I took a look at Seattle to see what there was to see from their first 3 matches this season, and then I looked back into 2016 in search of something that might help Atlanta on Friday.
Note: All the maps below are from MLSSoccer.com match recaps. Some additional stats stuff from AmericanSoccerAnalysis.com (@analysisevolved on Twitter).
Seattle’s Results to date:
- @Houston: 20 minutes level at 0-0, then behind. Lost 2-1.
- @Montreal: 17 minutes level at 0-0, then behind. Equalized at the death.
- vs NYRB: 28 minutes level at 0-0, then ahead. Won 3-1.
Seattle Chance Creation (ExpG: 0.99 @Houston, 2.28 @Montreal, 2.47 vs NYRB)
Pretty dangerous stuff here in terms of shot locations (central), and several key passes (shot assists, yellow) that are not deep crosses.
Chances Conceded by Seattle (xG: 1.62 @Houston, 0.71 @Montreal, 0.45 vs NYRB)
One caution against reading much into the above images is the impact of game states on Seattle’s opening 3 matches. For instance, in the first two, Seattle trailed for the last 70 minutes or so, while Seattle led for the last 60 minutes against New York. Seattle has only played a total of 65 minutes with the score drawn over the first three matches (which admittedly is more than Atlanta’s 38 minutes). But what if we looked at some maps of Seattle’s 65 minutes of even play?
Level Game State Chance Creation (@Houston, @Montreal, vs NYRB)
Not a ton of activity here, but what is here looks dangerous enough. Against Montreal, that key pass coming in from the right looks fairly dangerous from Shipp in the highlights – fortunately for Montreal, it was off target.
Level Game State Chances Conceded (@Houston, @ Montreal, vs NYRB)
Seattle’s opponents were much more active shooting and creating shots from passes in all three matches during the level game states, even if most chances were from distance (Piatti’s brilliant throughball for the opening goal in Montreal being the very big exception — if Martinez were healthy Atlanta might be licking its chops — alas.
One thing that jumps out to me from the even game states is defensive aggression. Blue interceptions, green tackles, gold ball recoveries for Seattle and against, below:
Level Game State Seattle Defensive Aggression
Compare to: Level Game State Opponents Defensive Aggression
This seems like a finding. It’s unclear whether it’s simply that Seattle sets up differently at home compared to the first two away, but while the games were still even, you can clearly see Houston and Montreal forcing the action further forward in Seattle’s half with the Sounders happier to sit back while the game was still 0-0. Hosting the Red Bulls, the Sounders were a bit more adventurous before the Dempsey penalty. Even in Seattle, it may very well be that the Sounders sit back against Atlanta who will be pressing early and often. If that is the case, Atlanta should be extra careful not to turn the ball over in dangerous areas as they are prone to doing on occasion. Lodeiro, if he plays, could punish those moments.
Level Game State: Seattle’s Passing
In all minutes played through 3 games, Seattle has averaged around 5.5 passes per minute and 81% completion, but in the 65 minutes of even game state, Seattle has passed at a lower rate of 4.8 passes per minute and 74% (with less possession). It would appear that when Seattle is trailing they have more of the ball and are passing at a clip of 5.8 ppm, and they’ve trailed for half of the minutes played so far, so their overall average passes per minute and % are higher than those played when the score is level. In the opening drawing 28 minutes, the Red Bulls passed at a clip of 5.8 ppm and 74% against Seattle. Lots of small sample size disclaimer here, but my guess is that even at home against Atlanta, Seattle will pass at a low 70%s success rate, and the away side will have plenty of the ball similar to Red Bulls, until “something happens.” Reminder: Seattle is happy doing this, so caution is recommended. I think this guess fits decently with this research by Jared Young at American Soccer Analysis, which found that Seattle is the least proactive team defensively during even game states. Interestingly though, Seattle were one of only a handful of teams that becomes more proactive defensively when they are ahead than when drawing. Further, Seattle is the only team in MLS that gets more direct in attack when leading… Some of this data may be stale since Sigi Schmid was fired midway through the 2016 season.
