A Look at the Bubble Screen

I don’t know how other fans take it, but a lot of USC fans seem to hate the bubble screen.  Personally, I am a huge fan of the play as it has been an integral part of USC’s offense since at least the early Pete Carroll days.  It will continue to be a staple of the offense for the foreseeable future as well.  Let us look at the bubble screen from a strategy perspective and also look at some stats from last season.

What is a bubble screen?

First off, a bubble screen is a play that the quarterback takes the hike, pivots, and quickly gets the ball outside to the wide receiver.  The receiver is waiting at or slightly behind the line of scrimmage when the ball is caught.  There are usually other receivers blocking ahead, but not always.

A slight variation to the bubble screen is when offensive linemen release their blocks to get ahead of the receiver.  In this case, the play has become a jailbreak screen.

Why use a bubble screen?

When utilized correctly, the bubble screen complements the rest of the offensive playbook and can lead to big plays.  Offensive play calling is all about complementary plays, and it is important to find how a play fits in to the overall scheme.  A bubble screen is used for a number of reasons, which we will go over here.

First, the bubble screen is almost guaranteed positive yardage.  This is a high completion pass and should be caught at least 80-90% of the time.  For this reason, the bubble screen should very much be viewed as a running play.  Let me reiterate that as I believe this is a key concept that must be understood: a bubble screen is a glorified running play and should be treated as such.  Rarely does a bubble screen result in huge yardage gains as that is typically not the purpose of calling a bubble screen.  Similar to running plays, bubble screens are meant to gain positive yardage at a consistent pace and slowly chip away at the defense.

Second, a bubble screen takes advantage of the athletic abilities and potential talent gap of your skill players against opposing defensive backs.  The nature of the bubble screen gets the ball directly into your skill player’s hands in space.  Defenders must make open field tackles which are much harder than gang tackling you see in the middle of the field.  Most bubble screens are designed such that the player has lead blockers and one defender to beat.  When you have an athletic player such as Robert Woods or Marquise Lee, the advantage usually falls towards the offense.  Both these players are excellent candidates for the bubble screen as both of them are selfless blockers as well, so they may play the roll of the catcher or the lead blocker.

Third, the bubble screen reduces pass rush on a micro and macro scale.  The micro scale is fairly simple as the ball is released before the pass rush can effectively get to the quarterback.  However, the macro scale is not always obvious and was pointed out by Oklahoma State head coach Mike Gundy at the Nike Coach of the Year clinic hosted by USC.  When a bubble screen is called, it requires little effort and exertion by the offensive line as they must only hold their blocks briefly.  However, the defensive line must continue pursuit beyond the point of the ball being thrown, as they must try their best to catch the receiver from behind.  Because of this, Mike Gundy uses the bubble screen early and often as a way to tire out the defensive line.  What he observes is that the defensive line is simply going through he motions and no longer actively rushing the passer in the 4th quarter, even on passing downs.

Fourth, the bubble screen sets up for bigger plays.  With an athletic ability advantage, opposing defenses will often play with a large cushion on the wide receiver to prevent getting burned deep.  A bubble screen will chip away at opposing defenses, gaining easy first downs or creating short yardage situations.  To stop the slow bleed of yardage, opposing defenses must eventually respond by closing the gap which ultimately opens up big passing plays.  On the flip side, if opposing defenses are too aggressive on their pass rushes, bubble screens can be used to punish the defense by getting the ball out before the rush/blitz can get there.  Have enough success in this and the defense will have to slow their pass rush as they look for the screen.  This will give the quarterback more time for deeper throws.

Lastly, the bubble screen is very versatile for play calling.  Many teams are known for their use of option plays such as the zone read.  It is not as well known that USC uses read plays as well, some of which that involve the bubble screen.  In these plays, Barkley will read the defense either pre-snap or shortly thereafter and either continue the play that was called or simply throw the bubble.  One example would be a running play has the bubble screen option on a pre-snap read.  Barkley will observe the coverage, most likely judging the amount of cushion on his wide receivers or the alignment of the linebackers, then decide to either hand the ball off or abort the run to throw the bubble.  Whatever choice that Barkley makes, the offensive linemen and running backs will perform their task as if it was a running play.  If you have a chance, go back to the 2011 USC vs UCLA game to see this concept in action.  Kiffin has noted that he wanted to maintain a pass/run offensive balance, but the Bruin secondary continually gave Marquise Lee and Robert Woods large cushions.  In response, not only did Kiffin call more bubble screens, but Barkley continued to abort running plays for the bubble.  In large part, this is why USC passed the ball on 69% of their plays against UCLA when their season average is 53%.

The bubble screen in 2011

Now let us look at some statistics on the bubble screen for the 2011 season.  I have tracked the use of the bubble screen in my personal notes and have the numbers for ten of our games last season (isn’t all of our games, but it should be enough to paint a general picture of what happened).  In those ten games, I tracked 43 bubble screen attempts (approximately 6% of USC’s plays).   Six of them were incomplete making an 86% completion rate.  In total, these 43 bubble screens gained 263 yards.  This comes to an average gain of 6.11 yards per bubble screen attempt (and 7.11 yards per completion).  Since the bubble screen should be thought as a running play, let us compare these numbers against USC’s running averages.  As a team, USC averaged 5.0 yards per carry with Curtis McNeal gaining 6.9 yards per carry and Marc Tyler at 4.7 yards per carry.  The bubble screen’s numbers are comparable to our running game.

