When will Madden get the ratings right?

[NHL_MLB4]

I always feel DHB is underrated

[stlpimpmonsta]

Considering the fact that ratings are objective, they will never get them right however I hope they do thembetter in the future.

[gilczbq]

theres simply too many players and ratings for one guy

[DCEBB2001]

I do not buy into the whole "Madden ratings are too big for one person" sentiment. To me it comes down to that one guy having reliable data to drive his ratings system, no matter what that is.

Hear me out on this one:

I run a website that rates 24000+ NFL players and free agents. It is a one-man show for applying the ratings. However, the difference between a guy like me and a guy like Donny Moore is the source material and resources. I utilize real NFL scouting data that I am privy to because of my link with NFL Draft Scout. That data drives the ratings. Now, because that data in its raw form is not scaled the same way that Madden ratings are scaled, you have to use data interpolation.

That is where the other resources come into play. I sought expert opinion from NFL scouts, a physics department at a top-5 research university, one of the top-ranked kinesiology departments in the world, and numerous academic articles that have attempted to define the application of athletic traits.

The problem I think we all have with Donny Moore, who only has to rate 3000 players every year (might seem like a lot to some, but to me that is a week of work), is that he does not back up his ratings with data, primary source material, or expert opinion. Instead, it is just a guy watching Youtube videos.

Now, on Youtube you can see how a player moves in his highlights, but you do not analyze every down like a coach or scout would do. You do not take into account (rarely do I hear people say this, anyway) that the outcome of a player is in fact due to 22 players interacting with one another. I often cite the analogy on this site as watching a DeSean Jackson TD catch. Sure you can see him score a TD and think, wow that player is great! He must have superior route running ability, speed, hands, etc. because he just scored this amazing TD.

What you may not consider is the fact that it took recognition of the coverage to adjust a route, good footwork to beat a jam, a perfect throw where only the receiver could catch it, and maybe a false step or two by the defender. Instead what you see is DeSean running with nobody near him and you think "wow, he's gotta be the fastest player in the game." But is this really true?

I have professed for seemingly ages on OS about how speed and acceleration, as well as other attributes, can be quantified. We have data to support this stuff. Please, spare me the "game speed" bologna. Game Speed is usually not attributed to speed at all. Game speed is a nice term given to a player who may lack actual physical talent compared to his top-tier competitors, but makes up for the deficit with good technique and skill. Jerry Rice is the prime example. Jerry ran a 4.57 in the 40. That isn't necessarily blazing for a WR or any other skill position, but it was functional enough. Why? Because Jerry Rice is probably the only player alive who can run a route, any route, at full speed. Even though that full speed may be 4.57 full-speed, it can still allow you to turn a 1 foot cushion from a defender into a 5 yard cushion through a few steps.

Watch this video from the 4:49 mark forward.

http://www.youtube.com/watch?v=ZLjgZc0sYIc

See how Jerry Rice created separation so easily? It was not because he was faster or more athletic than those DBs, it was because his technique was perfect. He was able to run routes without having to slow down as much as other WRs and was able to accelerate out of his breaks without being lazy.

So to me, giving a guy a speed boost because he "plays faster" is entirely illogical. We know what these players can run when they are in near-equal testing settings. We also know how to interpret data and interpolate it into useful ratings. The problem with Donny Moore's approach is that he is either ignorant to "real" football evaluation or he does not have the resources to rate players effectively. When have you ever seen him say the words "data", "distribution", "source material", or "interpolation" in a tweet or interview? I think that if a football nerd like myself can get access to solid info to draw ratings conclusions from, a guy whose employer has the backing of a multi-billion dollar business should be able to as well...and some.

[GiantBlue76]

