FBGRatings Seeks to Recalibrate Madden Ratings, Change Game

[michiganfan8620]

Quote: Originally Posted by charter04

There were 182 players rated 90+ on the latest madden 25 official roster. That's crazy.

I agree, that probably is too many, however, the formulas for each position are different in order to make it so that there are a few 90s at each position. Guys at different positions can't really be compared by OVR. What classifies as a 90 at different position are different standards. A 92 WR is not necessarily a better player than a 90 CB, but on the scouting scales, a 4.1 WR is most likely better than a 4.0 CB. It doesn't make sense, but it's the way the game was designed.

[charter04]

Quote: Originally Posted by michiganfan8620

I agree, that probably is too many, however, the formulas for each position are different in order to make it so that there are a few 90s at each position. Guys at different positions can't really be compared by OVR. What classifies as a 90 at different position are different standards. A 92 WR is not necessarily a better player than a 90 CB, but on the scouting scales, a 4.1 WR is most likely better than a 4.0 CB. It doesn't make sense, but it's the way the game was designed.

Your problem you seem to think EA's overall system makes since. Maybe that is the issue. They give way too much weight to stuff that doesn't make a player better. They are the ones who start with a desired overall(Sherman being 99) and just raise stuff like awareness or strength to get it.

You can get a 60 overall to be just as good as a 99 as far as play on the field in madden if you know how the ratings are weighted. I have done this. That is a lot of how Dan does it. The overall doesn't matter. It's how they play. The players play more life like with these ratings than that garbage EA gives.

[charter04]

Do you want good gameplay where ratings mean something or do you want them to be "right" as far as the way EA does it? I personally care more about how the game plays on the virtual field than if player A is rated right as I see it. Playing games with these feel more like a real nfl game even with default sliders. Especially as he has it now using equal interval.

[DCEBB2001]

Man, this thing really blew up while taking the GF out on date night...


Let me address a few things here.

1. The scale that Lindy's, PFW, Scouts, Inc, and the scouting department that utilize my data all use a similar structure: a linear method to rating players that utilizes equal intervals between each point on the scale. They set the worst of the worst possible to 0.0 and the best of the best possible (perfection, if you will - like Cris Carter's hands) to 5.0, knowing full well that there will be far more 0.0's out there than 5.0's. They draw a line from worst to best and "let the rest fall as it may" along that linear regression between those two points. The guys at 4.50+ are VERY, VERY, RARE.

If you don't trust how FBG Ratings rates players, then look at Lindy's, PFW, or Scouts, Inc. Scouts, Inc. used to rate NFL players on ESPN.com. In 2012, the last year they published online grades for NFL players, their top player overall (regardless of position) was Aaron Rodgers at 97, which translates to about 4.85 on the 5.0 scale. Tenth was Gronk at 92 (4.60). The last player over 90 (4.50), or "elite" was Suggs at #19. That means that Scouts, Inc had only 19 players in the entire league over 90. EA Sports, in Madden 25, had 136 players that were 90+. Meanwhile, Big Bad FBG Ratings (who apparently gets everything wrong) has currently only 6 players rated 90+.

Either way, both Scouts Inc and FBG Ratings have a significantly lower percentage of players that are rated highly in comparison to their EA counterparts. Now I pose this; either Scouts, Inc and FBG Ratings (with lower numbers of players who are considered "elite") are wrong, OR EA Sports with well over 6 times as many players rated 90+ are wrong. With such a wide gap, both cannot be correct.

2. The Madden OVR and non-Madden OVR's are exactly the same when you draw a point from worst to best. In Madden that is (for the OVR rating) from 12 to 99. In FBG it is 0.0 to 5.0. Sure the numbers aren't the same, but that is what the data interpolation is for! This way, we can MAKE them the same. You simply set 0.0 = 12 and 5.0 = 99. Now, draw your line and see where players fall on the scale.

This is not the problem, however. The real problem is not the scale itself (as I just solved it rather simply), but where players FALL on the scale. EA says that 136 players fall above 90+ on that scale. FBG says it is 6. Scouts, Inc. says 19. Who is correct? 136 vs. 6 is pretty wide of a variance and 19 falls a hell of a lot closer to 6 than it does to 136. You are asking the wrong question when you ask about the scale. The scale doesn't matter - they can be made equal. What matters is the DISTRIBUTION of the data. THIS is where you determine how evenly or unevenly skewed your population is.

