Diving Deep: Birmingham Legion 3-2 FC Tulsa

So the Legion has now come back from a losing position to win a game 3-2 3 times in its history. This was the first time at home as well as the first since 2019. Those who follow us on Twitter already know which games the other two were: the first ever win in team history at Louisville City and the first ever playoff win at North Carolina FC. All were dramatic in their own ways. The Twitter poll we ran ended up a tie between the NCFC and Tulsa games as the best, which really resolved little.

Personally, I would say that the Tulsa game was probably the better of the two, given that both the Legion and Tulsa are better teams than the two in that playoff game three and a bit years ago. For sure, it was a much more complex game tactically. And that is true of Tulsa as well as the Legion, as it turns out.

We’ve spoken before about the formational flexibility that the Legion has over the past year or so begun to employ, and this was definitely the case in this game. The team was allegedly in a 4-2-3-1 formation. Tulsa theoretically played the same announced formation. For parts of the game, both teams did precisely that, but for large parts of it, they did not.

Moreover, it was a very spread out game. Here’s the heatmap:

The Legion are playing from left to right in this graphic, not that it’s very important, as you could rotate it 180° and not notice much of a difference. Two things stand out, though. First, the game is very widely distributed over the field. That’s a complete turnabout from week 1 against Pittsburgh, where the game got stuck in midfield. Second, both teams favored the left flank for the attack. As far as The Legion is concerned, that continues on from the previous game, but is a change from last season where everything ran down the right, thanks in large part to Jonny Dean.

What the map doesn’t show is that the two halves of the game were not significantly different. Which is interesting in itself: since both sides changed formations a fair amount, does that mean they were rigidly flexible?

Well, perhaps not rigid, but certainly organized. Let’s investigate that a bit. Here are the average positions for both starting elevens:

Well, clearly that doesn’t easily break down into a 4-2-3-1 on either side. As noted above, it’s also fairly spread out. But what is it? Well, I think you can see potentially two other formations here. On the Legion side, if you count Gabriel Alves (#16) as left fullback, then it’s a 4-3-3. Or, more specifically, a 4-3-2-1, the so-called Christmas Tree formation, with Diba Nwegbo (#7) up top. But Alves played pretty high for the 80 minutes he was in the game. He booted in 4 crosses but had only defensive action. He was therefore more of a wingback than a fullback, which means the formation is more a like a 3-4-3.

Again, with Diba as a solo striker, it’s a 3-4-2-1. Now this is important. We have talked before that the Legion is closing in on a Total Football concept. This is evidence of that. Total Football is the brainchild of Dutch legend Johan Cruyff and he also developed the formation as a means to exploit it. I also have to note that it has been more recently used by Arsenal under Arsène Wenger, so obviously it has its drawbacks. Tulsa, on the other hand, look like they in a 3-3-3-1 formation. That’s a completely whacko setup, used by almost nobody except, you may not be surprised to hear, El Loco himself, Marcelo Bielsa.

 A couple of things to note about the 3-4-2-1. First, a team can slide into it fairly easily from a 4-2-3-1. Notably, both lineups are based around a double pivot. In this case, that would be Anderson Asiedu (#6) on the left, and in a somewhat unlikely choice, Prosper Kasim (#10) on the right. Alves moves up from left back, and Collin Smith (#4) hangs back from right attacking mid to right wingback (a more natural position for him anyway). Jake Rufe (#13) also leaves off attacking to become the third centerback. Second, the formation uses the rather out-of-fashion “libero” (also known as a sweeper). Here that would be Alex Crognale (#21). Another slightly odd choice, given Alex’s hobby of collecting yellow cards.

So what are the benefits of this formation? Well, it’s good for counterattacking teams who are okay with giving up possession. That certainly happened last Friday, with the Legion having just 38% possession overall and a meager 31% in the first half. You can do that because of one of the other strengths – namely defensive ruggedness.  With the libero covered by two other centerbacks and at least one half of the double pivot, it’s very hard to penetrate. Of Tulsa’s 10 shots in the game, 5 were from distance, including both their goals, and both also had to thread the needle between a pile of defenders. Great goals, by the way.

Also, the two “inside” forwards – Juan Agudelo (#9) and Tyler Pasher (#15) are given a lot of freedom to roam, especially laterally. This is designed to create defensive confusion and thereby create scoring opportunities. Juan didn’t drift much from his left side position, where he really isn’t comfortable. Tyler did. Big time. Here’s his touchmap:

It’s hardly a shock then that his two goals came from opposite sides of the field.

In this formation there is no true central attacking mid. And until Enzo Martinez entered the fray 58 minutes in, the Three Sparks didn’t have one on the field, so from that point of view it was a good option. It’s also good against teams that have a strong midfield. That’s probably Tulsa’s best part, with Rodrigo da Costa and Marcus Epps creating mayhem. And let’s not forget they were effectively operating with a 6-man midfield.

But it didn’t work perfectly. That double pivot is absolutely key to its success. Which is why the positioning of Prosper is a real headscratcher. OK, putting him on the right where his left leg can produce nasty inswingers makes sense, and we have seen him on that flank plenty of times. But he is an intrinsically attacking player. Frankly, he looked out of sorts. He was pulled after 58 minutes and only 18 passes. The average positions above indicate he was basically sharing space with Collin. Ando, in contrast, was completely in his element and thrived, having a great game.

With Enzo, Neco Brett and Ben Reveno on the field earlyish in the second half, things naturally changed. Neco at striker played even higher than Diba. Enzo of course was the CAM option and stayed central the rest of the way. Reveno was a more logical defensive mid/pivot. Brett and Reveno were involved in both late goals (although I question crediting Neco with an assist on Tyler’s goal, which was largely a solo effort). So was that a change back to a 4-2-3-1? With Gabriel remaining high, not entirely. Call it a 3-2-4-1.

That’s not to say that the 4-2-3-1 was not evident at any point in the proceedings, because it was. The 4-1-4-1 popped up from time to time in the game as well. The 4-man back line was especially obvious when Moses Mensah replaced Gabriel with 10 minutes left in regulation. It was, all told, a rather confusing evening as much as it was entertaining.

Of course, you can only be flexible if you have flexible players. The Legion likely has more of those than is typical in the USL Championship, and that is a big deal. It also includes Mikey Lopez, who has not featured yet. look for more of this as the season progresses.

One last point that needs to be highlighted: All three Legion goals came from wide play, either moving out wide from a central position or moving in from a wide position. The team did not try to run it up the middle. At all. A complete deviation from much of the teams’ history. You can see that in the heatmap, where there is a lack of red in zone 14 (the area just outside the opponent’s penalty area). It’s also a bit odd, as from the average position chart, you can plainly see that Tulsa left a yawning gap in the middle. But if the Legion can continue to draw defenders wide as they patently did here while at the same time having serious central attacking options, that will make the team very dangerous indeed.