In search of a Game Plan
I looked back at the 2016 Sounders in an attempt to find instances where they both 1) gave up a high amount of chances, and 2) also lost the overall chance creation battle 3) at home (using expected goals data from AmericanSoccerAnalysis). I found a handful of games, but none in the back half of the season after Roman Torres returned from injury (late August) and they signed Lodeiro (late July) and changed their manager. There are a couple of very uneven games with Seattle playing away (at LA on 9/25, and at FC Dallas on 10/16), but there wasn’t much to work with in terms of home games – they were simply good at home during the final stretch. In the end I remembered something about Seattle winning the MLS Cup without creating a single good chance in the final against Toronto. While this one was also on the road for Seattle (played in a cold Toronto), it feels like an important game to look at. Further, it makes sense to me that while Seattle won that game in penalties, Atlanta might want to replicate what Toronto did for 120 minutes leading up to those penalties. So we’re gonna focus on that — also I would say a full 120 minutes of even game state at 0-0 is precious decent data to work with.
In the MLS Cup Final, Toronto created plenty of chances both with short incisive passing just outside the 18 yard box and wide with Rickets (notably Frei’s amazing save off the Altidore crossed header). They also had a few chances off of turnovers/interceptions in Seattle’s half. Toronto outshot Seattle 19-3 and 7 shots on target to none. 11 key passes to 2.
Select Seattle unsuccessful passes (GK is filtered out).
If Atlanta can replicate this sort of untidy passing in Seattle’s own half, they can create chances in transition — the best chances. And the kind of chances Atlanta is aiming for.
Back to the Cup Final. In the end, Toronto produced 1.1 expected goals worth of chances to Seattle’s 0.1, but Seattle were crowned champions. Such is life.
Looking at some pass maps, it also looks as though Torres had to play more passes than Chad Marshall. Perhaps Toronto did something to deny distribution to Marshall? Torres was also dispossessed a couple times in his own half.
Only two of these passes from Torres make their way over to Marshall, and NONE go back to the keeper. Makes me think Toronto’s attacking players disrupted this sort of ball cycling at the back. The reason I looked specifically for this is…
Centre Back Distribution:
Chad Marshall is a relatively tidy passer (averaging in the low 80%s his career). Roman Torres is noticeably worse than that (low 70%s). It seems to me that while pressing, Atlanta United should shade the ball back to Torres (and not Marshall) and force him to hit it long back to ATL whenever they can’t trap and force a turnover in midfield/fullback areas. To illustrate, below are pass maps (Marshall playing LCB and Torres playing RCB) and while they are small and hard to see, just look at where the red passes are, mainly. From left to right, the MLS Cup Final @ Toronto, Week 1 @ Houston, Week 2 @ Montreal (didn’t include game 3 cuz Torres was suspended):
So… I guess gameplan-wise, since we know Martino will press/attack relentlessly, when pressing, try to force Seattle to play back to Torres instead of Marshall. Be cautious of turning balls over in midfield for Lodeiro (and Shipp) to create with. And an editorial comment: watch out for Joevin Jones on the left finding 2on1s with Mears if Villalba stays high then that 2nd goal (OG) conceded against NYRB could be recreated by the Sounders on the overlap methinks.
That being said, if Torres doesn’t play on Friday night due to resting after the international break, you can throw all this out the window. Others who played on Tuesday and so might not start: Joevin Jones, Dempsey, Kenwyne, Almiron (only 45). Josef Martinez definitely will not play…for a while.
Weak but specific predictions: Look for a midfield triangle of Carmona–Gressel-Almiron, and Jacob Peterson (?) leading the line. Martino has also used Asad at forward in preseason….Predicting 1-1 with lots of pressing from Atlanta to open the match. Asad opening the scoring for ATL after a spell of 65% possession, and Shipp creating a goal to equalize for Seattle in the second half. Good first look at how a long cross country trip will impact pressing intensity.