I had previously written about how the Madden NFL video game series from Electronic Arts has failed to simulate football by using a shortened quarter length to keep games around 30 minutes long. These shortened games lead to a rushed pace of play, fundamentally change the strategy of football, and also affect other aspects of balance and game design that are not easily fixed by simply setting the game to 15-minute quarters.
This time, I'm going to move away from the rules of the game, and look at more specific game mechanics that fail to simulate how real football players actually play football. This installment, and the next, will look at how real NFL quarterbacks make reads and go through progressions, and then at how defensive pressure packages are used to disrupt those reads and progressions to force the quarterback to make bad decisions. Then we'll look at how Madden completely fails to model these aspects of football, and the various ways that EA and Tiburon have tried to fix or cover up these problems over the years. Some have worked; others have been little more than band-aid solutions.
See this blog in video essay format on my YouTube channel!
How Madden succeeds at simulating football: pre-snap reads
Let's start with some good faith towards EA and Madden and talk about the things that the game actually does get fairly right: pre-snap reads. As a QB in Madden, you'll be looking at whether the middle of the field is open or closed before the snap, and this will give you a reasonably accurate idea of whether the route concept that you called will be successful. If you call a cover 2-beating post or dagger concept, but the defense comes out in a Cover-3 look, with a safety in the middle third, then you will be well-served to either adjust the routes using hot routes, audible out of the play entirely, or call a timeout to regroup and come up with another play.
Madden players can make sure
that a blitzing Mike LB is blocked.
Over the past few years, Madden has also gradually introduced concepts such as reading the Mike linebacker. This determines who the blockers will prioritize blocking, which can be important if the defense sends multiple blitzers. Identifying the most dangerous blitzer as the "Mike" ensures that someone on your offensive line will try to block him. Usually, this will be the inside-most blitzer (the one lined up closest to the center). You can also slide pass protection left or right to deal with an overload blitz, and can also assign a double team in order to neutralize a particularly dangerous pass rusher.
To Madden's credit, it gets most of this stuff right. Hopefully all the mechanics that I just mentioned are still in the game by the time you're reading this, and they haven't been stripped out by Tiburon in order to make room for some new gimmicky feature...
A Madden user can make many of the same reads that a real NFL quarterback would. The game will even highlight the key reads before the snap on certain plays to remind the user how to execute the selected play. Good stuff. I don't have many complaints here. Defenses can even disguise coverages, can fake blitzes, and use other similar tactics to try to fool the human user and force a bad read. Again, good stuff. The problems begin when a CPU QB steps on the field, and only get worse when the ball is snapped.
... until the CPU QB steps on the field
Madden doesn't really have any mechanics in place to simulate the CPU team making pre-snap reads (either on offense or defense). On the highest difficulties, I'm also fairly confident that the CPU knows your offensive play call, and picks a defensive play that is supposed to be effective against that concept. It then executes the called play, regardless of what the human player shows the CPU at the line of scrimmage. The CPU rarely make their own audibles, nor does it ever call timeouts if the user offense has an advantageous mismatch or if the defensive play is leaving someone uncovered. This is a big reason why audibles and hot routes are such an effective part of Madden's meta. Sure, the CPU QBs won't "see ghosts", so to speak. But the CPU also won't adjust to what it sees at the line of scrimmage. which means you can easily exploit the A.I. by modifying the play at the line of scrimmage.
Audibles and hot routes can easily break the defensive A.I.
In a human vs human game, you can play mind games against the other user. Even so, offensive adjustments can still be exploitative because defenses still aren't great at reacting to user adjustments. Even if a user-controlled defender makes a perfect read, a perfect pre-play adjustment, and perfect play execution, that user can still only control one player at a time. You can't control the other 10 players on the field, which means their A.I. and play awareness has to be good enough for them to do their damned jobs. But I digress. That topic will have to wait for another essay on another day.
Progressions are designed into the plays
Once the ball is snapped, a real quarterback (at any level of football -- not just the NFL) will have a series of reads that he will have to make. Every play is going to have a target "primary" receiver. The whole play is designed to get the ball to this player. Which play is called will depend on what defense the coordinator or head coach is expecting, and they'll call a play designed to get the primary receiver into a soft spot in the selected defense -- or for the primary target to beat his assigned defender one-on-one if the play-caller suspects man coverage. That might mean calling a post route to exploit the hole between the safeties of a Cover 2, or combining a post with a wheel route to force a corner in Cover 3 to have to chose which receiver to run with, and so on. In many cases, the routes that the other receivers are running aren't so much to get those receivers open; rather, they are designed to draw defenders into positions that give the primary receiver as much room as possible. They are essentially decoys.
