The problems with EA’s (Spot Dropping) zone coverage explained

[Oneback]

I know I have gone over this before in a previous thread but I wanted to explain the issues with spot dropping zone coverages.

While many of us hardcore players have begged for pattern reading coverage in recent years EA has decided to stick with spot dropping zone coverage, so what exactly is spot dropping?

On a pass read, pass defenders drop back to a predefined area or landmarks on the field. Now there are two fundamental problems with EA’s spot dropping zone coverage’s, defenders do not break on the ball and they do not gain enough depth.

In real life when defenders are taught how to spot drop the very first thing they must learn how to do is drop to their area while reading the QB’s eyes. Defenders in the game do not do this as there is nothing for them to read and therefore just drop to their area and wait for the ball to be thrown and then react (I’ll get into this more later).

The number one threat to a zone defense is a pass to the intermediate area, thus drop defenders are taught to gain depth. For example, a hook to curl linebacker could have a landmark of 12-14 yards deep. Next time you are in the game take note on the depth of the linebackers drops…8 yards, maybe 10…there is a big difference in those 2 to 4 yards. Think about where you are completing those intermediate routes.

Now, why does EA use spot dropping zones? First of all it’s easy to program: Cover 2 – corners cover flat, outside linebackers to the hash, middle linebacker to the middle and safeties have deep halves. There is nothing to it and thus it’s simple to have the defenders execute these drops time after time after time. There are more eyes on the ball which equals better pursuit and with EA implementing the programming where a defender must be able to see the ball to swat the ball this style of defense works great. Poor throws can easily be intercepted (I’ll get more into this later as well).

All of those positives sound great, right? Well I think with the outcry for pattern reading coverage’s we all know there are negatives as well. The problems with spot dropping zones are that defenders can be stuck covering grass, your defenders are assigned to guard an area and read the quarterback (there’s that again)…not cover a receiver. There are many throwing windows in the defense, ever tried to cover a slant route with a spot dropping zone defense? This style of defense requires a better than average pass rush (another problem with NCAA). Proficient passing teams can tear apart the zones (I think we’ve all seen this) and this style of defense is VERY susceptible to intermediate routes even when the drop defenders gain sufficient depth (most of us are complaining about this, passes over the middle).

To my key issue with spot dropping in NCAA…in real life spot dropping is centered around reading the quarterback’s eyes/shoulders, but what do are digital defenders have to read? Both the human player and CPU can throw to any receiver on the field without making any movement at all with the quarterback (his eyes and shoulders are always straight down the field).

So how did EA get around this? Early on they programmed zone defenders to react the moment you pressed the button to throw to a receiver (remember the first year of next gen). They have since toned down how quickly defenders react but I am certain this is still how zone defenders react to a thrown ball and why balls thrown further down the field and to the outside have more of a chance to have coverage effect the play than balls thrown over the middle, the defenders simply have a longer time to react. Years ago I believe EA was attempting to get away from this type of reaction from the defense with the passing cone but users hated it saying it made the game difficult to play. A passing cone or some sort of double tap on the receiver’s button I believe would solve some of the problems.

Adding another feature to the game aside how do we fix zone coverage today and going forward?

Zone coverage deep down the field and to the outside isn’t really a problem although it could be tweaked some as well. Why is this though? The first reason is obvious: The ball takes longer to get there and thus the defenders have a longer time to react, the second is corners and safeties have much higher zone coverage ratings than do linebackers.

There are two main issues that hinder our digital linebacker’s ability to play effective zone coverage. There is nothing for them to read and the ball gets to them too fast for them to react. Would higher zone coverage ratings help?

Improving the pass rush would help as well, remember one of the negatives of spot dropping coverage is that it requires better than average pass rush? This is one area were offline and online dynasty players have an advantage at least initially over online players (thinking that EA may use live tuning updates to up the pass rush).

The biggest thing that is needed though with the current construction of the game is pattern reading coverage’s. I say the current construction of the game as I do not see the passing cone making a return and I don’t know that EA would ever add a simple passing cone feature by making the user tap the receiver’s button twice to make a throw.

So what exactly is pattern reading coverage?

Rather than the drop defender taking a drop to a landmark and wait for receivers to arrive, pattern match coverage involves taking coverage to the most dangerous threat a defender recognizes in his zone. In this coverage defenders are taught to key certain receivers when they read pass. Usually the #2 receiver to their side, from this receiver’s action, they can diagnose whom the most likely threat to their zone is. Now after a certain point, pattern-match coverage turns into man coverage (it would take another thread to get into all the specifics).

