible skills. At this point we just have competing claims with no one providing any evidence, so unless someone starts linking some academic papers (or at least media articles covering them) it's all hearsay :-)

In the game, I haven't really seen significant regression at the age 28 season, though--are we sure it's not the age 29 season? Recall that your players get a year older during the offseason, so if they were 28 years old while you were playing with them in a given season, and they regress in the ensuing offseason, they actually regressed while they were 29. I've also seen regression hit fewer attribute categories in the age 29 season sometimes.

I would love to see variable regression curves. The current system reduces a lot of complexity I'd like to see when negotiating contracts because I know almost exactly what a guy's attributes are going to be in year X of the long term contract I'm signing. I wouldn't mind the curves being hidden from the user so that you aren't using information that would be unknown IRL to take advantage of a contract negotiation.

Finally, I think it's been mentioned, but players really shouldn't be losing many intangible attribute points as they age. I doubt a guy like Revis has just forgotten how to cover man-to-man; he just isn't as fast and quick as he used to be, so it's harder for him to keep up, even though he's one of the best of all time at that skill. Off the top of my head, I think only speed/agi/acc/str/throw power/inj/stamina/hit power should really be subject to regression.