If the primary receiver is covered, then the QB will have to go down the list of prioritized targets to find someone who is open. Usually, if the defense covers the intended receiver, then it is because they pulled a defender out of another assignment in order to take away that primary target. If so, then the play will have a secondary target that is usually going into the part of the field left vacant by the defender who decided to cover the primary target. If the secondary target is taken away, then the QB will look at a tertiary target who is similarly running to a part of the field that is hopefully left vacant by the defender who moved to cover the secondary target. And on down the line.
A real QB cannot see the entire field, and has to go through a series of reads to find an open receiver.
(image source:
Brett Kollmann's Film Room)
Going through these progressions takes time. A QB can't see the entire football field at once. He has to go through the process of turning his head or his body in order to look at each of his reads, then also spend time watching the defense to see where the defenders are going.
This may be part of the reason why bunch formations are so popular in football today. Not only do such formations create good spacing for outside and underneath routes, and generate a lot of natural picks and rubs, but they also allow the QB to make a single read (in a single part of the field) that informs 3 different receivers. This is opposed to a more traditional formation that requires scanning from one wideout on the far end of the field, to a wideout on the opposite end of the field, then back to a tight end in the middle of the field, before finally settling on a checkdown to a running back out of the backfield.
A defense hopes that, if they defend the first and second target, then their pass rush (or blitz) will get to the QB before he even has time to look at his third or fourth read. The purpose of a blitz is to rush more defenders than the offensive line can block, thereby pressuring the QB into throwing the ball before he's even had time to finish making his second read. After all, blitzing an extra defender leaves one less defender in coverage, which means there's going to be a hole in the coverage that the offense can exploit, if the receivers have enough time to get there, and if the QB has enough time to see where that hole is. And this assumes that the defender in coverage doesn't lock his man down or make a great play on the ball -- which is something that can thwart even the best-conceived offensive play. If the blitz is quick enough, then the defense may only have to defend the QB's first read. And if that read is covered, then the QB may be forced to throw into that coverage and risk an incompletion or interception, or else that QB is going to take a sack. The sack or interception is a best-case scenario, and kind of a bonus for the defense. Really, all they are trying to do is force the QB to have to make a bad throw before the play is fully set up, in order to prevent the offense from gaining yards, and force them to punt.
This is why the first rule of pass blocking is that you block the inside-most threats first. Protect the A-gap first, then the B-gap, and so on. If you don't have enough blockers and you must leave a defender unblocked, then your pass protection should leave the outside-most rusher free. That rusher will have to run around the blockers, taking the longest path possible, which gives the QB more time to complete more reads.
Madden's Bird's Eye View
Now you could make the argument that this is how Madden works. The human user has to scan the field with your eyes to find the open receiver. If the defense gets to the QB before you can do that, then it's a sack.
That's true at a surface level, but it breaks down in practice. First of all, scanning the field from a bird's eye view is a lot quicker from a simple body mechanics perspective. The entire field is available to you at a glance, allowing you to make split-second decisions that even an NFL QB might have trouble making. Any receiver on the field is just a button press away, no matter where they are on the field or in the progression tree. The player doesn't have to turn the QB's body to face the receiver before throwing. This motion is handled automatically, in a split-second. Much faster than in real life. To make matters worse, QBs in Madden can regularly throw across their bodies, into blind spots, with pinpoint accuracy.
To be fair to EA and Madden, this same view is used by most other football games as well, and is just as much of a problem in most of those games. But since other companies aren't currently making big-budget football games at the moment, I don't really have any other game to complain about except for Madden.
Most football games use a bird's eye view for passing.
For example, if I make an overload blitz against a bunch formation by manually bringing my Nickelback into the blind-side B-gap at the last second, I might be leaving a quick route (such as a hitch) open from a bunch formation. Because of the bird's eye view, the opponent (whether human user or CPU) can instantly diagnose that and throw hot with a single button press. This is true even if the tackle lets the blitzing backer through the B-gap, rather than blocking the B-gap and forcing the end to take the longer path around to the QB. This is because the QB (whether a human user or the CPU) can effectively see the entire field and does not have to spend time going through progressions. Or Madden will just motion-shift all the blockers over to magically pick up the overload blitz. It's becoming less common in later years, but it still happens.
