A Critique of the 3 Man Rush

Submitted by Enjoy Life on September 24th, 2009 at 2:01 PM

Let me start this by saying that I HATE the three man rush. Every time it happens there are only two possible sounds coming out of my mouth: “Nooooooo” or “God Dammit!” But aside from my emotional response, I truly do not understand how the three man rush makes any sense in any football situation. My reasoning is based on the following:

1) In just about every football game (college and pro) you hear: “If you give any quarterback that much time, he will pick you apart.” By definition, the three man rush gives the QB more time and violates this axiom. How is the extra time given to the QB offset by having just one additional defender? It just doesn’t make sense.

2) Basic math tells us the three man rush is a bad idea. You reduce pressure on the QB by 25% (1 less rusher out of 4) but increase the defenders by only 14% (1 added defender out of 7).

3) The three man rush is the completely wrong approach to defend the Hail Mary. There is only one way to guarantee the Hail Mary will not work (and, no, it is not “knock it down”) – prevent the QB from having enough time to set up and heave the ball 50 yards! The three man rush almost always results in giving the QB enough time to set up and throw the ball. After the ball is thrown anything can happen as evidenced by the three man rush in the Michigan/Colorado game in 1994. (I was there and have never heard the stadium so quiet and in such disbelief.)

Here is a table showing every three man rush in the Michigan/Notre Dame game.





1.00 1st

3rd &5


Floyd open in end zone, no pressure but pass is overthrown – INC.

4:00 2nd

3rd &10


Screen, 19 yard gain. Penalty ND negates the play.

3:30 2nd



Short pass for 8 yard gain. Tackle by ROH who dropped out of rush.

6:50 3rd

3rd &11


Screen INC.

13:04 4th

3rd &10


13 Yard Pass, First down

12:14 4th

2nd & 15


10 Yard Pass

9:56 4th

1st & 10


21 Yard Pass, TD

Anyone have some good reasons why the 3 man rush (and in the NFL, even the 2 man rush) is used?


Dhani Bowtie

September 24th, 2009 at 2:24 PM ^

Because it gives you the ability to be more aggressive in coverage. You can play man with 2 deep or a 1 deep and a robber. I think you really need to look at the situations where the 3 man rush is being used. If it's being used on 2nd and 4 than there is a problem but if a D coordinator is calling for it on 3rd and 18 it makes a bit more sense.

Ali G Bomaye

September 24th, 2009 at 6:46 PM ^

The counter-argument, of course, is that on third-and-long, if you rush more guys the QB will have to dump it off quickly and won't have time to get it downfield.

I don't think that one strategy is strictly dominant over the other. The key is to mix it up - sometimes you will rush more guys successfully, when the offense has a longer-developing play, and break it up. However, if you do that a couple times, the offense might call a safety-valve or quick-developing play, and if you only rush three you can sit back and tackle the receiver for a short gain.


September 24th, 2009 at 2:24 PM ^

I have the same gut feeling you do whenever I see the 3-man "stand in place" (becuase it rarely is confused with a "rush"). But you see it all the time by every team so it must work...at least to a degree.


September 24th, 2009 at 2:24 PM ^

EXCEPT for hail mary's. The 3 man rush is the one time when I like it because it is less likely that they will score a TD based on less blockers downfield.

This obviously depends on where you are on the field. If he can get it into the end-zone then yes more pressure is better but if he can't then yes the 3 man rush makes sense.


September 24th, 2009 at 2:39 PM ^

everytime i see 3 rushers on a 4th and long, all i think about is the CU v. M game in 1994. I'd never thought about it much before that, but I cringe everytime I see 3 guys rushing.


September 24th, 2009 at 2:44 PM ^

So you're saying that if you're winning a game with 1 second left and the opposing team on the 50, you wouldn't rush 2 or 3?

Some of your points may have merit, but the 3 man rush has some value.

Your data sample is tiny and without context. I would be interested to see data on a larger scale.


September 24th, 2009 at 3:36 PM ^

So you're using 1 example of it failing to say that it should never be used?

It's not 100% effective, but if you're trying to prevent a touchdown and there's only one play, the 3 man rush with a prevent defense makes the most sense.

If it's a hail mary situation there will be at least 4, and probably 5 receivers. A 4 man rush means that there are only 2 more defenders than receivers, and a lot of open space for them to make a play.

Either way, the rushers will be out numbered, and as another poster said, you can scheme to get a good pass rusher singled up.

What would you have done at the end of the ND game?


September 24th, 2009 at 3:52 PM ^

Like I said, you're only looking at 1/2 of the issue. How many hail mary's were knocked down/tackled before the endzone (read failed) because there were 8 or 9 defenders back?

The fact that this list even exists is because the prevent defense is so good at stopping hail mary's that when one works it's dubbed a miracle.

