How Atlanta United finally broke through Montreal’s defensive wall

American sports fans love statistics. This we know. If not, what else would we have to talk about on Monday mornings? But statistics can be deceiving, and sometimes don’t tell the true story of what actually happened. To a significant extent, that was the case on Saturday.

Consider: Atlanta United controlled 73.2% of the possession, and in fact controlled possession in every five minute segment of the game. United outshot the Impact 19 to 6, and had a passing accuracy of 88% as compared with Montreal’s 69%. All of that suggests that a blowout was a forgone conclusion, and the final scoreline of 4-1 reinforces that suggestion. But Atlanta did not break through until the 70^th minute of the game, and trailed for 57 minutes. So whose game plan really held the upper hand for that initial 70 minutes?

Well, think back to March 31^st. That was the game against Minnesota United. There are some fascinating parallels. Atlanta took an early lead. But Minnesota outshot United 13 to 5 and had 71.2% of the possession, holding the possession advantage for all but a brief 10-minute stretch in the first half (and that only by a small margin). They also had 87% passing accuracy vs. Atlanta’s 74%. Add to that the red card to Leandro Gonzalez Pirez, and you would think that the final result would have ended up much as Saturday’s game, with an eventual Minnesota blowout. Well, that didn’t happen, of course. But then, Adrian Heath isn’t Tata Martino, and Minnesota’s attack isn’t Atlanta’s.

The big difference between the two games was that Atlanta was forced into a bunkering situation, whereas Montreal came in with that as their primary game plan. And, for the most part, it worked.

It was, in fact, a gutsy strategy. Montreal is not exactly endowed with a wealth of defensive talent, as the 17 goals allowed in the 7 games prior to this weekend indicate, and they were also without their best defender for this game, Victor Cabrera having been sent off against Los Angeles FC the prior week.

Bunkering, parking the bus, whatever you want to call it, requires constant vigilance and discipline. Bobby Warshaw of mlssoccer.com has his own excellent analysis of this game and raises this very point. But doing it for a full 90 minutes is exhausting, especially against an attack as relentless as Atlanta’s. Mistakes are inevitable, especially as the game wears on.

Montreal came out in a 4-5-1 shape, and held it very well. Take a look at this shot from the 22^nd minute:

That’s holding your shape taken to literal extremes. Note that all 20 field players are in this picture, but Montreal is covering 6 Atlanta attackers with 9 defenders. In situations like this, the 3-5-2 that Atlanta played for much of the game is less than optimal. With 3 centerbacks, it was going to prove very difficult to penetrate a double wall like that.

It got worse as play progressed up the field:

In this play, Montreal has 8 players(plus goalkeeper Evan Bush) in the box to cover 4. Note also how tightly the back 4 are positioned; all 4 of them are within the width of the 6-yard box. Again, Atlanta’s 3 center backs are not involved in this play (although they did penetrate deep from time to time), but all 11 Montreal players are in the frame. By the way, the player on the edge of the penalty area at the bottom right is Ignacio Piatti, Montreal’s primary striker.

Another key to this game plan is communication. In a noisy stadium like Mercedes-Benz, that is going to be a challenge, but for the most part Montreal managed to do it. Watch this stretch of play from the 36^th minute, paying careful attention to how the Montreal players are constantly watching and directing each other:

Let’s give credit where credit is due. This was pretty impressive. Now let’s take a look at the heatmap:

Atlanta ran into a wall in the attacking third, in exactly the same way as they have doing to opponents lately. They did manage to get closer, of course, but refer to the pictures above. Try finding the gaps.

Let’s break it down a bit further. Here’s Miguel Almiron’s personal heat map:

And the same for Ezequiel Barco, who moved from secondary striker to attacking mid at half-time:

Not a lot of difference. In fact, both of them spent a good deal of time going sideways in vain hopes of finding that gap.

This is also illustrated in the midfield passing chart. Here are the passes for all of Atlanta’s midfielders:

Note that red arrows indicate unsuccessful passes. Only 5 passes are into the penalty area, and 4 of those were unsuccessful. For the most part, passes were lateral, and a good percentage of those were towards the touchlines.

The result? Well, there’s really one if you are forced to go wide. Crosses. Lots of them:

29 of them, all told. An exercise in futility. I have discussed in previous posts how crosses are a low-percentage scoring opportunity, much as corners are, and with a bunkering defense that percentage drops even further. That being said, Josef Martinez got on the receiving end of a ball from Julian Gressel in the 5^th minute that forced a double save from Bush. Which indicates that Atlanta felt there was a possibility of success from this tactic and that near-miss almost certainly encouraged the team to keep trying it.

Once again, this is eerily similar to what happened in the Minnesota game. The Loons ended up sending 49 crosses into the box in search of a goal only to come up empty.

So, the 3-5-2 didn’t work in the first half, and the 4-3-3 took 25 minutes to produce a PK in the second after Jeff Larentowicz was pulled and Tito Villalba inserted. But in the 74^th minute Tata went all in. He took Greg Garza out and put Kevin Kratz in, at the same time switching to the old reliable 4-2-3-1. Of those 29 crosses, only 1 was attempted after that tactical shift. The first goal was created by a handball in the box from a short-range overhead pass (it doesn’t really count as a cross), and the other three following Kratz’s entry all came from attacks directly up the middle. Going Route 1 has always been the preference for Martino, and in the end he got what he wanted, and it paid off.

Was Kratz the difference? Well, yes and no. Obviously, his free kick abilities were more than useful, and he is a capable midfielder, but certainly no more so than other players on Atlanta’s roster. Rather, it was the combination of Kratz and the formation shift that really made the difference. Also, that shift was made to capitalize on the fact that Montreal’s defense were extremely tired by that point. They were not really up to the task of making the necessary adjustments. Tata sensed it, and pounced for the kill.

This is not to say that there wasn’t an element of luck in Montreal’s strategy. In addition to Josef’s chance, Miggy hit the woodwork at one point and came very close on another shot. After all, Atlanta did rack up 19 shots in this game. But only 7 of those were on target (including 3 set pieces), and 10 of the shots were from outside the penalty area. If a team with a stronger defense than Montreal’s were to attempt this on the Five Stripes, they may not need to be so lucky.

So, is this the end for the 3-5-2? Probably not. There are more than a few strong attacking teams on the schedule and there will be games where defensive strength is going to be needed. But my guess is that for the next couple of games we’ll be seeing the 4-3-3 or 4-2-3-1. That is almost a certainty for this weekend’s game against the Chicago Fire, who have been playing with 5 at the back, including Bastian Schweinsteiger in the middle. After that is Sporting KC, who are best known as a defensive team, albeit somewhat less so this season than last.

Finally, kudos to Montreal for almost-not-quite figuring out Atlanta. Without bench depth, tactical flexibility and savvy managing, the result could have been rather different.