Pro Football Focus to adjust player ratings

[YogiGarebear89]

Quote: Originally Posted by SStory1993

He is an 89 to begin this season History of his ratings based on my HOF scale is as follows at the end of each season: 2017:89 2016:90 2015:87 Rookie Rating: 77 Things I would like to point out: - OBJ best season so far is his rookie year, but being that he only played 12 games, I am still taking his rookie rating into context which brings down his rating to 87. If I based my ratings solely on one season his rating would've been significantly higher, however the other side to that story is if everything is based on one season it will have major negative results on players that might have a down season such as JJ Watt's 2016 season. Where his rating based on that one year is a 67, which I feel doesn't nearly reflect the rating JJ Watt has earned/deserves.

Interesting

Sent from my SM-G950U using Operation Sports mobile app

[Prilosec]

I only edit to get sim stat realistic. great work and efforts you are doing.

[SStory1993]

Quote: Originally Posted by chunt04

What is your theory on rookies and younger players in the league? How do you handle the incoming rookies from drafting? I'd love to share ideas, as I may tinker with a roster once the 53 man roster comes out. Sent from my iPhone using Tapatalk

I hope this answer your questions correctly.

Generally speaking my rookies are on a 80-20 scale, most rookies I get the information from ESPN, NFL, and now PFF due to their rookie scouting reports that came out recently. Players that are 80 are few and far between, but generally speaking each draft is going to have a player higher than 76. For example Myles Garrett in this year's draft is the highest rookie player and the only 78 overall rookie. While most rookie are going to fall under 70.

I DO NOT base my ratings on when a player was picked, but what grades he has been given. For example Mitch Trubisky (75), Deshaun Watson (75), Patrick Mahomes (73). Regardless of what round they got picked this would be their rating.

Generally speaking a rookie at the end of the season will never be in the 90's, highest from last year is ezekiel elliott (88), Dak Prescott (86), and Joey Boss (85) who would've been higher had he played a full season.

The incoming rookies in franchise mode I have struggled with. It is my goal to make this roster as compatible as possible, but when it comes to generated rookie ratings I generally take the overall that they have an adjust it as follows:

M N
85 80
84 79
83 78
82 77
81 76
80 75
79 75
78 74
77 74
76 73
75 73
74 72
73 72
72 71
71 71
70 70

From the 70 ovr point I tend to not adjust. Probably something that I should change, but just haven't since I get more fixated on the roster side of things.

I use most pro day and combine information from different websites (mainly the website FBG used) to get physical attributes assigned, and than use the date from PFF draft guide to translate as much of their advanced grading into ratings as possible.

[Kingbailey87]

Are your rosters shared on PSN?

[SStory1993]

I work on xbox, but that doesn't mean we couldn't get it on the PS4 platform.

However, most of my ratings have been based on what I would consider to be data analysis converted into a madden format on a HOF scale. HOF scale meaning that players that have 99 in a certain category be it OVR, STHA, SPD, CTH, etc... are the best at that position or top 3 in all of football history. When I apply these metrics to just the QB in madden the outcome doesn't look good. The overalls I showed for my players on the first page is the OVR I believe they should be based on HOF scale ratings using PFF as my main data. The problem I have come across is EA has the QB scale for what makes an OVR in madden 18 way backwards.

What makes Tom Brady, Tom Brady?

Is he the fastest QB?

No

Does he have the a rocket for an arm?

No in fact last year his furthest completion he made in the air was 57 yards, which is about the league average.

Is he the most accurate?

No when it comes to yards under 10 there are at least 7 qbs better than Tom Brady in accuracy last year. Yards 10-30 Tom Brady was the most accurate last year, and tends to be in the top 5 year in and year out. However, balls thrown 30+ yards Tom Brady is one of the worst starting QB's in the league, throwing 27% in 2016 and 14% in 2015. (League average is around 35%)

Can he throw on the run? Yes, but I wouldn't say he is top 5 in that category when you have Aaron Rodgers, Ben Roethlisberger, Andrew Luck, Russell Wilson, etc.

Is he a master of play action? Ill give him that, but Brock Osweiler was consider a QB that was good at play action and look what that did.

I say all of this to point out that Tom Brady is not this golden goose when it comes to how a QB is made, let's say in all these categories he breaks top 5 in the NFL in 2017, that means he MIGHT be top 50 of all time in those categories.

BUT! What makes Tom Brady, Tom "GOAT" Brady?

It's that his intelligence as a QB has been rivaled by only a handful of QB in the history of football, and arguably no one has the game down at that position like Tom Brady has in NFL history.

