June 10: Chicago 2 – 0 Atlanta

Dig the Long Ball

Coming into this match, Chicago were playing 15% of their passes long, which put them in the top third of the league in terms of short passing (Crew & Sounders ‘lead’ at 12%). Against Atlanta, the Fire played 25% of their passes long at home, and often successfully over the top of Atlanta’s isolated centre back(s). For reference, the highest long pass average in the league is Vancouver at 20% who coincidentally also upped the long balls above their season average against Atlanta (and won at home). Before that, New York City stubbornly did not increase their long ball usage, and were wiped out at Bobby Dodd Stadium. But before that, NYCFC played longer in Yankee Stadium (home) against Atlanta and succeeded. Between those two matches, Portland played significantly more long balls than their season average and managed a point (with a potentially missed offside call)¬†at home.

See what I’m getting at? Like I said last week, with Atlanta coming to town, teams have a choice to make around how they progress the ball into their opponent’s half, and the correct choice seems clear. Playing short means Atlanta will create good chances off turnovers. Playing long means you minimize that problem plus you may even create something nice yourself.

I don’t want to be too presumptuous with the causality. The guy who watches one team and is bad with data is the king of small sample sizes. The above matches could be a coincidence. Or, higher percentages of long balls might be a symptom of game state, that is to say teams that are leading hit the ball long more and more as the match progresses (although there are some pretty good teams at the top of the short passes table). I’d love to see a broader analysis of which teams drive other teams to deviate away from their general trends. Is it a complement to Atlanta that each team they face is shifting it’s style of play significantly? Or is it evidence of a glaring weakness that’s easily exploited with just a few days of game film preparation? Anyhow, I’d love to see a statistical approach to examining which teams drive change, and which ones their opponents historically may just play straight based on style preferences and goals.

I should say also, as we segue back into the match, that Chicago were more clever and complex than simply playing route one against Atlanta. They mixed in their fair share of nice build-up, pressing, long countering balls, and just general balance in a way that seems to have optimized the result once the game state was on their side. Here’s are @11tegen11‘s pass maps for the match. Chicago’s is interesting.

11tegen11CHI
Chicago with lots of conservative passing and a few repeated long paths towards goal. For Atlanta, this looks like the usual very aggressive positioning, but you’d love to see some links to Tito up top. For what it’s worth, there were countless times where he made great runs to split the centre backs but was rarely rewarded with a throughball.

Own-half Possession & Pressing

Atlanta came out pressing high (surprise, surprise). Chicago picked their spots to press — waited for vulnerable moments, like if an Atlanta defender turned back towards his own goal with the ball. Chicago were successful in turning Atlanta over at times in their own half at the beginning of the game, but dropped off some late as they played with the lead.PressureGame

Atlanta, for their part created right around the exact number of turnovers they’ve forced of their opponents on average this season, both on a pass per giveaway basis (4+), and also in total (32). Here’s the full table (w/ 11tegen11 xG subbed in for the latest match instead of ASA’s numbers):PressureTable

In one such play, Garza slid in and won the ball back and laid off to Tito who shot wide, one of the better chances of the match for United. Chicago’s chances on the other hand came less from pressure in Atlanta’s half, and more from opportunistically exploiting the aggressive Atlanta positioning. With both fullbacks high, and at times a bombing Michael Parkhurst flying up the middle, Chicago played balls over the top several times, mostly catching Atlanta out.

Goal Focus ūüė¶

For the goal that changed the game, in a comedic turn of events LGP and Parkhurst are passing each other like strangers on a cross walk when we join live, LGP having gone on his tour of the attacking half, and Parkhurst beginning his ill-fated voyage with the ball.

CHI GOAL AWhile this was ¬†fun for a second (like the last time Parkhurst did this), it was a poor choice. You can see the general structure of the team completely broken with Parkhurst dribbling into trouble even with 5 or so Atlanta players open for a pass at this point. In the mean time, Pirez, now making his way back to position, has no centre back partner but is tasked with containing Chicago’s 2 most dangerous attackers.CHI GOAL BOnce Parkhurst is dispossessed, it’s kind of over.CHI GOAL D

Chances

I should mention that while Chicago did have the better of the chances in the first half, they were by no means dominant. @11tegen11 shows Chicago with a modest lead at the half, almost all of Chicago’s total (something like 0.8?) coming from the chance they scored. Tito’s 1v1 chance off the odd Gressel sideline clearance/lob makes up the bulk of Atlanta’s total at half (which looks like something slightly higher than 0.5 goals). That number in and of itself should tell you something about how harsh you should be towards Tito for the miss. On average that chance is converted a little less than half of the time (a fantastic chance, as good as you’ll find really, but not one that’s converted 100% of the time – it’s just the nature of the sport). Anyhow, here are the chance maps for the two teams at the half:

To the eye, it’s an intriguing half of football. Atlanta controlled the game and had the volume of chances — many of which don’t show up as shots in the stats because they were narrowly snuffed out. There are plenty of crosses from the right with Gressel/Mears, and more lively action from the left with Garza/Asad. There’s a great ball played in from Kratz off a long set piece before the half that doesn’t connect, and a Pirez header from the spot. What I’m getting at is that while the goal changed the game, Atlanta’s attack didn’t look inept to me, by any means, despite Almiron and Martinez not playing. Chicago had many such almost shot type chances as well, narrowly cleared by defenders.