Quote: Originally Posted by DCEBB2001

I do not buy into the whole "Madden ratings are too big for one person" sentiment. To me it comes down to that one guy having reliable data to drive his ratings system, no matter what that is. Hear me out on this one: I run a website that rates 24000+ NFL players and free agents. It is a one-man show for applying the ratings. However, the difference between a guy like me and a guy like Donny Moore is the source material and resources. I utilize real NFL scouting data that I am privy to because of my link with NFL Draft Scout. That data drives the ratings. Now, because that data in its raw form is not scaled the same way that Madden ratings are scaled, you have to use data interpolation. That is where the other resources come into play. I sought expert opinion from NFL scouts, a physics department at a top-5 research university, one of the top-ranked kinesiology departments in the world, and numerous academic articles that have attempted to define the application of athletic traits. The problem I think we all have with Donny Moore, who only has to rate 3000 players every year (might seem like a lot to some, but to me that is a week of work), is that he does not back up his ratings with data, primary source material, or expert opinion. Instead, it is just a guy watching Youtube videos. Now, on Youtube you can see how a player moves in his highlights, but you do not analyze every down like a coach or scout would do. You do not take into account (rarely do I hear people say this, anyway) that the outcome of a player is in fact due to 22 players interacting with one another. I often cite the analogy on this site as watching a DeSean Jackson TD catch. Sure you can see him score a TD and think, wow that player is great! He must have superior route running ability, speed, hands, etc. because he just scored this amazing TD. What you may not consider is the fact that it took recognition of the coverage to adjust a route, good footwork to beat a jam, a perfect throw where only the receiver could catch it, and maybe a false step or two by the defender. Instead what you see is DeSean running with nobody near him and you think "wow, he's gotta be the fastest player in the game." But is this really true? I have professed for seemingly ages on OS about how speed and acceleration, as well as other attributes, can be quantified. We have data to support this stuff. Please, spare me the "game speed" bologna. Game Speed is usually not attributed to speed at all. Game speed is a nice term given to a player who may lack actual physical talent compared to his top-tier competitors, but makes up for the deficit with good technique and skill. Jerry Rice is the prime example. Jerry ran a 4.57 in the 40. That isn't necessarily blazing for a WR or any other skill position, but it was functional enough. Why? Because Jerry Rice is probably the only player alive who can run a route, any route, at full speed. Even though that full speed may be 4.57 full-speed, it can still allow you to turn a 1 foot cushion from a defender into a 5 yard cushion through a few steps. Watch this video from the 4:49 mark forward. http://www.youtube.com/watch?v=ZLjgZc0sYIc See how Jerry Rice created separation so easily? It was not because he was faster or more athletic than those DBs, it was because his technique was perfect. He was able to run routes without having to slow down as much as other WRs and was able to accelerate out of his breaks without being lazy. So to me, giving a guy a speed boost because he "plays faster" is entirely illogical. We know what these players can run when they are in near-equal testing settings. We also know how to interpret data and interpolate it into useful ratings. The problem with Donny Moore's approach is that he is either ignorant to "real" football evaluation or he does not have the resources to rate players effectively. When have you ever seen him say the words "data", "distribution", "source material", or "interpolation" in a tweet or interview? I think that if a football nerd like myself can get access to solid info to draw ratings conclusions from, a guy whose employer has the backing of a multi-billion dollar business should be able to as well...and some.

Sigh -

Reading this made me depressed. This is the kind of thing that someone in charge of ratings at EA should be doing. DCB, have you ever applied for a job at Tiburon? This is what I mean about the difference in talent and mindset at Tiburon compared to the other sports development studios. I bet DM doesn't do ANY of this. It's all about putting in the effort and actually taking it seriously to produce the most realistic results.

[ryan36]

EA loves DCB lol.

Seriously, this is the problem with a stats driven system.

Different schemes. A guy can have all the tools, and then you trade for him. IRL, he's a great fit for your offense and would become a star. In madden, because he's only had 15 rec/yr he's rated really low. There would be no "accurate" sleepers sans some wonky progression BS.

[msdm27]

Just a heads up for those who think Madden Share will help with ratings....

It's basically USELESS once you start a CFM!

Because all the incoming rookies would not be graded accordingly, thus being in a great advantage.

I asked J. Looman why rookies were rated so high on madden custom rookie classes and he mentioned that the reason for this was that otherwise rookies would have no shot at competing for starting jobs.

Since the CPU AI doesn't "recognize" that playing a 67 rated rookie 1st round pick with lots of potential might be more benefitial than starting a 75 rated veteran... Madden's solution to this is bump up the start-point ratings for rookies, thus allowing them to start/compete for jobs.

In short, modyfing the ratings pre-CFM would only result in rookies becoming monsters from the get go.

[infemous]

Quote: Originally Posted by msdm27

Just a heads up for those who think Madden Share will help with ratings.... It's basically USELESS once you start a CFM! Because all the incoming rookies would not be graded accordingly, thus being in a great advantage. I asked J. Looman why rookies were rated so high on madden custom rookie classes and he mentioned that the reason for this was that otherwise rookies would have no shot at competing for starting jobs. Since the CPU AI doesn't "recognize" that playing a 67 rated rookie 1st round pick with lots of potential might be more benefitial than starting a 75 rated veteran... Madden's solution to this is bump up the start-point ratings for rookies, thus allowing them to start/compete for jobs. In short, modyfing the ratings pre-CFM would only result in rookies becoming monsters from the get go.

YUP!

I just wish that we got rid of OVRs and roster decisions were based off of individual attributes, production, contract length and age...