When will Madden get the ratings right?

[Kushmir]

I agree that simple statistics do not tell the whole story and can be extremely misleading. Which is why you have to delve just a bit deeper.

Totally agree that some things you need a scouts eyes for. Combine times, how well a WR blocks, raw strength and a QBs arm are all things that a scout can tell you. But IMO their rating should be 90% production. Because many of the things you mentioned hardly matter when guys cant perform on sunday. Football is full of Paul Bunyan tales and other hyperbole where Kyle Boller can throw through the the uprights from one knee at midfield...Every draft there are Superman workout warriors that show great athleticism but turn back into Clark Kent when the pads come on and someone blows a whistle.

And I wouldnt worry about more relevant stats becoming available. This is the information age, they're coming...great sites like PFF provide a wonderful look into tracking production with data. 15 years ago no fans cared how how many PBUs or TFLs an elite CB or LB had. Now these can be found in a number of places. Knowing how many pancake blocks a RG had and how many sacks or TFLs he gave up will be here shortly...be patient.

And to answer your question, THATS how you'll we'll be able to generate those ratings. When we see a FB who leads the league in pancakes, has the fewest missed blocks and gives up zero TFLs it will be pretty easy to give him a rating based on performance...dont you agree?

Mohammed Sanu? Derek Anderson Hmmmm...I doubt you were being serious. I'm pretty sure even the most basic of rating systems subscribes to stat minimums to disqualify guys with a tiny body of work...kind of illustrates my point regarding knowing WHERE to look and looking INTO the numbers instead of AT them.

Lastly, your Montana/QB Accuracy comment kind of drives home my point. "A QBs completion percentage, for example, is not only based on the ability to throw an accurate, catchable ball, but it is also dependent upon the WR actually hauling the thing in!" sadly, Montana WAS and IS one of the most accurate passers in NFL history...The math eventually tells the truth. While I agree that football is a very teammate dependent sport, these are PROFESSIONALS and the accurate QB in your example wont always have bums catching the ball. They will get cut/benched and eventually he will be throwing to guys who CAN catch and the fact that his placement is uncanny will be reflected in his numbers. NFL leaders in accuracy over the last three years? Brees, Brady and Rodgers...see how that works? Talent and Production always win out....every elite rated guy who is underperforming isnt always hurt/in a bad scheme/the victim of bad coaching.

Sometimes they were just overvalued or of average talent level when compared to the rest of the league.

[justud]

and the endless logic bombs just keep on coming. this guy is hammering u DCEB let it go bro.

[DCEBB2001]

Quote: Originally Posted by Kushmir

I agree that simple statistics do not tell the whole story and can be extremely misleading. Which is why you have to delve just a bit deeper. Totally agree that some things you need a scouts eyes for. Combine times, how well a WR blocks, raw strength and a QBs arm are all things that a scout can tell you. But IMO their rating should be 90% production. Because many of the things you mentioned hardly matter when guys cant perform on sunday. Football is full of Paul Bunyan tales and other hyperbole where Kyle Boller can throw through the the uprights from one knee at midfield...Every draft there are Superman workout warriors that show great athleticism but turn back into Clark Kent when the pads come on and someone blows a whistle. And I wouldnt worry about more relevant stats becoming available. This is the information age, they're coming...great sites like PFF provide a wonderful look into tracking production with data. 15 years ago no fans cared how how many PBUs or TFLs an elite CB or LB had. Now these can be found in a number of places. Knowing how many pancake blocks a RG had and how many sacks or TFLs he gave up will be here shortly...be patient. And to answer your question, THATS how you'll we'll be able to generate those ratings. When we see a FB who leads the league in pancakes, has the fewest missed blocks and gives up zero TFLs it will be pretty easy to give him a rating based on performance...dont you agree? Mohammed Sanu? Derek Anderson Hmmmm...I doubt you were being serious. I'm pretty sure even the most basic of rating systems subscribes to stat minimums to disqualify guys with a tiny body of work...kind of illustrates my point regarding knowing WHERE to look and looking INTO the numbers instead of AT them. Lastly, your Montana/QB Accuracy comment kind of drives home my point. "A QBs completion percentage, for example, is not only based on the ability to throw an accurate, catchable ball, but it is also dependent upon the WR actually hauling the thing in!" sadly, Montana WAS and IS one of the most accurate passers...The math eventually tells the truth. While I agree that football is a very teammate dependent sport, these are PROFESSIONALS and the accurate QB in your example wont always have bums catching the ball. They will get cut/benched and eventually he will be throwing to guys who CAN catch and the fact that his placement is uncanny will be reflected in his numbers. Talent and Production always win out....every guy who is underperforming isnt always hurt/in a bad scheme/the victim of bad coaching. Sometimes they were just overvalued or of average talent level when compared to the rest of the league.

