Madden 2010 Ratings

[soonertf]

Well the ratings are based on what the playters did last year, and I doubt they take into consideration OTAs yet, though weekly updates should take care some of that. I noticed a few players did seem to take a fairly decent hit (Ben Leber) on my team for instance. But without scrapping the grading system and coming up with something new, I don't see how you properly adjust some players and not others. Plus I notice there are 99 rated players still, What would we do to adjust them accordingly? The 4 players you mentioned didn't have big seasons last year so they would naturally not be ranked that high (not sure a D+ is fair for Watson). I'm not sure what the answers is, but I do agree we need to wait to see what the rest of the grades are. I haven't had a lot of players yet, but outside of Leber and Avril to an extent, I didn't notice anything that really stood out as unfair. If we do make adjustments, I think it's going to have to be a system across the board, and can't be a (well this guy is going to play more type of thing).

[Goodell]

There was previous discussion about some stated grading changes, and reaction that we'd have to see how that looked when we had more info and that it was a good time to make some adjustments to the grade scale. I don't think anything too major but shifting the ranges for groups a bit as the other grade system supposedly has shifted.

How big or little, I'm not sure and certainly something to be discussed-- but the intention as I would see it would be to keep it as initially setup with our grades corresponding to set accomplishment levels and trying to roughly match up category groupings of players from that grade scale into ours. We'll probably also have less separation in our grade scale translation than before where 1 point difference in Madden isn't as large of a difference here any more.

Hopefully when their entire list is out and sorted by value, there will be some natural groupings within their ratings where some of those lines are apparent. When you go down the list and start to get into clear backups or marginal/unproven starters of circumstance, we can start to see where those lines are drawn numerically to match up to our levels, etc. That examination wouldn't go down to an individual level but more generalized groups of players and natural numeric breaks that make sense, but those ratings are always moving as players improve or regress for players on the borderline who will eventually find their place as their career develops that's more for them to decide than us. We just want a grade scale that translates those numbers into something reasonable for our grade level expectations.

There's not a lot to do in the off-season, but that would be one important topic that'll probably be highly debatable and impactful on the game. For me I want to see more info and really put a big chunk of time toward an enhanced grade scale that provides the fairest representation of value based upon our letter assignment definitions. A "B" might have a different numeric translation from Madden score to our play this year, but ideally will be the same caliber of player fitting that description of an established, solid, dependible starter who would have that same prominent role with most any team -- and so fourth for each level and their set group description. Those types of players in that group from before might have lower Madden scores this year, but if so we'll adjust slightly to make sure our groupings make sense.

[Andrew P]

I have been looking at the grades and while I will admit they are grading tougher this year in Madden, I like it. I remember past years players like Jeff Backus were getting 89 Ratings and that is just crazy. I think we should go off the grades from Madden and not change a thing. The scale has changed but it has not just changed for one team, it has changed for the whole league. Hell I just paid a bunch of money for Charles Tillman who is now rated an 82….ouch but I will live.

[Goodell]

Just depends upon how it all comes out. We don't want legitimate established starters to be graded C's or D's. They should be B's by our scale. If players were over-inflated before and now it fits our structure better, that would be fine too. We just need to make sure our assigned letter grades make sense where the super stars are A's, established starters are B's, backups or marginal/circumstancial starters are C's, guys struggling to stay in the league are D's, etc.

We'll just make sure that the majority of guys who reasonably fit into a category of accomplishment have their appropriate grade letter here match that. That may need only slight tweaking (hopefully) or perhaps more. We'll have to see how they all lineup.

[Royce R]

Our scale was made to make star players stand out because madden didn't do a good job of it.

Now madden has made a bigger difference between average players and star players.

Then you put that into our scale and it is a huge difference. Some tweaking is needed for sure.

Look at it this way.. A 99 rated player in madden vs a 80 rated player. So the 2nd player is 80% as good as the first.