Man coverage is obviously already programmed for the game so all that needs to be programmed is the diagnosis portion of this style of defense. Not being a programmer myself I am unsure on how difficult this would be to program, although diagnosing which offensive threat to pick up in coverage seems akin to diagnosing which defender to block in a zone blocking scheme so maybe there is hope, for next year.

[fsuprime]

so would u basically end up with a defensive option route...programming might be a problem when offense runs option route vs defensive zone option route, but other than that it does not seem too complicated (i have no programming xp)

[Oneback]

One day when I have the time I will devote a thread to pattern-matching coverage, but the simple answer: There is an outside linebacker in pattern-matching zone coverage he see's #2 (TE in this case) run an arrow route (short route to the flat) he would then look outside for a in-breaking route by #1 and play in underneath "man" coverage on him.

[baseballplyrmvp]

i like the double tapping the button idea.....i think that we should also have to choose a receiver pre snap (or after the playcall) to be our primary read. that would help with zone defense too, as then the defenders would immediately have something to react on since your qb would be making his first read.

[gschwendt]

While it won't help with what you describe Oneback, I believe that the NCAA10 gameplay camera should come back simply because it's too easy to pass with the NCAA11 camera. Take a look at Rudy's blog where he has screenshots of each: http://www.thegamingtailgate.com/for...-Camera-Angles

Yeah everyone loves seeing more of the field, but on NCAA11, we're seeing too much and thus it's too easy.

[Oneback]

There is yet another thing that could be implemented...bring the camera in tighter and as you progress through your reads the camera rotates as well to show your reciever in full veiw. I wouldn't want huge swings of the camera but subtle movements.

[JeffHCross]

Oneback, four things.

1. I'm surprised, but you didn't seem to address the issue with spot dropping and the hash marks. It does not appear that the spots designated for the defenders to drop to shifts with the ball. So the spots for LB Hook Zones, for example, are the same whether the ball is in the center or on the hash. Personally, I have never seen a defense designed in this manner. Most defenses will have one of the LBs drop further to the field side if the ball is on the hash, but not to the point where they would be if it was in the middle of the field. And, as you said, normally these players will be reacting to threats in their zone. They're not going to shift field side if there's a heavy concentration of receivers on the boundary side.

2. I haven't been able to put my finger quite on this one ... but one observation I've had is that the defenders "spots" seem to be dictated by the zones of the other players around them. This does suggest some semblance of knowledge as to how they fit into the overall defense. However, it is strange, at best. Let's imagine a 3-4 defense with all 4 LBs dropping into coverage. Essentially, what I've observed is that the ILBs zones will be in a different position dependent on whether the OLBs are in hook zones or buzz zones. If the OLBs are in hook zones, the ILBs will remain largely in the center of the field (depending on the hash issue as noted in #1). If the OLBs are in buzz zones, the ILBs will abandon the entire middle of the field. It is almost as if they are playing Cover 2 at hook zone depth, and splitting the entire field between the two of them.

Actually ... damn ... that may be exactly it. I hadn't thought about that before writing it, but that describes practically all of the problems I see.

3. Shift Coverage Left / Right does practically nothing. It does have a limited effect on the safeties in a Cover 2 man defense, but as far as I can tell it has no effect on the underneath coverage in a normal zone defense. If, as a playcaller, I was able to call a Cover 2 defense on the hash mark and tell my defenders to shade to the boundary side, I think this may resolve some of the issues we see.

4. This is probably the most important one. I was watching the Mississippi State - Auburn tonight, and I saw example after example of what you termed "pattern reading coverage". There were numerous plays where the defenders (most notably the LBs) were not reading the QBs eyes or shoulders, but were just reacting to the most obvious threat in their zone. This, going by memory, is what NCAA did in zone defenses for years; however this may be an example of nostalgia clouding my memory. This is the central read that high school level defenders are taught ... if they learn to read eyes, great. If not, at least react to the closest threat.

If nothing else, maybe they could adapt the zone defense logic (or something like it) from NCAA Basketball. Essentially "read your biggest threat" is how zone defense is programmed (and played) in basketball games.

[Oneback]

Zone drop landmarks do not change in real life when the ball is on the hash...Cover 3 has the hashes and flat covered underneath, Cover 2 has the hashes, flat and the middle covered by the MLB. The entire field still needs to be covered, you cannot bunch your zone drops to one side of the field as you would open up the other side.

