Friday, September 14, 2012

(Week 2) Madden 13 Roster Update #3: QB's, RB's and (newly added) CB's - What really happened, looking deeper than OVR.


For months now, I have preached to Madden fans about looking deeper than OVR and into the attributes.  I have also asked EA (Donny Moore) several times to show fans the changes he makes in individuals attributes, not just OVR.  I believe fans deserve to see what is being manipulated each week to adjust  OVR.

So far, EA and Donny Moore have continued to just release the OVR changes (Click Here) and not include the attribute changes.  I've decided to do it myself.  Unfortunately, I don't have the updated spreed-sheet that EA is using, so I have to go back and forth from roster to roster to track down the changes.  Due to limited time, I won't be able to do every position each week, but I will do my best.  I will add to this blog as I get more positions done.

The first position I completed was QB.  The following table shows what individual attributes were changed in this week's (9/14/12) Madden13 roster update:


* PLEASE NOTE - The carry (CAR) attribute has no effect on OVR for QB's; the formula doesn't factor it in for this position.

Observations:

1.  If you have followed me for any period of time, you know that I do not support weekly "kneejerk" attribute changes (See more here).  That said, I want fans to see what EA is doing each week when it comes to attribute changes.  So, I will continue to provide some of that information for the rest of the season, or until EA starts providing it to fans.  Here is my EA Game Changer Idea for adding attribute changes to their Roster Update Details blog:  http://mgi-gc.easports.com/a/dtd/Include-attribute-increases-decreases-in-weekly-ratings-blog./145374-18503 (If you support this idea, please sign in and vote.).

2.  A few weeks ago, I gave Donny Moore a ton of credit for doing a massive correction in the QB carry (CAR) attributes.  Now it looks like two steps forward, one step back.  I don't believe the CAR attribute is something that should change weekly - not to mention only select QB's are being penalized for a fumble.

Get this, 17 QB's fumbled in week 1, for a total of 19 fumbles.  Of those 17, only five QB's received a decrease to their carry (CAR) attribute.

These QB's fumbled and received decreases to their CAR:  Locker, Ponder, Wilson, Dalton, and Cassel.

These QB's fumbled, but did NOT receive a decrease to their CAR:  Vick, Weeden, Gabbert, Luck, Newton, Skelton, Bradford, Tannehill, RG3, Gradkowski, Flacco, and Big Ben.

Vick and Weeden were the only QB's with 2 fumbles in week 1, but were not one of the five that received a decrease to the CAR attribute.  This is exactly the type inconsistency in Madden player ratings that I have been talking about for months.  

What exactly is Donny Moore doing here?  It's just a mess and anyone with eyes can see that.  I would support adjusting the CAR attribute every four weeks, if necessary (I think the same way about all attributes).  Let things play out for a few weeks.  The other thing you should notice, is that Donny didn't give anyone credit for not fumbling.  If you are going to have a procedure for decreasing players in certain attributes, you must also have a procedure for increasing them.  Is Donny really going to notice if a QB goes 8 games without a fumble?  I don't think so.

3.  Here's a good one.  Why in the hell did the AWR of Weeden and Tannehill get increased?  The answer is simple, the OVR formula is broken.  If you are given them an overall decrease because of their horrible week 1 performances, than you can't justify increasing AWR.  The fact is, AWR had to be increased to balance out the decreases in other areas.  This adjust to AWR prevents OVR from dropping too much.  LOOK DEEPER THAN OVERALL!

4.  This one will blow your mind.  According to www.ProFootballFocus.com, Matt Cassel was 3 for 3 in deep accuracy (passes targeted 20 or more yards down-field), yet he received a decrease of 3 points to his DAC.  Mark Sanchez was 2 for 3, but he received a 3 point increase to his DAC.  I am sorry, but this type of inconsistency is too obvious and makes no sense.  If you must make weekly changes, at least change attributes that match the player performance.  Wow.

One more. Flacco and Gabbert were both 2 for 4 in deep accuracy, Flacco gets a +2 and Gabbert gets a +1.  Again, what about Cassel?

I have to say, that I don't think any adjustment can be justified on a mere 3 or 4 attempts.  Talk about a small sample size.