Bye our scale 99 = 9
80 = 5.3

58% as good of player as the allstar?? That doesn't seem right to me. All that is going to do is make people pay 20-30 million/year contract on the good players, if our scale doesn't adjust they will have to big of an impact.

[Ben C.]

After noticing that Troy put the Madden grades into the rosters, I compiled a spreadsheet of all the Madden grades that have been released to do some statistical analysis. I found that the average rating is 71.84 with a standard deviation of 11.46. For those who aren't very good at statistics, this means that about 68% of the league falls between 60 and 83. 95% of the league has a rating between 49 and 95. Therefore, I tried to design a scale that would match this a little bit. I figured that a good way to do the scale would be to have the division between a C and C+ be the average of the league. This follows Troy's original hope of following a scale like this:

A+ : Hall of Famer, peak years (one of the truly elite top 25 type players in the game playing his best now. Brady/Manning types with multiple all-pro recent recognitions or MVP or other awards).
A : All-Pro, Pro-Bowl, etc. (and not just making the pro bowl because 5 other QBs were hurt, nor ancient history -- playing at an all pro level today)
B+ : Above Average Starter (starting player with some kind of documented recognition of his abilities, perhaps an older player who was recognized as an all-pro in the past but not quite amongst the best anymore but not too far down the list)
B : Legit Starter (not just a starter due to injury but could start on most teams in the NFL)
C+ : Below Average Starter (starter on the depth chart but perhaps due to injury or team weaknesses mostly)
C : Average Backup
D+ : Below Average Backup
D : Fringe Backup
F : Worst, liability on the field

As a result, I have come up with the following scale. Out of all the players in the league, 1.83% would be graded A+, 5.19% A, 10.14% B+, 13.56% B, 17.16% C+, 21.70% C, 17.92% D+, 7.9% D, and 4.6% F.

I also suggest that in light of Royce's point that the superstars seem to be given a final rating that is significantly higher than a comparable player, that we give a final simulation rating of the numbers listed in parentheses; however, it seems to raise the overall average simulation rating for starters from approximately 6 to about 6.5.

A+ = 96-99 (8-9)
A = 91-95 (7.5-7.9)
B+ = 84-90 (7-7.4)
B = 78-83 (6.5-6.9)
C+ = 72-77 (6-6.4)
C = 66-71 (5.5-5.9)
D+ = 60-65 (5-5.4)
D = 54-59 (4.5-4.9)
F = 0-53 (0-4.4)

This of course is up to debate, and I'm open to whatever we end up going with - just wanted to get some ground numbers out there that have basis on the final Madden figures.

[RyanM]

A+ = 96-99 (8-9)
A = 91-95 (7.5-7.9)
B+ = 84-90 (7-7.4)
B = 78-83 (6.5-6.9)
C+ = 72-77 (6-6.4)
C = 66-71 (5.5-5.9)
D+ = 60-65 (5-5.4)
D = 54-59 (4.5-4.9)
F = 0-53 (0-4.4)

Nice work, Ben. I must say that after putting the letter grades with the Madden number grades for my team, this scale works about to what I would expect given my team's talent (or lack thereof). The closest to a Pro-Bowler for me is Tatupu, who would be a high-end B+ (Madden=90), while most of my starters are low-end B's or high-end C+'s. That's pretty accurate when you look at Troy's description of what the letter grades should represent.

[Goodell]

Great! Thanks for the thoughts above, as all discussion and suggestions will contribute to the ultimate solution. I have been working on a revised scale looking over the new Madden scores the past few days and will put that out also below.

On the online roster pages now you'll see a numeric Madden score from recently released grades for the upcoming season. Let me know if you have a player without a Madden score who should have one. You'll also see the old letter grades that haven't changed since the season ended. Now that we have the info and various takes on it, we can begin to focus on where the grade scale should be now. Once that's decided this week, I'll adjust all the letter grades accordingly.

Here is a collection of all the Madden scores in one spreadsheet (versus the various team spreadsheets they have online). You'll also see another tab for suggested grade scale modification.

