Pattern reading is one of the most crippling tactics a defense can employ against an unprepared or poorly organized pass offense. Even a successful passing team can suddenly find themselves unable to get anyone open. Think about what it is like trying to throw against your own defense when they all know your plays.
I was actually kidding: if your own defense can always identify what you're doing, then that is probably a sign that you need to rethink things. The issue is that as often as possible the defense should not know for sure where the receivers will wind up until the ball is actually thrown. This is done by making routes look alike.
The first thing is that your eligible receivers must learn how to explode off the ball and make every route look like they are going deep. If a receiver can explode and make the DB think he is going deep, and then has the skills to stop and change direction in two (or sometimes three) steps, then he can always get open versus one defender. Imagine a pass play with four (or five) vertically releasing receivers: each could go deep, stop, or go in or out at any time. Further, no single receiver can afford to be double teamed until the intentions of some of the other receivers have been given away, at which time the ball should already be thrown.
In fact, most 3-step drop passing games and many timing routes look just like this. Below are some diagrams of possible combinations and routes put together (some put together quite casually, not all are vouched for as great plays).
Drawing the route tree can be thought of as showing what it is like for a defensive back: he is backpedaling and he sees all the potential ways that receiver could go. However, if he can pattern read, even if he doesn't know with 100% probability what a receiver is going to do, if he can narrow it down to one or two things, then the offense is on a slippery slope downhill.
So, if defensive backs have a hard time pattern reading vertical releases, then what can they pattern read? Well, they read routes that immediately show where the receiver is going, as shown below:
I.E. Shallow crosses and flat and shoot routes. Or in other words: the bread and butter of many, many football teams' passing games.
Homer Smith wrote a few articles that very convincingly decimated the usefulness of some of these routes, particularly the flat route. You can threaten no more than one defender, who can always take it away. In the world of football strategy, this is not how you win football games. You win football games by isolating your players in one on one matchups that they can win and score against and occupying two defenders with one receiver, thereby creating those one on one matchups. Defenses will trade one for one every single time. Think about what bunching or stacking receivers is intended to do, or the multiple threats that tight ends and H-backs can present.
Moreover, these routes a) are easily covered or "squeezed" by the LBs, b) they usually signal exactly what the vertical releasing WRs are going to do (i.e. if the slot flies to the flat, then the corner knows he will only run a go or a curl, but will not run an out route), and c) does not threaten the deep coverage so the safeties can help double team the downfield receivers.
I used to be a bit more sceptical, but I think it is safe to say that the NFL, from a strategy standpoint, is obviously the most sophisticated football being played at any level. (Though certain strategies, such as many option football strategies, are deemed too risky because of injury to the quarterbacks.) There is simply so much more time, money, and experience at every level of the teams--from players to coaches to technical assistants, and there is too much incentive and reward for success for it not to be.
Thus, I think it would be a fair test to say that if these routes are as useless as they seem to be on paper under careful analysis, then someone would have realized this, whether explicitly removing them on purpose, or implicitly they would get phased out when they analyze piles of play results and use the most successful rotues and packages.
A quick scan of any NFL game will show you that lots and lots of shallow crosses and flat routes are being called, completed, and used successfully. While running the risk of asking the obvious (or simply making the simple complicated), but why?
The first, and probably most important reason, is that these routes are simply shorter and easier to complete than other quarterbacks. Most are less than 6 yards, compared with the vertical stem routes, which are 10-15 yards (at least) downfield, and, knowing that we can use the Pythagorean theorem to determine how far the pass actually needs to go, this could be a 15-25 yard difference between a 5 yard route and a 12 yard route. This is important both because of simple success rates, but also because football teams and quarterbacks are human, psychological beings, and I am a big believer in getting QBs established early with easy throws to get them comfortable.
Second, they do focus on the LBs and the undercoverage, who are often weaker defenders. This is a less strong reason because routes with vertical stems can and still do attack these underneath defenders, but there is no mistaking it with the shorter routes.
Lastly, while you do immediately lose deep threats from the route, you can still create new route trees off these pass releases, creating new uncertainty for defenders, which is what I will discuss in my next article.