In the second half, Chicago managed the game state very well. They are a good team I’m afraid to say. In general, they never lost their balance between attack and defense:

Lots of crosses for Atlanta in the second as Chicago appropriately managed the lead and extended the lead off of a corner kick originating from a great piece of build up play where Kann made an outstanding save. It’s hard to break a team down who’s leading, smart, ¬†and dangerous on the counter. Atlanta’s learning that the hard way this year, but not really applying it to situations when they have the lead — they’d rather maintain the aggression and take risks.

The continued question of where the goals will come from if not through transition

I’ll continue this discussion from last week. There’s a growing concern that if in any given game, Atlanta aren’t converting transition chances, or aren’t getting said chances because a team is bunkering, or a team is “managing the game well” with a lead, then Atlanta won’t be able to generate goals through possession and buildup. And in response, my basic theory is that very few teams can count¬†on calm buildup type goals that aren’t transitions or set pieces (another story altogether with ATL) — certainly no team in MLS. I have a thought experiment. Think of the last five or so soccer matches you’ve watched. For me it’s CHI-ATL (transition goals + set pieces), MEX-USA (transition goals), USA-T&T (transition goals), VAN-ATL (set pieces, 1 ATL buildup goal), ATL-NYC (1 transition goal, 2 ATL buildup goals). So in this span, 3 out of the 5 matches had no “normal possession goals” and the two that did featured Atlanta. My hypothesis remains that no team can count on possession/build-up-created goals as it’s “bread and butter.” The other ways to score are much more common. Possession/buildup (if executed correctly) is first and foremost (at least in a league like MLS) a tool to control the game and minimize the amount of time your opponent has the ball. And it comes with risks. It’s just the futile act of watching your team chase a game down 2 goals that makes you feel like they’re lacking in ideas or unable to break teams down. It happens to the best of us, but from what I’ve seen, Atlanta has the skill/talent/tactics to create the odd possession-based goal from scratch, and “occasionally” is a good enough rate for this in my book. Set pieces are maybe a problem? I’m not sure. I defended the short corners last week, and I think there were some shots created off corners this week, but pretty soon I’ll be out of defenses for the zero set piece goals created. Oh, and while we’re at it, it occurs to me that Atlanta hardly ever draws fouls just outside the box, where some of the more dangerous set pieces are taken. For a team with dynamic dribbles and quick speedsters, where are the set piece opportunities? Something to keep an eye on. One would imagine Almiron and Asad would be drawing fouls just outside the box often.

That thing Mears does

That all being said, if we are to take a deep dive into some possession/buildup stuff…I know I’m not the only one bothered when the ball is switched over to Tyrone Mears, only for everything to halt and as the defense shifts back into place before the ball is played back to Parkhurst. My game notes from the last several matches are filled with stuff like “ball is switched to Mears, who plays back to Parkhurst.” I stared at a couple of these instances from this week’s game. My immediate reaction is always that Mears is too slow, or timid, or too right-footed, and he’s ruining these possession for us. But as I watch these sequences over, I’m just not sure what’s going on. It’s clear that Garza/Asad have something nice going on over on the other flank that Mears hasn’t found with Villalba or Gressel, and so my inittal thought was that Mears has trouble finding those square passes into the advanced midfielders that he’s overlapping (say a Gressel, or an Almiron or Kratz). But here are some screenshots from one such sequence on Saturday, and maybe you can help me figure out who’s to blame (I mean it, it’s not clear to me really):Mears15aComments after all the pics.Mears15bMears15cMears15e

It looks to me like maybe at the beginning he’s too hesitant to keep driving forward (there’s still some space in front of him, and everyone is moving forward), but he’s also looking for someone to pop into some space and no one does — they just keep running. Further, he has to be pretty cautious about trying to take on a defender or take more space, given how many red shirts are in nice spots in the center circle area should he turn it over (with Garza also wide and high on the left). Then he stops the ball (the trademark Mears move it feels like), and every single player makes the exact same come back to the ball type move (or they’re stationary), except Kratz who does a clever little but perhaps overly nuanced overlap? So it’s unclear if Mears misses his chance to play someone through, or if no one is able to find actual space. Ideally you want someone running in behind and someone dropping in front of the centre backs, and someone else showing for theball in midfield etc etc.