The formulas for ATTAINING the overall scores, ARE different, however. But, when you think of how it pertains to the OVR scores, this also doesn't matter. Why? Because each OVR score no matter what system you use is dependent upon the variables and their sums. They all still use an OVR rating! So, because of this fact, we are still able to compare them equally. It would be one thing if one system didn't use an OVR score, but they all do. That is the point! Scouts use it, so why wouldn't we? Do you remember that leaked scouting report by the Pats on Johnny Manziel this offseason? Even THAT had a final OVR grade on it! People who fail to realize that scouts use OVR grades to compare players from one position to the other are completely ignorant of the true processes behind all of this.

So, on a scale of 0.0 to 5.0 where 0.0 is 12 and 5.0 is 99, then a grade of 4.1 is 83. This scale is unbiased toward any average and the chips are allowed to fall as they may. The only bias is at the max and min, but the slope is equal the whole way. That means that the difference between a player with a grade of 4.0 and a player with grade of 4.1 is the same as the distance between a player with a grade of 3.6 and a player with a grade of 3.7. This is the most realistic way we can determine where each player truly lies on the scale. Think about this: in your stats class in high school or college, did your x and y axis have intervals between 0.0 and 1.0 that varied in distance or were they uniform? They were uniform, of course! Otherwise the distance between 0.0 and 1.0 wouldn't be accurate! It would be something else! The notion that someone can fight this logic is just so far beyond me that I am expecting blood to just start shooting out of my nose at any point of this response.

3. The OVR rating is used to rate players from one position to the other. THAT IS THE WHOLE POINT! To say that the OVR rating is different from one position to another makes no sense! Of course it does! How else do you determine who the best player in the league is? Since EA Sports has completely brainwashed people into thinking that the best player at each position every year should be 99, it is no wonder why someone would erroneously think this.

If a QB is a 99, no matter what formula used to derive it, is he better than a WR rated 40? According to the phoney logic, the QB wouldn't be any better because we "can't use the OVR to evaluate player across positions". Do you realize how illogical that sounds? If the best WR in the game is an 81 and the best QB in the game is a 99, who do you think will perform better on the field given their position? The QB of course, because he will have higher attributes that gave him the higher OVR in the first place! The OVR is dependent upon the attributes, not separate from it!

4. I cannot believe the lack of thought process about this sometimes. I utilized professional opinion, data, and logic in all of this. Where have you heard Donny Moore mention any of that? Where have you heard him say anything besides Youtube, Stats, and PFF? Come on. The way he rates players is a JOKE. When Lindy's NFL magazine comes out, count the number of elite players listed in that book. I bet you can count them all on your fingers and toes. Do the same for PFF and Scouts, Inc. You will get results that are a hell of a lot closer to what FBG Ratings offers vs. what EA offers as far as rating the OVR for NFL players.

But don't take my word for any of this...after all, I'm just a guy making this crap up as I go along, right? Woah! A QB just went 20 of 24 for 294 and 2 TDs! He is sooo accurate! I better bump up his SAC, MAC, DAC, and AWR all 10 points to mimic that one awesome game! He is a super star! What is his name again? Oh yeah....

Rob Johnson.

[charter04]

I got the Cardinals re- rated with the equal interval ratings and I played a half Cardinals v Cardinals. Default AP. The only changes was QB acc for user and CPU at 35.

I now can't wait to get Atlanta rated. The game plays even better with these.

When Atlanta is done I'll play on twitch

[azdawgpound]

hey dc are u guys updating your site or something I was updating my ps4 roster got to the RT's for bengal's then page froze and now all I get is waiting for site.


I can go to any other sites no problem sites load right up sorta through wrench in my plans I was making good process today.

[DCEBB2001]

Quote: Originally Posted by azdawgpound

hey dc are u guys updating your site or something I was updating my ps4 roster got to the RT's for bengal's then page froze and now all I get is waiting for site. I can go to any other sites no problem sites load right up sorta through wrench in my plans I was making good process today.

If you ever see the site say that, it means we are doing updates. Just started about 10 minutes ago. Should be done in a couple hours.

[sword1986]

Do you think they will let us use your rosters for CCM? I am definitely looking to purchase Madden here, last Madden was the one with Peyton Hillis on the cover and before that was Madden 09. Your rosters are LEGIT, by the way.