Yuo are correct though in saying that defense's change when the ball is located on the hash, this is because teams will call a rotation coverage (Cover 3 Cloud). The problem we have though as playcallers in NCAA 11 is that EA has only given us a limited number of these types of plays (which is a big problem as the ball will typically be on one hash or the other most of the time):

In the 4-3 playbook there are only 6 plays that are designed hash zone defenses.

4-3 Normal: Zip Shoot Gut (Blitz)
4-3 Under: Over 3 Strong
4-3 Under: Zone Blitz (Blitz)
Nickle Normal: Over Strong
Nickle Normal: 3 Weak Roll
Dime Normal: Strong Roll 3 (Blitz) will work but you will need to hot route the boundry slot corner to a buzz zone.

What EA has done with the defenses or at least zone defense isn't funadmentally flawed, the problem is that 1. there is nothing for the defense to read and 2. the offenses in the game have become too complex for the style of defense we are given.

[JeffHCross]

Hm.

I'm not sure if I'm confusing my memory with what other "Xs and Os gurus" have said elsewhere (I've seen it insisted that zone coverage should be different on the hash), but my basic memory is that the defensive zones will shift, not drastically but still do so, with the placement of the ball. I.e. a MLB is covering hook/curl in the dead middle of the field when the ball is in the center of the field, but he's covering somewhere between hash-and-middle when the ball is on one hash.

Maybe this is just a function of the way LBs drop while they're reading the receivers. Or that they're playing men that go in/out of their zone while they drop, so the drop looks different.

In the game, meanwhile, an MLB bails as soon as the ball is snapped on a drop directly to the center of the field, regardless of ball placement. Last night with a 4-2-5 I saw the two LBs drop straight to the hash, leaving the center of the field completely unguarded.

Maybe you're right that a coverage roll is what I'm thinking of, but that doesn't strike me as covering all of it.

[Oneback]

Here are Saban's Field Coverage Rules:

A. Field Defense - Front is set to the wide side of the field.
B. Coverage is weak rotated to Boundary.

Stack Coverage Rules:

A. Stack Defense - Front is set to Boundary
B. Coverage is Strong Rotated to Field.

When the offensive strength is to the boundry side the defense will check sky coverage.

[JeffHCross]

Alright, that does it, I need to completely refresh my defensive knowledge.

[gschwendt]

I just had an idea occur to me... it may be something that every football gamer has thought of before today but it just hit me: why not target based passing? That is, instead of icons over receivers, there would be a target laying flat on the field and you would move that target around using the R-stick. Based upon the QB's awareness would determine how fast the target would move around the field, and based on the QB's Throw Accuracy, the target would be larger or smaller (larger means more possibility for errant passes). Positives I can see from it...

• Since you don't have receivers assigned to buttons (L1 would be throw to target, R1 would be fake), you could then find other purposes for the face buttons.
• Let's you throw the ball away how you want to, just get out of the pocket and aim to the sideline.
• This would allow players to use timing routes much better along.
• Finally, this would resolve the problem that Oneback described above since the QBs head/body would turn as the target does.

Obviously this would be a huge change in philosophy for the game... hell button passing has been in since Madden 1. I can hear the non-thumb-jockeys complaining now (which I understand) but let me know any thoughts.

[Oneback]

I think that would be too passing cone-ish for the masses and it would be exteremly difficult to elude the rush, move a target and properly lead a reciever all in 4 seconds.

[JeffHCross]

Side note: I don't mind the idea of the passing cone returning, though they should be able to tweak it to be more acceptable to the masses. The more I think about it, the more I think something along those lines is the direction we're going to go in eventually.

Initially, I don't like your idea, G, both because that requires shooter-like stick skills and it just seems ... overly-gimmicky, for lack of better term. That's not quite the term I mean, but I'm imagining how annoying it is to move the camera target around on replay, then imagining that while the replay is actually moving ... yeah, no.

I mean, if you ask the average to above-average guys here, I think we can all agree that while we have moments that we're not paying attention to the rush and we're looking downfield, that's a far cry from putting a target down where we want to throw the ball. I mean, forget thumb-jockeys ... now you have to actually have a level of football knowledge because you need to know your route, know where the break is, know how the coverage will react ...

The other thing to consider is that Backbreaker introduced passing that uses the right stick to flick between the receivers. In my opinion that was an absolute failure, because the controls weren't very responsive.

[AustinWolv]

I like the concept, G. The problem is I'm not a quick-twitch FPS guy, even though I try. So, seeing the coverage, thinking I know where to go with the ball, trying to find the ball 'reticule' and get it where I want it.....