The primary point I was attempting to make is that statistics are not mutually exclusive. Stats derived during game situations cannot tell the entire story and can be extremely misleading. What you see on game days are the product of individual who have characteristics. When these characteristics are put on display (literally millions of factors), you get game outcomes, and statistics will never be able to quantify all of these factors.

How do you measure a player's toughness through statistics? How do you measure a player's stamina? His agility? His recognition skills? Are you able to get into his mind every play? Are you able to see exactly what he says and explain why he did what he did on any given play? You know who does? Coaches during film sessions certainly do. Scouts ask players these things in interview sessions. There are some things that stats acquired on game day cannot measure.

I think you also missed the point with my Montana example. My Montana example simply outlines how you have to isolate individualized talent. You have to rate players on an individual level because Madden, the system for which we are taking into consideration here, isolates individual players and their talents at the most basic levels (the attributes). Stats acquired on Sundays are the result of team play. Let me ask this: if a QB can throw a ball 80 yards and 70 MPH, would you give him a high THP? Well how would you rate him if his offense doesn't allow him to throw a pass further than 20 yards downfield? What statistics do you use to measure his THP?

Here is something else to consider: how do you rate players who are on practice squads or who do not get to produce on Sundays? Do you rate them all at 0? Do you use their college production? NFL GMs freely admit that college production is a poor indicator of NFL success, so to me that seems less reliable. This day and age Madden gamers want more players...they want full squads they can test out and then cut when in franchise mode. However, if you are unable to give stats for them, how do you evaluate them.

One of the most simple ways I can argue for using scouting and coaching opinion to rate players as opposed to a purely statistics based system is because NFL offices use the same system in evaluating talent. They employ scouting services and their own regional talent scouts to do this stuff. When you watch a show like Hard Knocks you constantly hear about the coaches talking about the talent of these players and their limitations. Most of the decisions they make are based on things they do on the practice field. If a guy has what it takes to make it, they usually expound on the traits that give them that opinion. The inverse also holds true.

Until there is a statistic that takes into account all the interactions of every single player, his ability, his thought process, and everything else that happens on any given play, whereby all outcomes are quantified, I must remain skeptical. Stats are just not mutually exclusive enough in my opinion to contain every answer to every outcome.

[DCEBB2001]

Quote: Originally Posted by justud

and the endless logic bombs just keep on coming. this guy is hammering u DCEB let it go bro.

I can't say I share your opinion. Ratings are a pretty big deal to me, so if I feel it necessary to defend my opinion, rest assured I will.

Thanks for that great contribution to the discussion though.

[Kushmir]

"Until there is a statistic that takes into account all the interactions of every single player, his ability, his thought process, and everything else that happens on any given play, whereby all outcomes are quantified, I must remain skeptical. Stats are just not mutually exclusive enough in my opinion to contain every answer to every outcome."

My only issue is that contrary to what you wrote NO ONE can do this. Not even scouts. They also cant measure toughness, stamina or get into a player's mind on every play. What scouts do during interview sessions doesn't allow them some God-like insight into each player. FACT: Most scouts thought Emmitt Smith would never run away from NFL defenses yet he rarely got caught from behind. Again...production/what happens on the field much higher priority than educated guesses.

And thats largely what scouting is...not to demean it but its informed guessing. Want proof? You have access to scouting data correct? Give me your top 10 rookies of this class in terms of NFL potential. We'll revisit it yearly to see how you did. Remember I dont forget anything...and I hold people to their word. If you get more than half correct i'll concede to your argument. Deal?

We both agree that the process by which real players are rated to simulate their effectiveness in Madden is flawed. We just differ regarding what counter process would be most effective. I defer to production and you to scouting...no harm in it.

[DCEBB2001]

Quote: Originally Posted by Kushmir

"Until there is a statistic that takes into account all the interactions of every single player, his ability, his thought process, and everything else that happens on any given play, whereby all outcomes are quantified, I must remain skeptical. Stats are just not mutually exclusive enough in my opinion to contain every answer to every outcome." My only issue is that contrary to what you wrote NO ONE can do this. Not even scouts. They also cant measure toughness, stamina or get into a player's mind on every play. What scouts do during interview sessions doesn't allow them some God-like insight into each player. FACT: Most scouts thought Emmitt Smith would never run away from NFL defenses yet he rarely got caught from behind. Again...production/what happens on the field much higher priority than educated guesses. And thats largely what scouting is...not to demean it but its informed guessing. Want proof? You have access to scouting data correct? Give me your top 10 rookies of this class in terms of NFL potential. We'll revisit it yearly to see how you did. Remember I dont forget anything...and I hold people to their word. If you get more than half correct i'll concede to your argument. Deal? We both agree that the process by which real players are rated to simulate their effectiveness in Madden is flawed. We just differ regarding what counter process would be most effective. I defer to production and you to scouting...no harm in it.

