Re: FBGRatings Seeks to Recalibrate Madden Ratings, Change Game
So I finally had some time to play Madden a bit this weekend, and I spent some time messing around in a Madden 25 CFM, basically doing roster moves and simulating seasons.
I'm going to have to look into it further, but I'm not entirely convinced that that FBGratings true mean ratings are completely incompatible with CFM based on the rookies I'm seeing the game spit out. The players on the FBGratings website right now are rated on the equal interval method, thus are obviously not going to work with CFM because the OVR ratings are all over the place and that's likely going to throw the CPU contract / free agency / drafting logic for a loop (regardless how the game may play with them, I don't think that's been tested by anyone yet unless Charter04 has completed some teams). That's not faulting the way Dan is rating the players based on the scouting data, not at all; just a fundamental incompatibility.
Granted, how Madden 15 generates the random draft classes obviously will have some effect on this as to how much of this applies to this year's game, if at all. But my initial impression is that the custom-made rookies in M25 aren't rated on the same scale as the one Donny Moore uses for the base roster file.
At the very least, I wanna take some time and look at the rookies from the M25 draft classes in detail and see if the attributes come close to matching (i.e. the mean and standard deviation is close; I don't know what I define as "close" yet, though, any ideas?).
Definitions for refresher:
- FBGratings equal interval ratings: every player is graded in every field on a 0-100 scale, ignoring entirely how the game functions. This is how the website is currently rating players.
- FBGratings true mean ratings: the equal-interval ratings are scaled to better match expected values within the Madden NFL game. This is how the website was rating players earlier this summer, back in May and June.
So I finally had some time to play Madden a bit this weekend, and I spent some time messing around in a Madden 25 CFM, basically doing roster moves and simulating seasons.
I'm going to have to look into it further, but I'm not entirely convinced that that FBGratings true mean ratings are completely incompatible with CFM based on the rookies I'm seeing the game spit out. The players on the FBGratings website right now are rated on the equal interval method, thus are obviously not going to work with CFM because the OVR ratings are all over the place and that's likely going to throw the CPU contract / free agency / drafting logic for a loop (regardless how the game may play with them, I don't think that's been tested by anyone yet unless Charter04 has completed some teams). That's not faulting the way Dan is rating the players based on the scouting data, not at all; just a fundamental incompatibility.
Granted, how Madden 15 generates the random draft classes obviously will have some effect on this as to how much of this applies to this year's game, if at all. But my initial impression is that the custom-made rookies in M25 aren't rated on the same scale as the one Donny Moore uses for the base roster file.
At the very least, I wanna take some time and look at the rookies from the M25 draft classes in detail and see if the attributes come close to matching (i.e. the mean and standard deviation is close; I don't know what I define as "close" yet, though, any ideas?).
Definitions for refresher:
- FBGratings equal interval ratings: every player is graded in every field on a 0-100 scale, ignoring entirely how the game functions. This is how the website is currently rating players.
- FBGratings true mean ratings: the equal-interval ratings are scaled to better match expected values within the Madden NFL game. This is how the website was rating players earlier this summer, back in May and June.
Comment