Out of those 43 bubble screens that I tracked, only two lost yardage.  This was a loss of one yard against Minnesota and a loss of four against UCLA (Marquise Lee caught the bubble screen on his knees).  In total, there were 15 plays which gained three or fewer yards (one loss of four yards, one loss of one yard, six incompletes, one gain of one yard, two gains of two yards, and four gains of three yards).  Comparatively, there were nine plays which gained at least 10 yards (one gain of 11 yards, two gains of 12 yards, and one gain each of 13, 14, 15, 16, 17, and 19 yards).  This is also comparable to a typical running play’s histogram.


I love the bubble screen.  I like it because it is part of a larger system and when utilized correctly, it can do wonders for the offense.  You won’t see many huge plays come out of it, but it will be used to set up plenty of touchdowns. One reason I think the bubble screen gets a bad rep is because people expect it to be the same as any other passing play.  However, the fact of the matter is that the bubble screen will never gain as many yards per attempt or yards per completion as a typical pass.  It should instead be viewed as a running play.  Like any other running play, you can expect to see a decent amount of short gains.  These gains should be consistently positive and grind down on the defense.  The point is to get closer to those first down markers all while tiring out the defense.  It is in this manner that I hope that fans will learn to appreciate the bubble screen as a part of a larger offensive system.


3 responses to “A Look at the Bubble Screen

  1. If bubble screens tire out defensive lines, why did USC perform less well in the 4th quarter than in other quarters?

    In addition, USC was considered a poor red zone team. I assume that bubble screens work less well in the red zone, because the DBs don’t have to give the WRs such a large cushion. Is it possible that intensive use of bubble screens in lieu of running plays prevents the OL from fully developing its run blocking capabilities and rhythm, leading to poor red zone run blocking? Therefore, when bubble screens become ineffective in the red zone, USC has no viable high-percentage offensive plays to run.

    Alternatively, is it possible that running plays wear down a defensive line more than do bubble screens, so that success in the red zone and in the 4th Q would be greater if a power running attack was used a lot, averaging 4.5 yards per carry, rather than using bubble screens that typically gain 6.5 yards per play? The offense has to run more plays, which causes greater defensive fatigue.

    Furthermore, traditional power running consumes more clock time, because the clock is never stopped. That provides indirect assistance to defenses.

    It’s all tradeoffs. Is the bubble screen a great tradeoff? Maybe, but I would have to see more statistics than have been provided, addressing some of the above-stated concerns.

    • Thanks for the feedback and questions, Ben.

      The lower offensive output in the 4th quarter could be due to a number of factors. First, we may not have done the bubble screen enough to see the benefits in the 4th quarter. Using my numbers, 43 bubble screens over 10 games is only 4.3 per game. While I don’t have hard numbers, this probably isn’t to same level as Mike Gundy uses it at Oklahoma State or Mike Leech uses it in his air raid offense. Secondly, while it does tire out the defensive line, our players may have just tired out at a quicker pace due to low roster numbers leading to less substitutions. And finally, the last reason I can think of off the top of my head is that lower offensive execution in the 4th quarter could be attributed to any number of areas on the team, such as the defense getting tired out leading to less time that our offense has the ball.

      Like you said, bubble screens lose their effectiveness in the redzone because the corners can play tighter and also it will bring the safeties up as well. There is a point where you are so close to the endzone that the corners get completely isolated from safety help in which the bubble screen could be used as a touchdown play, but that is another issue altogether. Anyways, it is possible, as you suggested, that relying on the bubble screen can lead to lesser abilities of the offensive line in run blocking. I’m not sure how much this would affect USC itself as we averaged 4.3 bubble screens per game, so that is taking away 4 potential run plays. Another contributing factor to consider is that USC had no power runner, or “big back.” This makes it harder to get those red zone yards as the safeties play much closer to the line.

      I completely agree with you that a solid power run game will wear down the defense more so than bubble screens. Having your offensive linemen driving the opposing defense backwards against their will definitely tire them out faster than making them run around to the perimeter essentially unblocked. I am a big fan of the power run, but not sure our offensive personnel is geared well enough to run this type of play all game long like what Stanford uses. USC still definitely uses power runs, just not to the same level. I hope to someday do a write-up analyzing USC’s use of the power run game.

      Traditional running does use a bit more clock, but given the high completion rate of the bubble screen, it shouldn’t be too big of a gap. The other consideration is how often the receivers went out of bounds or were forced out by defenders.

      Unfortunately, it is hard to pinpoint the contributing factors to these issues without having the necessary data on hand. Without this data, like you said, it is hard to know the tradeoffs. However, these are some things that I could think of. I would love to keep this discussion going.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s