Right now your argument is akin to saying that we shouldn't wear seatbelts because some people die in car accidents while wearing them.

Well how many lives were saved bc of seatbelts? Wouldn't more people die if nobody used them?

Enjoy Life

September 24th, 2009 at 4:12 PM ^

Well, I supplied 10 examples of Hail Mary's. Can you supply 10 examples to support your theory?

Notice in the first video (4 man rush), the receiver gets behind all the defenders. This should never happen at the end of the game no matter how many you rush.

Also notice that on virtually every Hail Mary there are at least 4-5 defensive players around the ball. So, sheer number of defensive players around the ball is not really the solution.


September 24th, 2009 at 5:11 PM ^

failedhailmarrys.com last week, going thru hours of uneventful footage of balls falling harmlessly to the ground or getting intercepted as the seconds ticked off to half time or end of game. My kids love this site!

Went back to grab some vids today to serve up as no sugar coat Data, and they've taken it down for lack of advertiser and user interest. Cue Allanis, wedding days and rain, and all that.

Enjoy Life

September 24th, 2009 at 6:15 PM ^

And, I wish there was was a data base of all 3 man rushes (not just Hail Marys) to analyze the results. But, even Brian does not collect this information. The only way I did the table was to go back and review the entire game on DVR.

I was just pointing out that the idea that the 3 man rush is the ONLY way to defend the Hail Mary does not make sense. Again, if anyone has the data, what % of Hail Marys are successful against a 3 man versus 4 man rush?


September 24th, 2009 at 3:02 PM ^

Sending 3 rushers does not automatically transfer to 5 blocking 3. The way in which those rushers line-up and the pre-snap protection call will determine the matchups after the snap. Watch the Patriots defense in obvious passing downs and 10 or more to go. They do it better than anybody. For example, if the offense lines up 5 wide, QB in shotgun and your DE successfully gets matched up one on one with the tackle, then this is a win for the defense. Quite often in a three man rush you will see offensive lineman free in the middle blocking no one. This creates an advantage for the defense in simple numbers as someone in coverage is free to double team or rob.

OSUMC Wolverine

September 24th, 2009 at 3:08 PM ^

Pressure Pressure Pressure Pressure Pressure

The only defenses that choose to sit back instead of pressure are those who do not have the ability to pressure no matter what they do. I agree with you, the prevent with rushing 3 is frustrating to watch as a fan and a sign of hope if your opponent deploys it.


September 24th, 2009 at 3:10 PM ^

Response to 1) You're using a oft-used quote as an established fact? Weak sauce. And I think it's kinda offset by the fact that there's one more guy out there to confuse the QB in coverage or, you know, simply just covering a guy.

Re: 2) Oh dear. You're using basic arithmetic to try and bolster an already shady argument? You're taking nothing into account except the sheer bodies on the line or in coverage. This is stupid. I would 3-man rush all day if all 3 guys were excellent pass-rushers - which does happen, ya know.

Re: 3) (slams dick into the wall) You're right. Let's just bring the house on the last play of the game with the offense in 4 or 5-wide. Nothing bad could possibly come of that, but a 3-man rush is suddenly heretical? And you're right, once the ball is in the air, anything can happen - and you're reducing your chances of batting it away, blanketing a receiver, coming up with a pick, or making a tackle downfield when you follow your logic.

I don't think the 3-man rush is sliced bread, but I think your reasoning sucks balls from a logical/analytical perspective.


September 24th, 2009 at 3:12 PM ^

That's why you 3 man rush because the guy can create pressure by himself. All joking aside Michigan has been excellent at actually creating a decent amount of pressure with the 3 man line but the underneath coverage of our back 7 is so atrocious it hasn't been effective. Brian pointed out a couple times laast year where teams had 3 or 4 guys in the pattern and they were still wide open. 3 against 8 should = throw away but many times last year it equaled 1st down. With Graham, Roh and Martin we should have pretty good success with a 3 man rush, but it has been terrible ineffective.


September 24th, 2009 at 3:18 PM ^

Because I'm sure that is exactly the reasoning GERG deploys the 3-man rush on occasion: Brandon Graham. If you have a world-class pass-rusher obliterating his side of the line on every play, maybe he sees an opportunity to drop Roh or Mouton into coverage while he suspects Graham creates enough pressure on his own? Who knows. But a great, great rusher like Graham makes a world of a difference.


September 24th, 2009 at 3:21 PM ^

Like many, I tend to agree, sort of, except where it comes to Hail Marys. You damn well had better be only rushing three on the Hail Marys. Might wanna rush two. It's simple numbers. You simply want to outnumber the offense, who can legally send no more than five to catch the ball. Rush three and you send eight to knock it over. Eight on five is better than seven on five is better than six on five....period. Let the quarterback have all day to throw - what's he going to do, run?