5.  People keep telling me that it doesn't matter how often EA does roster updates if the changes are not going to be accurate (and consistent).  Donny Moore has done some nice things this off-season, but Madden fans deserve better than this.  I've only touched the surface here, I can only imagine how inconsistent the other positions will be.

------------------------------------------------------------

** UPDATED 9/16/12

The 2nd position I completed was Running Back.  The following table shows what individual attributes were changed in this week's (9/14/12) Madden13 roster update:

Pleaes note - CIT and SPC do not contribute to the OVR for running backs.




Observations:

1.  The most questionable change is this week's update are the decreases for James Starks.  Why you ask?  Well, because Starks didn't even play in week 1.  So why did Starks get a decrease?  Your guess is as good as mine.

2.  What about the Ronnie Brown, plus 15 in CAT.  My best guess, is that this was a correction to an attribute that Donny Moore believed was inaccurate.

3.  Eight running backs fumbled the ball in week 1 for a total of 9 fumbles.  Here are the running backs who fumbled:  Greene (twice), Reggie Bush, Lynch, McCoy, McGahee, Daniel Thomas, Ryan Williams, and David Wilson.  Of those RB's, Thomas was the only decrease listed in the EA ratings blog, but McGahee received a decrease to his CAR as well, from 88 to 87 (that was not listed on the EA website).  You have to watch out for sneaky increases and decreases that EA is not showing the fans on their website.

   I have expressed my thoughts on the carry (CAR) attribute several times.  Donny did a large correction to the CAR attributes in the off-season for RB's, I applaud him for that.

I don't feel attributes should be changing weekly, and that includes the CAR.  If EA insists on doing weekly roster updates with attribute changes, then they need to be accurate and consistent.  Why did only 2 players receive a decrease and the other 7 who fumbled didn't?  Why did Starks, Brown, and Felix Jones receive a decrease in their CAR even though they didn't fumble?  The most likely answer is that it was used to bring down OVR quickly.  The CAR attribute is the one of the most heavily weighted attributes in the RB Overall formula.

   Personally, I would like to see the carry (CAR) attribute evaluated at mid-season and at the end of the season.  Of course, this is assuming they were accurately assigned in the off-season.  If you must change them more frequently, then look at them every four weeks.

4.  Both Bennie Wells and Michael Turner received a decrease in their catch attributes, even though neither of them was even targeted during week 1.  Both of them had an original CAT attribute of 55 in Madden 13, which they deserved by having horrible drop percentages the last three years (17.1% for Turner, 15.6% for Wells).  My issue with this decrease is that they didn't do anything to deserve it.  Since neither of them had a target, they didn't even have an opportunity to drop a pass.  The CAT attribute should be about the ability to catch the football, not the number of catches.  This decrease (while small), makes no sense after just one week of play.

5.  Besides the BIG change to Ronnie Brown, here are the other seven running backs that received an increase to their CAT attribute in this update:

Dwyer: +4 from 40 to 44.  Per ProFootballFocus, 1 drop on 3 catchable targets (33% drop rate)
C. Johnson:  +2 from 76 to 78.  0 drops to 6 receptions.
Ridley:  +1 from 64 to 65.  0 drops to 2 receptions.
Martin:  +1 from 73 to 74.  0 drops to 4 receptions.
Kevin Smith:  +4 from 66 to 70.  Smith had 2 drops on 6 catchable target (33% drop rate).
Spiller:  +1 from 76 to 77.  0 drops to 2 receptions.
Greeen-Ellis:  +1 from 40 to 41.   0 drops to 1 reception.

* From 2009 through the 2011 season, NFL RB's had combined drop percentage of 8.2%.

Does this type of weekly attribute adjusting make any sense?  Where is the consistency?  Hell, the players with the drops received the largest increases in week 1.

If you need reminded of just how bad the Madden13 catch (CAT) attributes are, Click Here

6.  Let's talk about elusiveness.  Donny Moore mentioned (on this week's Madden poscast), that he utilizes Profootballfocus and their elusive ratings for the RB elusiveness attribute.

Here are the increases and decreases for this week in regards to the elusiveness (ELU) attribute.  I have included the corresponding rating from PFF for week 1.


That is not a typo for Spiller, he had a huge rating in week 1.  That said, this is too small a sample to change any players elusiveness (ELU) attribute in my opinion.  I am not going to say anymore, just absorb this table and the inconsistency.

