Old Madden Ratings Site Reborn Under New Management

[at23steelers]

Quote: Originally Posted by Maelstrom-XIII

Here's where I think we have a philosophical difference about ratings. To me, ratings should be STRETCHED. And I mean elastic waistband, "don't let go that's really going to hurt" stretched. We have a ratings system that goes from 1-99. The median value in that system is a 50 (actually 49.5 but you get the idea). Therefore, it stands to reason that an AVERAGE rating should be a 50. Therefore, an average NFL player, the JAGs (Just A Guy) would be in the 50s overall. That may seem low, but it's the average here...I'd even say, "Sure let's go to 60 instead." So now we have our average NFL player at 60. 70s? Starter. 80s? Pro-bowler. 90s? Future Hall of Famer. We're talking a handful of 90s. But people would have a fit if they saw players on their favorite teams with overalls in the 60s, or speeds that are much lower than they perceive because they're used to everyone having high 80s or 90s speed...but alas, I'm writing up a blog about this very issue. And from my perspective, being effective, but not dominant, for one year does not equal an 80 rating.

This sounds good and all, but QB's with 60-70 accuracy in Madden, is pretty terrible. So, if the average QB has that for a rating, then he will be playing worse than what he does in the NFL. It still doesn't fix the AI, but just makes them much worse. The ratings now, correspond with the gameplay functions in the game, if that makes sense. It's hard enough to play Madden with a backup as a starter, but would be pretty difficult, if all your starters played like backups. I hope that makes sense.

[Maelstrom-XIII]

I understand exactly where you're coming from. I'm not saying we should re-rate every player for Madden 11, because the development team has tuned performance around their own concept of what should be "average" and how often bad throws occur...I know that lowering Trent Edwards' accuracy down to a 60 or so would make for an incredibly frustrating experience (which Bills fans should be used to)...I'm just advocating for the development team to re-evaluate what they're rating system looks like, and why it isn't stretched to account for the actual median value of their range. I mean, why bother having ratings go from 1-99 if you only use 30 values (69-99) for it? I'm being a bit facetious of course, but you can see where I'm coming from, yes?

It certainly isn't helping their stance on "We have a new ratings philosophy" when their stretched ratings have just as many 90 overall players as past Madden iterations. You have the whole range for a reason, let's use it.

Some tuning, like DCEBB is working on, is a step in the right direction, but my holy grail of rating systems would be a ludicrously stretched rating system, which is what I posted up above...

[at23steelers]

Quote: Originally Posted by Maelstrom-XIII

I understand exactly where you're coming from. I'm not saying we should re-rate every player for Madden 11, because the development team has tuned performance around their own concept of what should be "average" and how often bad throws occur...I know that lowering Trent Edwards' accuracy down to a 60 or so would make for an incredibly frustrating experience (which Bills fans should be used to)...I'm just advocating for the development team to re-evaluate what they're rating system looks like, and why it isn't stretched to account for the actual median value of their range. I mean, why bother having ratings go from 1-99 if you only use 30 values (69-99) for it? I'm being a bit facetious of course, but you can see where I'm coming from, yes? It certainly isn't helping their stance on "We have a new ratings philosophy" when their stretched ratings have just as many 90 overall players as past Madden iterations. You have the whole range for a reason, let's use it. Some tuning, like DCEBB is working on, is a step in the right direction, but my holy grail of rating systems would be a ludicrously stretched rating system, which is what I posted up above...

I agree with you 100% as long as they made the gameplay correspond to your rating scale, then it would be an excellent addition to the game. It's then easier to see who are the best players in the NFL.

[tlc12576]

Quote: Originally Posted by DCEBB2001

Relying on stats though will get you in trouble. You can't rely on them. I would think that you should rely more on scouting data. If you want stats though, try to find insider stats. Profootballfocus has some great insider stats. They review every game, player, and snap throughout the year and find out how players do in situations. For your DTs you are so concerned about, they break down how well a DT did against the run. They even rate DE/OLBs different depending on the system they play in! Not a bad source, but I still prefer the scouting data we have, then simply convert those findings into ratings.

DCEBB, this brings me back to the discussion we were having back on the other thread about how AWR and OVR relate in Madden. I dont have a problem with admitting when Im wrong and seems like I was in this case. I think since OVR is so widely accepted as the players complete skill set including AWR, they should be factored together on some scale. I still believe rookies and young guys should rarely, if ever, have high AWR and that NFL vets should. However, since alot of gamers will use OVR mainly to compare players, OVR should be skill set calculated with AWR, IMO. This would allow players OVR to stay balanced even if young players skill sets are high or NFL vets skill sets begin to decline.