September 24th, 2009 at 5:28 PM ^

Doesn't seem to validate your theory. It is a best College Hail mary because it worked against overwhelming odds. Also does't compute because in order to be a Hail Mary you have to get enough time to throw. So, for example, if there was a 4 or 5 man rush at the end of the game, the offense runs a slant (not hail Mary) and takes it all the way for a touch down. This wouldn't make the list of greatest hail Mary's but rather the list of bad coaching decisions.

The other person was right, you need to compare 3rd and 4th down hail mary type situations and calculate the percentage of time that the hail Mary works when running it against a 2/3 man rush versus 4/5 man rush.

If you knew which had a higher success rate of winning games then you would have a better analysis.


September 24th, 2009 at 6:08 PM ^

When virtually every Hail Mary attempt is against a 3-man rush, it follows that most successful attempts are against 3-man rushes as well, no? For instance: I HAVE NEVER SEEN A SUCCESSFUL HAIL MARY AGAINST A TWELVE MAN RUSH. Why does nobody do a 12-man rush?


September 24th, 2009 at 3:59 PM ^

I agree, the three man rush is sort of like no-mans land on a tennis court. It lies between rushing 4 or more, or only rushing two...yes...I said two. Teams should do THAT more often. Example? 1995 game between pass-o-matic, high TD Machine Florida State and Virginia. Virginia ran a two man rush most of the night, dropping NINE...completely flusted FSU spearchucker! In fact, when watching the rosebowl vs. the longhorns, as Texas embarked on the game winning drive, I was screaming at Herrman to "rush two and drop nine....make him throw!" But my voice did not carry through the Tv screen.


September 24th, 2009 at 4:20 PM ^

Of the 7 pass attempts against a 3-man rush that you cite, all but 2 were short of a 1st down, (would have been 3 but for the penalty against ND in your second example). Still, 3-for-7 (I'm throwing in the one called back) on long yardage situations isn't too bad considering we were playing against some world-beating receivers.

That doesn't explain putting a 3-man front in on 1st down, as in your last example. I agree with you there.


September 25th, 2009 at 11:20 AM ^

+1 to you; an excellent point. One, I think, that you can make even stronger by not discounting the penalty. Although the penalty we accepted was an illegal shift, there was also a downfield hold; one maybe slightly more likely to happen by having more defensive players downfield.

Additionally, if there's ever a play against which additional rushers might be a bad idea, it's the screen. There are rare cases where the rushers get to the QB so quickly that the screen is broken up from that end but the ball will get to the reciever the majority of the time, no matter how many rushers your bring, so it's a matter of who's back on D to break up the play.


September 24th, 2009 at 4:53 PM ^

Point #3 is majorly flawed. Even if you rush the passer and prevent the QB from sitting back there and heaving it 50+ yards, he doesn't NEED to throw it into the end zone. That is, if you rush 5/6 and leave only 6/5 back to cover 5, you're unnecessarily creating the opportunity for a deep cross down the field with decent numbers by the WRs. Why do that?

In other words - what's more dangerous to you - 8 defenders in the end zone trying to bat a ball down with maybe 5 (more likely 4) WRs hoping for a lucky bounce, or 5 defenders trying to make an open field tackle on a WR, when the defenders probably have already gone into a crazy deep drop and have to close a gap of 20-30 yards while avoiding WR blocks?

Edit: I realize that we, as UM fans, are a bit scarred by the Hail Mary. But as amazing as Kordell's throw was, that play is one of many ignored failed Hail Marys if the ball is batted down and not batted up in an attempt to pick it. Hail Marys are rarely successful because of how hard it is for 4-5 WRs to beat out 7-8 defenders for a jump ball in the end zone in that sort of traffic...

Enjoy Life

September 24th, 2009 at 6:22 PM ^

A) I never said to rush more than 4. I agree, that does not make sense. Rushing 4 should provide enough pressure to make up for the one less defensive back.

B) The famous AC catch/run against Indiana was against a 3 man rush and 3 (yup, 3) defensive backs missed AC as he ran the final 20 yards to the end zone.


September 24th, 2009 at 6:25 PM ^

#1 I don't buy into what the talking heads on tv say, and I'm surprised any reader of this blog would. Sometimes a QB with a lot of time finds an open receiver, sometimes he does not. I can make the same, totally unsupported argument but in the opposite direction - if you drop 8 guys back, the QB will have nowhere to throw.

#2 This is terribly flawed logic. As someone already pointed out, by your logic, a 6:5 ratio of rushers to coverage guys (or a 5:6 ratio thanks to the odd numbers) is always optimal. Besides, you're assuming the marginal benefit of the next pass defender is less than the marginal benefit of the extra pass rusher. Why?

