FBGRatings Seeks to Recalibrate Madden Ratings, Change Game

[DCEBB2001]

Quote: Originally Posted by charter04

even at default sliders QB's accuracy is much closer to reality. It's still Madden but, the game is more realistic with these rosters IMO

I must say, you don't see as much herky jerky player movement. They seem to run and interact a bit more smoothly.

[charter04]

showing Seahawks at Panthers now. No rookies.

[RogueHominid]

What do you guys do with/about the rookie classes once in a CFM?

[charter04]

Ill show the rest of that game later tonight. Looks like it's going to be a defensive battle

[ggsimmonds]

Quote: Originally Posted by DCEBB2001

Many things, I am sure, go into each category. For route running, how well the receiver gets off the ball without false steps, gains leverage, sells the DB on the stem, uses a pressure step, accelerates out of the cut, doesn't give away the route, etc. all factor in.

But would they break it down to the components, assign each one a individual grade and then add them all up? Could you speak to how rigid the grading is? How much variance is there between different scouts?

This next question you may not be able or willing to answer, but how much variance is there for the entire data set? How is the distribution? The scale is 0.0 to 5.0 correct? Would you mind providing something of a landmark? Like the typical all pro grades at 3.8, average starter would be what? If I am getting too close to asking for specifics or information that you would rather not share I understand. Sorry if I am asking too many questions.

[charter04]

Quote: Originally Posted by ggsimmonds

But would they break it down to the components, assign each one a individual grade and then add them all up? Could you speak to how rigid the grading is? How much variance is there between different scouts? This next question you may not be able or willing to answer, but how much variance is there for the entire data set? How is the distribution? The scale is 0.0 to 5.0 correct? Would you mind providing something of a landmark? Like the typical all pro grades at 3.8, average starter would be what? If I am getting too close to asking for specifics or information that you would rather not share I understand. Sorry if I am asking too many questions.

My guess would be 4.0-5.0 (blue chip) 3.0-3.99(red chip) 2.4-2.99 (purple)

Then the rest

[DCEBB2001]

Quote: Originally Posted by ggsimmonds

But would they break it down to the components, assign each one a individual grade and then add them all up? Could you speak to how rigid the grading is? How much variance is there between different scouts? This next question you may not be able or willing to answer, but how much variance is there for the entire data set? How is the distribution? The scale is 0.0 to 5.0 correct? Would you mind providing something of a landmark? Like the typical all pro grades at 3.8, average starter would be what? If I am getting too close to asking for specifics or information that you would rather not share I understand. Sorry if I am asking too many questions.

The only numbers I get are the aggregate for each category, not the components. There are components, so I have been told, but I do not get that specific data. Let me send an email and see if I can find out what they do exactly for each category and get back to you on that one.

What I do know is that the score they do come up with is based on all of the opinions by each scout. However, you must realize that the player personnel department may assign only a scout or two to a region, so there may not even be much overlap on prospects.

The scale is 0.0-5.0 for all categories, yes. You would need to pick a category as giving you an average number amongst all of them wouldn't say much. The entire database includes players from high school on up. The overall grade, however is based on a 0.00-10.00 scale. The highest I have ever seen a player's OVR was 10.0 and that was Aaron Rodgers in 2011. The lowest I have seen is 0.0 but due to an injury (all I will say is that it was a college kid who got hurt really bad in 2011/12). The average score in the database, including all players from all levels is 2.42 as the OVR score.

[ggsimmonds]

Quote: Originally Posted by DCEBB2001

The only numbers I get are the aggregate for each category, not the components. There are components, so I have been told, but I do not get that specific data. Let me send an email and see if I can find out what they do exactly for each category and get back to you on that one. What I do know is that the score they do come up with is based on all of the opinions by each scout. However, you must realize that the player personnel department may assign only a scout or two to a region, so there may not even be much overlap on prospects. The scale is 0.0-5.0 for all categories, yes. You would need to pick a category as giving you an average number amongst all of them wouldn't say much. The entire database includes players from high school on up. The overall grade, however is based on a 0.00-10.00 scale. The highest I have ever seen a player's OVR was 10.0 and that was Aaron Rodgers in 2011. The lowest I have seen is 0.0 but due to an injury (all I will say is that it was a college kid who got hurt really bad in 2011/12). The average score in the database, including all players from all levels is 2.42 as the OVR score.

Interesting stuff, thank you. I imagine due to encompassing all levels that standard deviation is quite large. Could you perhaps provide a ballpark figure for average OVR for NFL players and the SD? I am wondering how wide the gap is between players. I would guess that it is lower than one would think. This way you don't have to provide anything regarding specific players.

Also in your opinion what position allows for the greatest variance in terms of overall grade and could that be an argument for it being most important? I hope it is not QB but if it is what is second?

Do you ever do some just for the hell of it models? You know, just play with the data?