Madden Scores for 2009 and possible scale adjustment:
http://www.fangm.com/football/M10.xls

This was my proposed grade scale adjustment based upon a couple of factors:

• I wanted to keep it somewhat symetrical with some milestones to make it easy to get a handle on when a player might be moving up when he goes from a 70s score to an 80s score that some natural borders be understandable.
• More importantly I wanted to make sure that when players grouped that it fit the previously established letter grade scale when you looked at it on an individual level. That's difficult because of so much difference of opinion on players, but generally the grouping of players needs most of them to fit within their letter grade description for the most part.
• I wanted to reduce the gap between fairly similar Madden scores. In this recommendation, it generally goes .1 for every 1 Madden point throughout other than at the very top for B+ and above players who get .2 for each Madden point difference. Much smaller differences and more refined scores than previous groupings.

Although the madden scores seemed to be scaling down, this adjustment really would scale our overall averages up. I don't want to get too extreme with that to make it very different from past seasons and will have to look into that further, but also wanted to bring some of the starter-level grades closer to the above-average starters where not such a huge gap but still enough to recognize their greatness. Also keep in mind that there are many more players in our system than listed with madden scores so far so not a lot of lower rated guys on madden yet because they are not high enough on the totem pole to be listed with them while they are with us and would have lower scores at the bottom.

2009 PROPOSED SCALE:
98-99 : A+ (9)
-- 99 : 9
-- 98 : 8.8
93-97 : A (8)
-- 97 : 8.6
-- 96 : 8.4
-- 95 : 8.2
-- 94 : 8.0
-- 93 : 7.8
88-92 : B+ (7)
-- 92 : 7.6
-- 91 : 7.4
-- 90 : 7.2
-- 89 : 7.0
-- 88 : 6.8
80-87 : B (6)
-- 87 : 6.7
-- 86 : 6.6
-- 85 : 6.5
-- 84 : 6.4
-- 83 : 6.3
-- 82 : 6.2
-- 81 : 6.1
-- 80 : 6.0
70-79 : C+ (5)
-- 79 : 5.9
-- 78 : 5.8
-- 77 : 5.7
-- 76 : 5.6
-- 75 : 5.5
-- 74 : 5.4
-- 73 : 5.3
-- 72 : 5.2
-- 71 : 5.1
-- 70 : 5.0
60-69 : C (4)
-- 69 : 4.9
-- 68 : 4.8
-- 67 : 4.7
-- 66 : 4.6
-- 65 : 4.5
-- 64 : 4.4
-- 63 : 4.3
-- 62 : 4.2
-- 61 : 4.1
-- 60 : 4.0
50-59 : D+ (3)
-- 59 : 3.9
-- 58 : 3.8
-- 57 : 3.7
-- 56 : 3.6
-- 55 : 3.5
-- 54 : 3.4
-- 53 : 3.3
-- 52 : 3.2
-- 51 : 3.1
-- 50 : 3.0
40-49 : D (2)
-- 49 : 2.9
-- 48 : 2.8
-- 47 : 2.7
-- 46 : 2.6
-- 45 : 2.5
-- 44 : 2.4
-- 43 : 2.3
-- 42 : 2.2
-- 41 : 2.1
-- 40 : 2.0
0-39: F (1)

This isn't something that will necessarily be implemented, just one of the suggestions to go along with any others already posted or that others want to contribute. Things will be shifted with further discussion. I don't mean this to stiffle other thoughts, and really need intelligent GMs to look over all contributions and see what makes sense or what needs tweaked a little more. I want to get that resolved this week, though, to get things back up and going again with other things still to be done.

With hundreds of players it makes the rankings job difficult but the groupings need to make sense for how the letter grades are intended generally IMO, so you might look at some of the borders in this system and others to see how it would impact your team and what does or doesn't seem to make sense.

[Alex S]

I like the new proposed rankings on the excel spredsheet.

[whteshark]

I liked the proposed ratings too.