A.I. of defensive players in Madden needs to be overhauled, as does the rest of the game

Current issues

Man coverage is too good

If playing defense in "2 Man Under" (man coverage with two deep safeties) in a 4-3 or 3-4 personnel group for almost the whole game is your strategy, then you can stay in the game competitively. This shouldn't be the case.

Zone coverage isn't realistic

The way zone coverage currently works involves the players dropping immediately the zone area and ignoring the play as it develops so you predetermined position. Real zone coverage is complicated, and it is understandable that simulating it should be difficult. However, defenders should at least shadow the receiver across from them for a step or two before showing the zone coverage.

Defensive line play is awful 

Defensive line realism is the  part of the game most in need of improvement. D-linemen interact with blockers in a very rudimentary way, often unable to engage multiple blockers and showing a sharp demarcation between free roaming and engaged play. The line doesn't react to the play on the field until the ball is past them.

Sliders make for a intellectually dull game

Madden already has enough of a problem of not catering to knowledgeable football enthusiasts. The emphasis on making the game more of "Sticks Skills" sort is just lame. Football at its finest is all about strategy, and the strategy and skill aspects should be well balanced. Having a slider for "frequency of interceptions" is nonsensical, for example. Interceptions happen because of three factors: improper throwing decisions/poor throws by quarterbacks, exceptionally athletic plays by defenders, and reliable catching by ball-hawks. Rather than somehow moving a slider to make interceptions more frequent, one ought to be able to alter defensive reaction times, quarterback accuracy (and when playing against the computer, decision making), and most obviously, defensive catching skill. I once counted thirteen dropped interceptions in a normal length Madden game. That is preposterous.

Athletic quarterbacks are too dominant

When a player can preform better with the likes of Mike Vick and Colin Kaepernick compared to the likes of Peyton Manning or Aaron Rodgers, there is a problem. Skilled pocket passers have had far better success in the NFL than dual-threats, and the reasons for their success can be simulated in the game. Footwork and quickness of release should be more important skills. Also, ball placement, with the height and trajectory of the ball, are intricacies that pocket passers are more effective with.

Biased, opinion-based athletic ratings

Before they ever touch the field, The NFL runs almost every one of its players through their scouting combine, in which players perform drills that are designed to measure the same abilities - speed, agility, acceleration, etc. - that the drill scores and times assign numbers to. Additionally, we live in an environment with massive amounts of data, some of which strongly indicate specific strengths and weaknesses in players.

However, it is apparent that these measurables are often disregarded by those who assign ratings. The examples are numerous: 

Devin Hester, an outstanding kick returner for the Chicago Bears, has used a deadly combination of quickness, speed, agility, and instincts to run his way into the record books. However, his 4.50 40 yard dash time signifies a player who has just solid pure speed. In Madden 2008, Hester was rated a 100 in "Speed", making him the fastest player in the game. Considering the evidence at hand, and the fact that the NFL doesn't just sign Olympic sprinters to return kicks, assigning him the top speed rating seems unfitting.

Brandon Marshall is a fabulous receiver. His combination of strength, route running, intensity, and penchant for the spectacular play has made him one of the league's best receivers over the past seven years. However, avoiding drops has never been a strength of his: he was among the NFL's worst last season with a 13% drop rate. It is understandable to give him high marks for spectacular catching and catching in traffic, but one of the worst in the league with regards to drops should not have a 98 "Catching" rating.