#3 The 1994 game sucked, but that doesn't prove you're correct. Neither do 10 random youtube videos. Think I'm wrong? Apply some stats to those 10 observations and find out.

Elno Lewis

September 24th, 2009 at 7:09 PM ^

that episoded of The Simpsons where Lisa sells Homer that rock that will keep tigers away.

Lisa used the word 'specious' in her exchange with Homer.


learn it.

Blue boy johnson

September 25th, 2009 at 12:17 AM ^

Enjoy Life may be on to something. I found this footage of the 1976 Orange Bowl. 9:32 into the video Leach attempts a hail mary pass against a 4 man rush and it is intercepted.

If you watch the entire video you will see Gordie Bell score the lone M touchdown. Bo also called for a Halfback pass and Gordie threw an INT into the endzone, but that's ok.


September 25th, 2009 at 5:08 AM ^

I usually don't like 3 man rushes either, but based on your 7 instances above, it looks like 'paper' to screen plays 'rock' (paper wins that one in case you don't rochambeau).

From now on, whenever I see us in a 3 man front, I'm going to be hollering "Come on screen! Call a screen!"

Greg McMurtry

September 25th, 2009 at 8:49 AM ^

the 3-man rush because I don't think it puts enough pressure on the QB. I also think that having 8 DBs instead of 7 only provides marginally better field coverage. In the end, if it is a hail mary play, the main concern is getting position to where the ball is being thrown. I don't think it helps having more guys in the cluster of players standing in the end zone. I want my tallest guy out there with the best deep ball skills to bat it down. That's why you'll see Randy Moss out there in "jump-ball" hail mary situations because that's what he's great at. I would go rush 4 and have the players I described cover the end zone, with lesser quality DBs covering short.



September 25th, 2009 at 11:06 AM ^

Which is somewhat valid as long as the extra defender is one of the guys in the cluster. It's still better that he's there but not by a huge margin. I'd argue that Michigan-Colorado is actually a good example of why this is still better, though--if a ball gets batted it's not going to be the guy with the best position who makes the play, it's going to be a random guy in the cluster. The more of your guys in the cluster, the more likely your team makes the play.

But you're missing a hugely important case--what if that extra defender is not an extra guy in the cluster: what if he's the one who's able to get the best position? If he's not there then there's a HUGE difference in the chances that the hail mary will succeed. It may be more likely that the Randy Moss-esque guy is the one in the best position but I highly doubt it's always the tallest guy who has the best position, particularly in college football.

name redacted

September 25th, 2009 at 2:00 PM ^

I am no fan of 3 man rushes, but in a hail mary situation they are a must. Your argument FAILs miserably on your first bullet, and goes down hill after that.

A QB will pick you apart if you give them time. Agreed. A hail mary is not a "pick you apart" situation. It is a throw it as far as u can and pray (hence the name). You are thinking much more of say a final drive, where a team drops back into a prevent and the QB goes down the field with 10-15 yd passes. That, I agree with you on. I hate the prevent, and for some reason Michigan's historically bad prevent. I believe in pressure during those circumstances. But we are talking massive differences between that and a hail mary.

Your math on #2 is beyond flawed. Others have ripped it apart I prolly shouldn't. Just matching up a man vs a man and determining that x more man means x% more effectiveness is nuts. Against a 5-6 man front, how more often does a 4 man rush get at the QB than a 3 man? That's more of a metric. But still not complete. Next, I would say a hail mary doesn't take a lot of time. There's not a play developing, there's no break down, no need to create, no need to read defenses..its 5-7 step drop, 2 steps up throw it as far as you can. Every time. Same over and over and over. Its a high throw, your not going to bat it down. Literally, you have to sack or apply enough pressure to make the qb do something other than drop back, step up and throw. Not happening. I will take 2 or 4 more hands in the end zone. A hail mary is one of the most predictable, shoot I would say THE most, predictable throws as far as how much time it takes to get off. Coaches aren't dumb, some, but not every single one of them. If they thought by adding 1 more guy on the line they could get at said QB in the amount of time it takes, they would. They clearly don't think its even close, or more would try. Whats the % of a blocked punt? Yet defenses try that any number of times a game. Its dumb, it just takes one more person away from the secondary.

You also have to think about which hands are being taken away. Usually, its a taller receiver or special teams player, who can physically match up against the offenses receivers better. The normal secondary is back there cause they have the experience in the position, and the extras are the taller good hands guys. If you simply do a normal front, and drop your normal secondary, its a bunch of 5'10'' DBs going for a jump ball against 6'2'' receivers.

You are kind of using the logic that you see the sun come up in the east, and set in the west. It must revolve around the earth! Without taking other factors into play or thinking your argument through. Eh, good attempt.