This would prevent Sidney Rice's OVR from being to high just because he has good skill sets and great height. His AWR would keep him balanced till he begins to play really well, consistently. As I am typing this, Im really coming to the conclusion that AWR should be based on player consistency, since we dont have player position IQ test, IMO.

I think a good example of this in the NFL was Reggie White when playing for GB. Whites actually skill set peeked with the Eagles but he was still able to play consistently great in GB using his position IQ(AWR), despite his deminishing skill set. The best indicator of position IQ(AWR) is a players on-field consistency, IMO. This should be calculated in some manner with players position skill sets to determine OVR, IMO. For example with WRs, catching, route running and release are major position skill sets that should factor into OVR. However, AWR should factor heavily into OVR as well because all the talent in the world is of little benefit without consistent play.

Dez Bryant has high WR skill sets but until he gets in the game and becomes a consistent playmaker threat he will not be considered a good overall(OVR) receiver. I dont know the formula or equation to use but I think AWR in Madden currently isnt calculated right. IMO.

I still believe it would be ideal if OVR was calcualted with just the position skill sets. However,currently it seems that other gamers consider OVR to be the overall scale at that position and consider AWR to be a part of this. If that's the case, calculating position skill sets and AWR to determine OVR, using a different equation than EA, is probably best, IMO.

I have no doubt that you are already ahead of me on this but I just wanted to correct my position anyway.

[DCEBB2001]

Quote: Originally Posted by Maelstrom-XIII

Here's where I think we have a philosophical difference about ratings. To me, ratings should be STRETCHED. And I mean elastic waistband, "don't let go that's really going to hurt" stretched. We have a ratings system that goes from 1-99. The median value in that system is a 50 (actually 49.5 but you get the idea). Therefore, it stands to reason that an AVERAGE rating should be a 50. Therefore, an average NFL player, the JAGs (Just A Guy) would be in the 50s overall. That may seem low, but it's the average here...I'd even say, "Sure let's go to 60 instead." So now we have our average NFL player at 60. 70s? Starter. 80s? Pro-bowler. 90s? Future Hall of Famer. We're talking a handful of 90s. But people would have a fit if they saw players on their favorite teams with overalls in the 60s, or speeds that are much lower than they perceive because they're used to everyone having high 80s or 90s speed...but alas, I'm writing up a blog about this very issue. And from my perspective, being effective, but not dominant, for one year does not equal an 80 rating.

I think you need to draw the comparison between the stretching of attributes and OVR ratings. The attributes go from 12-99, but the overalls are very different. For instance, there is a point where the OVR rating will not go any lower. Take a QB, make all of his attributes the lowest possible (12 if you edit them in the game as it will not let you edit any attribute lower than that). His OVR rating with all of his attributes at 12 will still make him like a 28 or something overall. At least that is how it used to be.

Would anyone be willing to go into Madden 10 real quick and create a player at all positions to determine their rating with all attributes turned down as low as possible? I want to check this out.

If that is the case then the average is not 50, but perhaps (99-28)/2 = 35.5 + 28 = 63.5

All I did was take the highest rating possible (99) and subtract the lowest rating possible (28?). Then take that number and divide by 2. Then add that to the lowest number. Note that this is only the case if 28 is the lowest rating possible.

So, once again, the real "average" could be a lot higher than 50 depending on what the lowest OVR rating possible is.

As for the attributes, an average attribute would be 100-12 = 88/2 = 44 +12 = 56.


However, I would prefer to use 70 as an "average" for each attribute. It kind of follows a grading system where an average mark is 70 to 75 depending on the scale you like.

[cubsball899]

absolutely loving what i'm seeing with this thread.... if these ratings work well, madden 2011 might turn into a buy for me ... imagine how many things in the game will look better with significantly lowered ratings

i'm with maelstrom i think though and love this as a good start but i'd try to lower them even more... addai IMO shoudn't sniff the 80's as he's been one example

but a love what i'm reading here i'll stay tuned!!