The CPU QB isn't even limited by how quickly it can scan the field with its eyes (bird's eye view or no) -- or by the limited vision and reaction time of a human user. The CPU QB can instantly diagnose the defense with perfect information and identify the optimal receiver to throw to. It basically just comes down to the difficulty setting and a random number generator that determines whether the CPU QB will make that optimal decision or not. To prove this point, I would like to remind you of a late-season update for Madden 17 that broke the CPU QB throwing logic and introduced what the community called "Robo QBs". The QBs would throw the ball to an open receiver within just a couple seconds of snapping the ball with near perfect accuracy, resulting in even average QBs consistently achieving 80 or 90% accuracy ratings in game after game. Madden 17's QB weren't even pretending to go through progressions. That is, after all, exactly what the game is doing: pretending to make reads. This update rendered Madden 17 (which I thought was an otherwise solid release by modern Madden standards) virtually unplayable on the All-Pro or All-Madden difficulties.
A Madden 17 patch broke QB logic and lead to every QB completing 80-90% of passes.
This bug (or feature, or whatever the heck it was) serves to highlight the underlying problem, which is that QBs in the game do not have to go through progressions. They can instantly identify an open target and throw to him without having to take the time to put his eyes or body in the proper position to make the throw.
Compared to other games
Compare throwing the ball in Madden against throwing the ball in Backbreaker. In Backbreaker, you are locked into a ground-level perspective. Your can only see part of the field at a time, which (being at ground level) can be obscured by other players, including your own blockers. You have to press buttons in order to move your view to another part of the field. It is a hell of a lot harder to throw the ball in Backbreaker than it is in Madden.
Now, I'm not saying that Backbreaker is a better game. It isn't. Though I really do wish that Backbreaker had gotten at least one sequel to try to iron out the problems and flesh out some of its better ideas and mechanics. Ah well... In any case, Backbreaker's default passing mechanic was much closer to how a real QB would play. Because of that, Backbreaker basically expected the average video gamer to be able to read the field like an NFL quarterback, which is just far too much to expect from the average gamer, and I think ended up being a big part of why that game failed.
Backbreaker has a ground-level view, and Maximum Football 19 has a delay between pressing the button and the QB making the throw.
Maximum Football 19 has a slight input delay when pressing a button to throw to a receiver. I'm not sure if this is deliberate design on the part of Canuck Play, or if it's a flaw. In either case, the effect is that you do kind of have to identify an opening in the defense in advance; otherwise, the delay might cause you to miss the throwing window. You can't just tap the button to hit a receiver with some separation from his coverage, or in a last-second act of desperation. CPU QBs, however, don't seem to be burdened with this input delay, which is part of what makes me think it's a bug.
Madden 2006 may have solved the problem 15 years ago
Neither of those games have provided ideal solutions to this problem. In fact, to EA's credit, we have to look back to an older version of Madden, which may have stumbled onto a decent middle-ground solution 15 years ago. They just didn't stick with it. Madden 2006 introduced the QB Vision Cone mechanic. This mechanic projected a spotlight across the field, originating from the QB's head. If the QB threw within this spotlight, the passes would be accurate relevant to the respective QB's pass accuracy ratings. If the QB threw outside of the spotlight, passes would be less accurate than the QB's ratings. The width of the cone would be determined by the individual QB's awareness ratings. At the higher difficulty settings, receiver icons would even fade away and disappear if they are outside of the vision cone.
Madden 2006's QB Vision Cone mechanic attempted to solve the problems in this post
-- 15 years ago!
This mechanic solved or mitigated most of the problems that I've discussed in this essay. Even though the user still had a bird's eye view of the field, you had to scan the cone across the field to simulate the QB turning his head to go through his progressions. Turning the cone also turned the QB's body, and trying to throw across the QB's body would lead to an additional accuracy penalty. And if you would scramble, you might not be able to rotate the cone in the opposite direction of the QB's scramble, which mitigated the exploit of scrambling QBs throwing across the field.
The size of a QB's vision cone actually made a considerable difference in how the user would control the QB, and so the awareness rating of the user QB was much more relevant to a user, and also had a much more tangible effect that a user could more easily understand. Having a vision cone that was big enough to consistently encompass multiple receivers (instead of just one targeted receiver), reduced the number of inputs (and time) necessary to go through progressions and make a throw. A QB like Rex Grossman was much slower in making his reads than a QB like Peyton Manning because Manning could see so much more of the field at once, and had access to throw to multiple receivers without having to move his vision cone.