Yet, the ratings that matter to a QB in Madden are as follows. (Adjust rating by X amount of points affects OVR rating by 1 point)

Madden

THP 2.5
STHA 3
MTHA 3
DTHA 4.5
AWR 5
SPD 12
PAC 14
TOR 26
AGI 26
ACC 54

My opinion, but this rating formula is all wrong. Physical attributes should matter less for QB than knowledge and playing the QB position. Ideally it would fall more in the lines of:

AWR 2.5
STHA 3
MTHA 3
DTHA 4.5
THP 5
PAC 12
TOR 14
SPD 26
AGI 26
ACC 54

And that is if I just kept Madden formula (wouldn't). I mean literally EA had to up Tom Brady's OVR to 99, if you edit him by dropping anything 1 point down Tom Brady's OVR becomes 94-95.

Why does this all matter?

It really dampens my hopes to bring an accurate roster based off of any metrics much alone PFF. Taking actual numbers and converting it to successful ratings that can be translated to the game and than be used in CFM. If anybody has ideas, I would love to hear them. But this just doesn't look good.

Update: Ran the roster in CFM, where i have adjusted my rating scale for MED, & DEEP ACCURACY. In CFM ratings are coming out with positive results. Will continue to update.

[JRT2006]

Quote: Originally Posted by SStory1993

I am putting a roster together using pff statistics as we speak. My focus is to bring madden ratings to a HOF type system similar to 2k where players rated at 99 are the best at their position. At least with the data I have available. My ratings are based on a 32 game progression favoring the most recent 16 games. I am trying to incorporate the ideas the FBGrating used, but translate it where you can actually use the rosters in franchise mode. I am adjusting the speed, acceleration, agility, and vertical based on the best information I can retrieve and than using Signature stats to transform tackle, pass rush, run-stop, catching, etc. This is a project I have worked on for many years, and I believe this year is the year to share it. Would love to have help implementing it. Here are some overalls based on a HOF scale using PFF ratings. Tom Brady NE 98 Aaron Donald LAR 98 Julio Jones ATL 96 Khalil Mack OAK 96 Luke Kuechly CAR 96 Rob Gronkowski NE 94 J.J. Watt HST 94 Matt Ryan ATL 93 Aaron Rodgers GB 93 Marshawn LynchOAK 93 Le'Veon Bell PIT 93 Mike Evans TB 93 Marshal Yanda BLT 93 Brandon GrahamPHI 93 Cameron Jordan NO 93 Von Miller DEN 93 Calais Campbell JAX 93 Eric Weddle BLT 93 Andrew Luck IND 92 Russell Wilson SEA 92 Antonio Brown PIT 92 A.J. Green CIN 92 Travis Frederick DAL 92 Kawann Short CAR 92 Chris Harris Jr. DEN 92 Greg Olsen CAR 91 Trent Williams WAS 91 Andrew WhitworthLAR91 Joe Thomas CLV 91 Zack Martin DAL 91 Ndamukong SuhMIA 91 Fletcher Cox PHI 91 Geno Atkins CIN 91 Jerrell Freeman CHI 91 Devin McCourty NE 91 Drew Brees NO 90 T.Y. Hilton IND 90 Doug Baldwin SEA 90 Larry Fitzgerald ARZ 90 Ryan Schraeder ATL 90 Tyron Smith DAL 90 Rodney Hudson OAK 90 Michael Bennett SEA 90 Justin Houston KC 90 Damon Harrison NYG 90 Leonard WilliamsNYJ 90 Mike Daniels GB 90 Bobby Wagner SEA 90 Sean Lee DAL 90 Dominique Rodgers-Cromartie NYG 90 Casey Hayward SD 90 Kam Chancellor SEA 90 Reshad Jones MIA 90 Eric Berry KC 90 Love to here feedback on what you think.

Are these the signature stats you're using? Curious how you determine short and medium accuracy, from what I remember they only detailed deep passes as far as in depth distances.

Still really looking forward to this, wish Madden would revamp their whole ratings and get rid of overalls

[SStory1993]

Quote: Originally Posted by JRT2006

Are these the signature stats you're using? Curious how you determine short and medium accuracy, from what I remember they only detailed deep passes as far as in depth distances. Still really looking forward to this, wish Madden would revamp their whole ratings and get rid of overalls

Unfortunately, I do not have access to the STHA, MTHA adjusted for drop balls. As PFF for some reason doesn't allow you to see all of there sig stats that are available even when you pay for the $1,500 membership. I mean how nice would it be to know effectivness of zone cov vs man cov, or bull rush vs swim move vs spin move. They have all theses stats, yet not available.

So in the mean time I do the best I can with what I have. STHA and MTHA is data I gather from tallying up all the attempts and completions from each passer by yards. You can find this stat at several websites if you know what you are doing.

My first thing is to use signature stats, from there I go to other sources be it nfl.com, pro football reference, pro football outsiders, and espn to than determine what would be the best stat for each rating, come up with a formula that fits the mold of the game, and allows a HOF scale.

[rovert22044]

TheGame8544 and I are working on a Madden ratings calculator that is based off PFF Elite Signature Stats. It will be included in our roster project this year when we are finished with it.

Here's our roster thread:

Simulation Football Roster [Madden 18 / PS4]