Smith is an example of function speed. The guy had great vision and great agility while sifting through the trash. His pad level was near-perfect. He never ran sub 4.5 in the 40, but he didn't have to because speed alone is not a predictor of success in the NFL. There are a multitude of ways that a player can succeed when lacking the physical talent of more talented athletes. The fact that Madden doesn't incorporate this into the game is crazy. Instead, we get our unfounded speed boosts to make a player play faster...often times justified by "production", when clearly other factors were more important contributors to a good performance.

When I watch a highlight of Smith, I see lots of YAC, not breakaway speed. Breakaway speed is when a guy like Chris Johnson takes a toss and isn't touched for 80 yards. Smith was so patient with his blocks (behind one of the best OLs ever assembled in the early to mid 90s) and fought through contact so well that he didn't need breakaway speed.

https://www.youtube.com/watch?v=ggWrCJ5-nLs

In this Smith highlight cutup, I see vision, agility, and the ability to break tackles. I do not see great speed though because he is seemingly always being touched by defenders, not evading them by being faster. The run at the 3:40 mark comes close to this, but he is still nearly caught by a LB (#55). The run at 4:11 shows some good agility and a chance at a breakaway run...but he's caught.

https://www.youtube.com/watch?v=1B_Bg6MUvMg

In this run, there are 8 men in the box at the snap (7 on the line and one middle backer about 3 yards off). His burst on this run is AMAZING, but he is still almost run down close to the goal line. In fact, the CB is 4 yards back at the 40 yard line and narrows it down to only a yard at the 5 yard line.

https://www.youtube.com/watch?v=epVFxEPjtjM

Here's another one. 9 (almost 10) in the box. Once he gets to the second level he has a pretty big cushion on the two defensive backs chasing him. That gap is still closed on, but not enough by the goal line.

I see a lot of agility and great burst...just not great top end speed. In the highlights with the big runs he was facing a loaded box and only needed to beat some DBs that were already out of position after crashing in. In the cutup, I see more vision, burst, and agility, but still not great speed.

Speed is one thing that can be quantified, but the production is still the result of the skill. I still think that using the production to base the traits off of seems backwards. I think that is what Donny does now...sees a guy with a great game and gives him speed/agility/acceleration boosts.

Scouting is a ton more than guessing. I think you are thinking about it as refers to drafting players for the draft. It goes far beyond that. It is more about evaluation Scouts are not in the business of prediction...that is for mock drafters, talking heads, and ESPN analysts. Scouts don't bother themselves with this stuff...that's more for the media. The role of the talent scout is to report on the ability of a player, not make predictions. Every scout does this a bit differently, but when you have a wealth of data, you are able to make some conclusions based around the average consensus.

If you want prediction, turn to my site's owner, Rob Rang, or my webmaster, Brian Hitterman. I don't do prediction...I do analysis and interpolation.

[Kushmir]

well thats what i'm asking...can't you (using your wealth of scouting data) tell me based on seeing things like vision, agility, great pad level, burst, function speed and ability to break tackles (your words) you really did such a great breakdown with the smith clip...so do it again. Please?

Cant you do the same with recent tape on the current crop of rookies? Cant you make some conclusions? If not i'd have to assume that hindsight is 20/20 and you're only taking guys who produced in the league and working backwards. And lets be honest, anyone can do that...I've never scouted a day in my life and I can look at tape of Kearse at FLA and tell you what made him a good NFL DE.

I'm sure you can see why any reasonable person would be skeptical, no? Not to be rude but seems like this is the put-up/shut up point...where the rubber hits the road, so to speak...all i'm asking is that you Analyze and Interpolate the top 10 guys in this class.

I'm game if you are.

[DCEBB2001]

Quote: Originally Posted by Kushmir

well thats what i'm asking...can't you (using your wealth of scouting data) tell me based on seeing things like vision, agility, great pad level, burst, function speed and ability to break tackles (your words) you really did such a great breakdown with the smith clip...so do it again. Please? Cant you do the same with recent tape on the current crop of rookies? Cant you make some conclusions? If not i'd have to assume that hindsight is 20/20 and you're only taking guys who produced in the league and working backwards. And lets be honest, anyone can do that...I've never scouted a day in my life and I can look at tape of Kearse at FLA and tell you what made him a good NFL DE. I'm sure you can see why any reasonable person would be skeptical, no? Not to be rude but seems like this is the put-up/shut up point...where the rubber hits the road, so to speak...all i'm asking is that you Analyze and Interpolate the top 10 guys in this class. I'm game if you are.

Like I said in my previous post...I don't do predictions. I can give you some scouting reports and tell you what I think about some guys but the draft is a crap shoot. Scouts don't draft players...GMs and Coaches do. If you want prediction, go ask a Mel Kiper or Todd McShay, after all, they get paid to make predictions...I don't.