A good QB like Peyton Manning could see more receivers at once, compared to a poor QB like Rex Grossman.
Having a low awareness QB also influenced your play-calling and team-building. The limited field of view meant that it was much more important to call plays that would scheme a receiver open against the defense that you expect to see; whereas a more talented QB would be much more likely to just chuck the ball up to any receiver who happens to come open. You didn't have as much time to move your vision back and forth across the field with a less-talented QB, so having the right play called, making the right read, and identifying the open receiver before the snap became so much more crucial.
Further, you would actually notice a difference if you starting quarterback got hurt, and you had to rely on a backup with a much lower awareness rating. You might actually consider looking into signing a veteran free agent if it meant you'd get a higher-awareness QB with a bigger vision cone.
If your starting QB gets hurt, you might look to sign a veteran with high awareness.
QB Vision also had trickle-down effects on defensive play that were never fully-realized within the game. The CPU QB would also project a cone onto the field, so a human user controlling a linebacker or safety could actually follow the opposing QB's eyes to potentially jump routes. Hypothetically, a clever QB (whether CPU or user) could use this mechanic to "look off" a defender by baiting a safety or linebacker in one direction, then turning his field of vision in another direction and throwing against the flow of the defense. Sadly, I don't think the CPU defenders were programmed to react to the user QB's vision cone in this manner in Madden 2006.
Defenders in zone coverage could follow the eyes of opposing QBs to disrupt the play.
Abandonware
Despite the innovation and realism that it brought to the game, the Vision Cone proved to be unpopular with a large cohort of Madden players. Some of this unpopularity was for justifiable reasons. Madden 06 was not a perfect game -- not by any stretch of the imagination. Despite the "Precision Passing" mechanic that came alongside the Vision Cone, QBs had trouble putting the ball in the right spot on many route concepts. This lead to too many balls being swatted down by defenders, despite the receiver being [in NFL terms] wide open. CPU QBs would also get stuck rotating their Vision Cones back and forth, causing the QB to spin around in circles until he gets sacked. These problems combined to make the CPU ineffective at throwing the ball.
Many users felt the Vision controls were too cumbersome.
This was compounded by usability issues with the Vision Cone controls. Users felt that it required too many button inputs and was generally hard to use. I think the same complaints are the reason that Tiburon scrapped the targeted passing mechanic that was introduced in Madden 18 and abandoned by Madden 19. But yet the abysmal "Tackle Battle" QTE mechanic remained in the game for 3 or 4 years?! Who the heck through that was a good idea? In any case, a lot of players back in 2005 and 2006 just turned the QB Vision Cone off. The following year, it was disabled by default, and a year or two later, it was completely removed from the game. EA didn't try to improve the controls or add any assist features or further tutorials to make the feature easier to use. They just cut it.
I'll admit, at the time, 20 year-old me didn't understand the sport of football well enough to be able to use the mechanic effectively. I was a lineman in high school, so I never learned route trees or how to read defenses. The mechanic practically required the user to have a basic understanding of NFL pass concepts and progressions in order to make sure that the user was looking at the right receiver, at the right point in his route, in order to be successful at passing the ball. I bet a lot of players -- especially younger players like myself -- simply didn't know route concepts well enough (if at all). As a result, this mechanic made the game considerably harder -- unless you happened to be a Colts fan.
As I've grown up and come to learn the game of football better, I now look back at the QB Vision Cone with a nostalgic understanding and respect for what it was trying to do, and how it elegantly solved a lot of problems that Madden struggles to deal with even today. And apparently I'm not alone, as I've seen more and more support for the return of the mechanic (or something similar) within the Madden community.
Would the Vision Cone work better now?
If you didn't like the QB Vision Cone, then before you turn your nose at the proposal of bringing it back, please keep in mind that modern Madden has a lot more assist features for users who maybe don't understand the sport of football as well (or the game of Madden). It also has the extensive Skill Trainer, which systematically teaches users the basics behind each and every route concept that is included within the game, and the weekly preparation mechanic in Franchise Mode allows the user to refresh their memory and practice various concepts until you've mastered them -- all without having to leave your Franchise to go back through dedicated tutorials. That, by itself, could make the QB Vision Cone a much more viable feature in modern Madden games, as compared to a decade-and-a-half ago.