[Maelstrom-XIII]

Quote: Originally Posted by DCEBB2001

I think you need to draw the comparison between the stretching of attributes and OVR ratings. The attributes go from 12-99, but the overalls are very different. For instance, there is a point where the OVR rating will not go any lower. Take a QB, make all of his attributes the lowest possible (12 if you edit them in the game as it will not let you edit any attribute lower than that). His OVR rating with all of his attributes at 12 will still make him like a 28 or something overall. At least that is how it used to be. Would anyone be willing to go into Madden 10 real quick and create a player at all positions to determine their rating with all attributes turned down as low as possible? I want to check this out. If that is the case then the average is not 50, but perhaps (99-28)/2 = 35.5 + 28 = 63.5 All I did was take the highest rating possible (99) and subtract the lowest rating possible (28?). Then take that number and divide by 2. Then add that to the lowest number. Note that this is only the case if 28 is the lowest rating possible. So, once again, the real "average" could be a lot higher than 50 depending on what the lowest OVR rating possible is. As for the attributes, an average attribute would be 100-12 = 88/2 = 44 +12 = 56. However, I would prefer to use 70 as an "average" for each attribute. It kind of follows a grading system where an average mark is 70 to 75 depending on the scale you like.

Hmm...that's a very good point. I concede to your expertise.

However, I pick up my new charge. No more OVR at all...just attributes.

[DCEBB2001]

Quote: Originally Posted by tlc12576

DCEBB, this brings me back to the discussion we were having back on the other thread about how AWR and OVR relate in Madden. I dont have a problem with admitting when Im wrong and seems like I was in this case. I think since OVR is so widely accepted as the players complete skill set including AWR, they should be factored together on some scale. I still believe rookies and young guys should rarely, if ever, have high AWR and that NFL vets should. However, since alot of gamers will use OVR mainly to compare players, OVR should be skill set calculated with AWR, IMO. This would allow players OVR to stay balanced even if young players skill sets are high or NFL vets skill sets begin to decline. This would prevent Sidney Rice's OVR from being to high just because he has good skill sets and great height. His AWR would keep him balanced till he begins to play really well, consistently. As I am typing this, Im really coming to the conclusion that AWR should be based on player consistency, since we dont have player position IQ test, IMO. I think a good example of this in the NFL was Reggie White when playing for GB. Whites actually skill set peeked with the Eagles but he was still able to play consistently in GB using his position IQ(AWR), despite his deminishing skill set. The best indicator of position IQ(AWR) is a players on-field consistency, IMO. This should be calculated in some manner with players position skill sets to determine OVR, IMO. For example with WRs, catching, route running and release are major position skill sets that should factor into OVR. However, AWR should factor heavily into OVR as well because all the talent in the world is of little benefit without consistent play. Dez Bryant has high WR skill sets but until he gets in the game and becomes a consistent playmaker threat he will not be considered a good overall(OVR) receiver. I dont know the formula or equation to use but I think AWR in Madden currently isnt calculated right. IMO. I still believe it would be ideal if OVR was calcualted with just the position skill sets. However,currently it seems that other gamers consider OVR to be the overall scale at that position and consider AWR to be a part of this. If that's the case, calculating position skill sets and AWR to determine OVR, using a different equation than EA, is probably best, IMO. I have no doubt that you are already ahead of me on this but I just wanted to correct my position anyway.

First off, props to you man. It takes a real man to admit when he is wrong...especially here.

Now for the issue you bring up:

At FBG in the past we had a simple equation to determine a players AWR. Now, if you analyze last year's ratings there was a strong statistical correlation between a player's OVR rating and his AWR rating. This was a positive correlation of nearly .90. To those who are not stat-obsessed that means it was a positive correlation where 90% of the data was correlated. Most people who do stats accademically shoot for about .70 to show a statistically significant correlation between 2 pieces of data. This means that it wasn't just a strong correlation, but a VERY VERY VERY strong correlation.

So what does this mean? It means that as a players AWR goes up, his OVR goes up in nearly 90% of the cases. It also means that as a players OVR goes up, his AWR goes up in 90% of the cases. So what can we draw from this? We can conclude that the two are tied into each other very closely.

At FBG we used to actually have a formula for determining a player's AWR rating for rookies. We would take his OVR rating, which is pre-determined based upon his other skills including SPD, ACC, JMP, CTH, etc...and subtract 10.

Dez Bryant. As a rookie his skill set gives him a 60 OVR. So you subtract 10. His AWR is now 50 as a rookie.

So what about players who are not rookies? For them you do the same thing, but instead take the present year and subtract their draft year from it. Then add that to their OVR-10.

Randy Moss. Is currently rated a 90. Subtract 10. Now take this year (2010) and subtract his draft year (1998). Now add them.

90-10 = 80
2010-1998 = 12
80+12 = 92

It gives progression for each year a player is in the league, and rightfully so.

Peyton Manning

98-10 = 88
2010-1998 = 12
88+12 = 100 (or 99 because it is the maximum.)

This doesn't over-inflate the AWR rating and still takes into account experience and the player's OVR rating. Thoughts?