------------------------------------------------------

UPDATED 9/17/12


The 3rd position I completed was Cornerback .  The following table shows what individual attributes were changed in this week's (9/14/12) Madden13 roster update:

Pleaes note - CAT, CIT, SPC, and BSH do not contribute to the OVR formula for Cornerbacks.














Observations:

1.  Be sure to note that the catch (CAT) attribute does not contribute to the OVR of the Cornerback position (Click Here).  Therefore, you could see big changes from week to week in the CAT attribute, without having an impact on OVR.

Three CB's received a decrease to their catch attribute in this update, but nine received an increase to their catch.  Let's take a look at the increases:

Rogers-Cromartie:  2 ints = +1 CAT
Jennings:                 2 ints = +10 CAT (from 54 to 64)
A. Cromartie:          1 int = +1 CAT
Porter:                    1 int = +2 CAT
Jackson:                 1 int = +1 CAT
J. Jenkins:               1 int = +1 CAT
K. Wilson:              1 int = +9 CAT (from 45 to 54)
Finnegan:                1 int = +4 CAT (from 53 to 57)

M. Adams:  0 int = +1 CAT

In all of the NFL, 12 cornerbacks intercepted passes in Week 1.  Eight of them received an increase to their CAT attribute in this update.  The four who did not receive an increase were:  Joe Haden (CLV), Johnathan Joseph (HOU), Darrelle Revis (NYJ), and Richard Sherman (SEA) Video of week 1 INT.

So, does an INT get you an increase in CAT or not?

2.  We all know Donny Moore endorses ProFootballFocus.com as one of his most reliable sources for NFL player analysis.  Well, here is an example of him paying very close attention to their coverage (COV) grades.  12 Madden CB's received a decrease to their:  MCV, ZCV, or both this week.  Of those 12, 11 received a negative coverage grade from ProFootballFocus in week 1.  That is a pretty good correlation and if Donny must make weekly MCV/ZCV adjustments, at least he is using an accurate source to make those changes.

Nate Clements was the only CB who received a positive coverage grade (+0.7), but received a decrease of 2 points to his MCV.  Bengal fans might want to ask Donny for an explanation on that one.

On the other side of things, 13 CB's received an increase to their:  MCV, ZCV, or both.  Of those 13, 12 received a positive coverage grade from PFF.  Chris Cook was the only player to receive a negative coverage grade from PFF (-0.2) and received an increase of 1 point to his MCV and ZCV.  Viking fans got lucky on this one.

------------------------------------------------------------------------------

That's it for this blog post.  For the next roster update, I will select some different positions to look at.  My main goal with this topic, is to get EA and Donny to reveal these changes to the fans.  It would be very easy for Donny to note what attributes he is changing and include it in the Weekly Roster Update details.  There is a reason they choose not to.  I believe it's because many of the changes a complete guesswork just to increase/decrease OVR.  This type of blog takes a lot of time, because I have to compare the newest update to the previous one; I have to record both sets of attributes to find the changes.  It would not take that much time for the person who actually made the changes.

Madden fans (and players for that matter) deserve to see ALL of the attributes adjustments on a weekly basis.  The only reason for not showing fans this type of information is to avoid the criticism that follows.  It much easier for EA keep fans focused on OVR;  that way, the inaccurate attribute adjustments are rarely noticed.

Have a good evening.








4 comments:

  1. I would love to see HB and DE! Im dying to see what Stevan Ridley and Chandler Jones got for updates!

    ReplyDelete
  2. profootballfocus and football outsiders should really collaborate with EA to make a fine tooled ratings system or atleast EA should hire its own game charters to bring more realism instead of requesting arbitrary crazed fan-dized suggestions for ratings.

    ReplyDelete
  3. Hey I've been searching all over for a concise explanation of how ratings affect the players exactly, and how or what the OVR formulas are. Do you know if that is all compiled in one place? and if so what is that?

    ReplyDelete
    Replies
    1. http://maddenmanniac.blogspot.com/2012/01/madden-12-what-attributes-effect-ovr.html Try this blog that I did on Madden12 OVR formula. I don't believe they've changed the formula one bit for M13.

      Also, check out video for some explanation on what some gamers think they do - http://www.maddentips.com/madden-13-ratings-what-do-they-do/

      Delete