Modern Madden has more assist features and comprehensive tutorials that teach football fundamentals.
EA could further supplement the Vision Cone by adding additional [optional] assist features. The game could label the progression order in the play art (both in the play-call screen and on the field), either by numbering each receiver, or color-coding the receiver routes. The primary target could have a blue route, the secondary could have a green route, followed by yellow and orange, with a "hot" read having a red route. In fact, I mistakenly thought that this was what Maximum Football 2018 was doing, but it was actually just matching the color of the receiver's route to the color of the buttons on the XBox controller. I was playing on PS4, but Canuck didn't bother to change the colors for PS4 users, and so I didn't make the connection until it was pointed out to me in an online forum.
Play art could show the order of progressions.
Another wild possibility could be to tie the progressions to specific buttons on the controller. If looking at a PlayStation controller, the first read might always be the X receiver, the second read always the Square receiver, third read always the Triangle receiver, fourth always the Circle receiver, and the hot receiver is always the R1 button -- regardless of where the receivers are on the field. And there could be an audible chime, or a visual indicator on the screen, or a vibration of the controller (or all of the above) telling the user when to move onto the next read in the sequence.This way, following your QB Vision progressions would be as simple as rotating your thumb around the face buttons when prompted.
Madden could also highlight the defenders who you should be keying off of, just like it does in the Skill Trainer tutorials. If a particular route concept is supposed to create a high-low conflict for a particular defender, then Madden should highlight that defender as the primary read to remind the user of who they should be looking at.
Madden could also have an optional assist mode that would cause the user QB to automatically progress through his reads. A user who maybe doesn't understand the progression order wouldn't have to move the cone manually to the right places at the right time; he or she would only have to wait for the Vision Cone to automatically highlight over an open receiver, then press the "throw" button. In fact, EA's own NCAA Football games had a feature called "One-Button Mode", which operated similarly when passing the ball.
The "One Button Mode" of NCAA Football basically went through your progressions for you..
Abstract band-aid solutions
I hope I've demonstrated that Tiburon could make a feature like QB Vision more user-friendly. All of the ideas for assist features could be optional in case hardcore players want to turn it all off. Tiburon never tried any of this stuff 15 years ago because EA made them move on to new back-of-the-box selling points.I wonder, however, if there might still be enough lingering resentment towards the Vision Cone that EA might be hesitant to re-instate it. But it would solve so many problems!
In lieu of an elegant solution like the Vision Cone, EA has instead relied on various band-aid solutions to try to address these problems. The game applies arbitrary accuracy penalties for when QBs are throwing on the run or if the game decides that the defense is applying a certain degree of pressure on the QB. This mechanic is fine, especially in the cases of throwing on the run or throwing off of a back foot. But it's so abstracted that it's hard to tell what the thresholds for triggering these effects are. I've seen QBs get accuracy penalties for pressure even in a clean pocket, but then another QB can turn around backwards, slip in snow or rain, then pick himself up and throw a perfectly accurate ball completely blind, with a defender in his face, with no penalty at all.
Accuracy penalties for pressure seem to be arbitrarily applied.
Seriously, what pressure?! Every rusher is blocked in a clean pocket, and the passing lane is huge.
The X-Factor mechanics of Madden 20 went on to make pass rushers much more effective at breaking blocks and sacking QBs to try to limit the opportunities that users have to throw deep down the field. This, however, has its own problems. It doesn't accurately reflect how pass rushes generate pressure, and it leads to far too many sacks. But this essay has perhaps already gone on too long, so I might have to save that topic for another day...
If you want to keep up-to-date on new YouTube projects, please be sure to check out my Patreon page. Patrons will have access to previews of upcoming content (including future installments in this "How Madden Fails to Simulate Football" series), early access to certain content, and the opportunity to provide feedback or vote in polls of what content will be created in the future.
Your support won't only shape my content. It may also help shape the future of these very football games! No, realistically speaking we probably won't (or can't) make a dent in Madden's design, but with the support of the community, we can help shape the indie games into the simulation football games that I know we all want. In fact, a few weeks ago, one of the developers of Axis 19 stumbled onto my critiques. He invited me to help troubleshoot some of the bugs that I had pointed out, and to provide feedback on a patch to fix those issues in Axis 19. So in some small way, my content has already helped to make that game just a little bit better.
Stay tuned, and stay ready for some football!
The X-Factor mechanic of Madden 20 leads to too many sacks.