But I like the concept.

I just think it would need to be very carefully balanced, as it would piss me off to no end to know where I want to go with the ball and constantly fight trying to get some targeting reticule in the right spot or overshooting it trying to move the stick quickly somewhere, etc.

And good grief, give me the option to throw the ball at a RB's feet on covered screens.

[morsdraconis]

It's an interesting idea G, and definitely one I've thought of.

Personally, I REALLY liked the idea behind the cone for passing, I just didn't like the execution of it with having to hit extra buttons to throw the ball to the receiver like that. The problem was you could use the right analog stick to the move the cone around, but then you had to do that real fast and then hit the button to throw the ball to the receiver, once again, making it far too difficult to actually throw the ball with timing at a receiver (especially while getting rushed and having to make a split second decision).

In the end, the cone is definitely the best way to truly emulate the QB aspect of the game without feeling like you're in a tunnel (ala Backbreaker pre-patch) there just needs to be either a better way to implement it or more responsive in it's implementation (part of the issue with the previous cone setup was it seemed like any amount of lag at all would absolutely kill passing because you had to wait another second to see if the cone changed to the next receiver or not).

Maybe go the route of instead of it switching between receivers (again, ala Backbreaker pre-patch) maybe switch between sections of the field with also the ability to lock onto individual receivers (with the previous cone controls or even something different). That way you could flick the right stick left or right between 3-7 areas of the field (more areas for a lower awareness QB to simulate it being harder for them to progress to their 2nd or 3rd options in the play). You'd still have your cone (size defined by your QB's awareness) that you could use to lock onto individual receivers and such but also have the flicking of the right stick to make split second changes.

[gschwendt]

I think if they ever brought back the cone, they'd need to make the primary receiver targets the shoulder buttons (L1, R1, L2, & R2). I honestly never used the passing cone (didn't play Madden, wasn't required in NCAA), but I can only see trying to use your thumb on both the R-stick and the face buttons being a pain in the ass.

My idea is that the targeting reticule would move rather quickly so you should have enough time to get from the middle of the field to your intended receiver... that and start going towards your target before you have to.

[Oneback]

I still believe the best way to do this with the least amount of impact to the user would be by using the already existing passing icons. As your dropping back you tap X to look at yout X reciever if he's open you then hit the X button again as we do now to throw to him. If you don't like it you have the option to turn it off and the game plays much like it does now.

I think this would also help combat the run and chuckers that play online. As when you have a defender bearing down on you because your running around behind the line of scrimmage it'll be tough to hit X so the QB looks then hit X again to throw.

[rhombic21]

I don't think the target thing is workable. Back in the day you could choose an option to pass that way on some of the old PS1 games, and it was borderline impossible, especially on anything down the field.

Initially I thought that they should just tie drop spots to where the center is, rather than to field landmarks, but that would actually result in HUGE holes to the wide side of the field. Really the biggest problem here is that the short side hook zone defender doesn't look for an in-breaking route, so you get huge gaps in the zone that are right in front of the QB (and are thus easy to spot and hit).

The other thing, which is related to pattern reading, but is really just a much simpler version of it, is that they have some basics incorrect. For instance, in cover 2, the outside CBs have to carry a vertical release by #1 about 10 yards further downfield so that they take away the quick fade. Similarly, the OLBs/NBs in hook zones have to wall off a vertical threat by #2 (the slot) and force him to go wide, thus allowing the safety to be able to stay on top without getting beat to the middle.

Also, with flat zones (particularly in Cover 3), defenders give up too many plays behind them to the corner area, particularly on late developing drags that go all the way across the field and wheel routes. They should come up and guard the flat initially, but then sink back to get under those routes as the play develops, rather than just standing and guarding grass.

Full blown pattern reading would be a big development, but there are some major improvements that could happen simply by expanding the playbooks to include a better variety of hashmark defenses (or giving us some kind of defensive play creator) and by adding some fairly basic changes to the way that players in hook and flat zones behave, particularly with regards to vertical routes.

[gschwendt]

I don't think the target thing is workable. Back in the day you could choose an option to pass that way on some of the old PS1 games, and it was borderline impossible, especially on anything down the field.

I'd still like to see it done in today's games. I think back to when the PS2 was first released, a buddy of mine had some FPS game and that was the first time I ever used both sticks to control running & aiming. I thought there was no way that would ever catch on... it just seemed awkward at the time. Now, obviously it's become a staple of all FPSes.

It would take some good tuning but I think it could be done, though I doubt we ever see anything like it anytime soon.