My solution would be for when Almiron returns, for him to make a more concerted effort to not drop back into midfield to start play, and instead to remain high tucked between the lines, and then importantly, to commit to coming across field when the ball is switched to Mears. I would be OK with Mears playing a ball slightly forward and into Almiron’s feet with a higher risk for a turnover, if perhaps there was more cover deeper to handle a counter — maybe Garza drops back with purpose when the ball is switched? It’s all quite difficult to understand — that’s the game I suppose.

Let me know your thoughts.

Week 3: ATL vs CHI

IMG_0242.JPGThey said not to analyze the game where Atlanta was up a man for 80 minutes. I ignored them.

Summary: Another odd match for United. The game turned twice in Atlanta’s favor in the opening 11 minutes, once with a dangerous cut back from Gressel which caromed in off of a Fire defender (1-0) after the play was started by a Joseph Martinez recovery and Villalba playing Gressel in on the right, and then second when Gressel’s ambitious¬†interception of a throw-in¬†and then smart through ball to Martinez forced a last man foul and a red card all in a matter of seconds, fundamentally altering¬†the game. Gressel should have 2 assists instead of zero for those two plays — we’ll see what we can do about that further below.¬†The home side¬†was slightly shaky to close out the next 30 minutes even if Martinez had another great through ball chance denied, with the hosts giving the ball away in counter-ish type spaces a few times, but ultimately, comfortably closing out the half as I sprinted to beat the beer lines in stoppage time (sorry not sorry, anyone at opening day knows what I’m talking about). Atlanta came out in the 2nd half bossing the game, and cracked it wide open with a beautiful throughball from Almiron to Martinez in the 60th. At that point Chicago was in a bad spot, not able to sit back as much. Atlanta poured on 2 more including a delicious throughball from Asad to Martinez¬†which we break down further below, and Tito’s first.

Match Analysis: Julian Gressel lovefest:

That’s the 2nd excellent¬†performance¬†from Gressel in as many games. Here’s his chalkboard from MLSSoccer.com which I modified to more properly credit him for the¬†opening goal and the ball that did the sending off. He covered space well, recovered balls well, and played dangerous passes, again. Also narrowly missed a hard driven shot from deep (the red circle).

3Gressel

Match Analysis, for real this time:

Well, one of the reasons many are pouring water on this win is that we only have¬†somewhere between¬†4 and 11 minutes of normal game state to take a look at. Let’s give it a go I guess.

Opening segment (0′ – 4′). Atlanta on the left. Chicago on the right:

That recovery and shot (the only official shot of the opening 4 minutes)¬†from Chicago’s David Accam deep in Atlanta’s half doesn’t look great for the hosts– it was the result of a ball recovery from Almiron that was then turned over immediately in his own half (red backwards pass on Atlanta’s map) to the Chicago striker. ATL may not have registered a shot in this opening segment, but its opening goal was not completely unearned. Martinez wins the ball in Chicago’s half¬†after the ball changes possessions a few times, and he’s¬†dispossessed with Tito running on and playing Gressel in. Gressel makes the right decision and plays the ball across the six yard box — he has two attackers making runs, whether he can see them or not.

1-0 lead, before the red card (4′ – 11′)

Neither side were able to generate much more than a half chance between the first goal and the red card. A little more activity from Atlanta (left). I mentioned the red card in the intro. In my opinion and¬†from where I was standing, it seemed like the right call. Martinez was passing the last defender, even if the non-fouling centre back was near. I know I was howling for it in the moment but not expecting it just because it’s such a big decision early in the game. There was no unbaking this cake.

1-0, (11′ – 45′)

After the elation of waving goodbye to the departing Fire defender wore off,¬†I was worried Atlanta may over-commit and struggle to break down the compact 10 man Fire, leaving themselves open to swift counters from Accam, the one thing you cannot do up a goal and a man. I didn’t calm down until half time (which is where I’m stopping the clock on the images below). It wasn’t just the chances Atlanta was giving up (see the three in the box below, right)…

It was the turnovers in midfield troubled me the most. First, the ball recoveries and successful dribbles for Chicago in Atlanta’s half during this period of play (11′-45′):

3-cardHTrecoveries

Then, the mirror image of unsuccessful passes from Atlanta in their own half after going up a man. At the top of the image is Almiron drawing a yellow card for a good professional foul to stop a Chicago counter that looked troubling.

3-cardHTATLunssuccessfulpass
all of these green passmaps from MLSSoccer.com

A lot of the buildup during this period just seemed needlessly risky to me. I kept thinking, if you’re up a goal and a man, kill the game until half time, but they were intent to press on. By stoppage time though, I was comfortable enough to dash up the stairs and grab some beers. Anyhow,¬†the game looked much different after the break.

2nd Half

Below are the ball recoveries (yellow triangles), interceptions (blue triangles), dribbles (green or red) and crosses, key passes, assists, and shots for the second half. Chicago (right) simply didn’t create much at all.

Almiron’s pinpoint throughball to Martinez came off an interception and a great timed run through the heart of Chicago’s defense. It wasn’t so much that the turnover created a numbers-advantaged break. Martinez simply confounded the centre backs and Almiron found him with precise timing. Martinez calmly rounds the keeper again (twice in two goals) and not for the last time….

The one good chance for Chicago came in the 67th minute (3-0), an early cross to a¬†half-marked Accam who had good positioning on Pirez for the header. Well saved by Kann. This doesn’t worry me too much. 11tegen11’s excellent passmaps and expected goals materials for this game below. A lot has been said about the high possession figures and pass %s, but I think¬†the most important thing is the chances comparison in the second half.

Highlight:¬†The 4th goal, Asad’s throughball to Martinez is worth breaking down since it’s similar to Almiron’s 1st goal last weekend in that it was created through fluid movement on the left side during a sustained buildup phase. Very pretty (even against a wounded team).

First, the ball is played out wide¬†left to the right-footed Asad, and what do know about right-footed Asad? Aside from being the surprise of the season for me, we know he likes to come inside from the left. Chris McCann provides the “underlap” setting this whole thing up. That blurry guy heading towards the corner flag is him.gifAsadstartAs Asad takes his first touch to come inside, Atlanta already has a 2v1 and a nice pocket of space between the right fullback (who has gone with McCann), and the centre back. Almiron notices this and looks to move into the space (I think Martinez probably likes it himself and considers occupying it as well), but wait…3gif almiron moves into spaceAs Almiron moves into the half space, the nearest CB holds close to him, which provides Martinez what he needs, space in behind. Notice Peterson chilling on the right and occupying two Chicago Fire players, for no reason whatsover. This helps things. Also of note: Asad immediately has line of sight into the danger area, whether it’s going to be Almiron or Martinez who receives the ball.3gifmartineznoticesMartinez makes his run and Asad looks up again and sees the better option. As¬†you watch the highlight of this play, notice¬†the pace he puts on the ball, not only the required amount to thread the many defenders, but also a cue to Almiron that the ball is not for him but for someone else, if there was any doubt.3gifasadpass¬†To control the driven pass and round the keeper in one touch is something special.3gifconnectionIs that “something special” finishing skill¬†or Luck, statistical variance, the hot hand?…It’s a difficult question, but to me, any narrative that attempts to suggest Atlanta’s success to date is unsustainable because of their “clinical finishing” (implying primarily “chance”) so far is unfair. Either we need to recognize that for whatever reason Atlanta is repeatedly creating 1v1-on-the-keeper situations for their striker and that these chances are crazy good or we need to admit that finishing is something Martinez is better at than the average forward. We might need to admit both these things. After he rounds the keeper for the third consecutive goal (3 in 2 games), I’m not saying this pace will persist, but any insistence that these games should be draws and Atlanta is just lucky seems wrong.

The part where I went way too long but¬†I don’t have the discipline to edit:

I think of #finishing as both broader¬†and narrower than “goals minus expected goals” (the metric most commonly associated with it)– broader because I enjoy¬†Mike Goodman’s discussion¬†of finishing as possibly being several subskills all wrapped into a thing we sometimes¬†call “finishing” (i.e. more than just the physical ability to convert a given shot profile into a goal) — but also narrower, because to me, finishing isn’t something that encompasses all shots. Finishing is what happens (or doesn’t) in the 18 yard box or in otherwise “good” scoring situations. I don’t think half volleys from the D (like Almiron’s last week), or curlers from 25 yards out (like Feilhaber’s this week) should be factored in in any way to something like “goals minus expected goals” to track finishing. When finishing is evaluated, or its existence as a concept proven or disproven, players should neither get credited nor penalized for “converting”¬†(or not) those¬†types of chances – and I realize that’s got some sample size problems. Admittedly, Martinez’ sharp angled finish on the above play is an odd place to launch into this treatise (speaking of sample sizes, 3 games), but perhaps his¬†#finishingskill¬†was on display not when he rolled the ball in from Upside Down dimension at Bobby Dodd, but instead the finish was the moment he touched the ball on this play (when he turned a rocket of a pass deflected out for a goal kick into an open net chance from an acute angle).¬†Or, perhaps this is all statistical noise and Josef Martinez and David Accam and Carlos Bocanegra all have the same finishing skill.