Houston Dynamo Stats Scouting Report

Atlanta want to play out of the back (especially at home) to control the game, AND they want to press high up the field to win the ball back close to (or within one throughball of) their opponent’s goal to create high quality scoring opportunities.

Houston want to bunker on defense, invite Atlanta into over-committing in attack, and break quickly with fast counters. With the ball, they’re happy to play long balls (18% of passes are long) and they average 46.7% possession, which is on the low side.

Both teams should get their wish, in terms of overall architecture of the game.

Buildup and Pressure

I’ve been tracking some stats around possession and giveaways in a team’s own half, partly to better evaluate Atlanta’s progress in learning to play out of the back, and partly to assess the importance and vitality of their high pressing tactics and their ability to disrupt the opponent’s buildup on the other end of the field. While I’ve been focusing on Atlanta, I was curious about Houston, and so I ended up compiling their stuff as well this week. For the discussion below, I define “giveaways” as the sum of the following actions, all taking place in a team’s own half: unsuccessful passes, unsuccessful dribbles, dispossessions from tackles, unsuccessful touches.

When we adjust for the record-breaking Chicago game where Atlanta was up a man for most of the game, they’re averaging 183 own half passes per game, completing 84% of them, and giving the ball away there 38 times per game (which, while I don’t have this data for all teams, seems extraordinarily high). Conversely, Houston is spending less time passing in their own half (125 per game), completing 88% of them (which if I had to guess might be pretty standard), and the dynamo are giving the ball away just 23 times per game.

On the other side of the coin (measuring high pressure), Atlanta’s opponents this season are averaging only 116 own-half passes per game (which I would guess is low), completing 83% (also low), and giving the ball away in their own half 31 times per game. Conversely, Houston’s opponents (who are enjoying more possession) are averaging 168 own-half passes per game, completing 88% of them, but also giving the ball away 31 times per game.


Where I’m headed with this is that unless we witness significant tactical departures on Saturday, Houston will bunker and look to break, and Atlanta will build out of the back, push numbers high, and press when Houston is in possession. Both sides will have opportunities to turn the other side over. And while this all suggests Atlanta will have the lion’s share of the ball, I can’t help but think we could have a track meet at times when Houston gets to running, and Atlanta pressing the counter to thwart these moments.

I find all of soccer to be difficult to predict (MLS especially), but this complementary set of styles (possession vs counter) may exacerbate the extreme outcomes. I would guess that Houston is completely comfortable in this equilibrium. Further, I might be undervaluing Houston’s transition threat by focusing on the “high press turnovers” (opponent’s half) when Houston is also dangerous countering from deep. Interestingly, through 11 games, there’s a high correlation between Atlanta’s chance creation numbers and their opponents’ giveaways (ATL putting up bigger xGs in games where their opponent is sloppier in the defensive half), but for Houston this relationship doesn’t really show up.

If I had to suggest a specific tactic for Atlanta it would be to focus some pressing on the hometown veteran Rico Clark. My stuff suggests he is more prone to the “giveaway” than his teammates, right around 3 per game. Also, right back DeLaGarza is sitting on a quite low 80% pass accuracy in his own half. So, perhaps shading the ball his direction could be fruitful, to the extent that Houston isn’t already bypassing their own half with long balls.


And Houston certainly are dangerous. Despite not having a lot of the ball, they are creating chances. AmericanSoccerAnalysis has Houston 2nd in MLS in expected goals for (underlying attacking output) and expected goal differential (+0.53/gm). Here are some shooting stats (shots for and against on the left Y, and ordered by shot diff on the right y):Shots.PNG

Houston are generating some of the highest shooting totals in MLS. And interestingly, set pieces are a big reason for this. They’re taking 1 or so more per game than the average MLS team and 3 more than Atlanta PLUS they’re converting them at an exceedingly high rate (though one would not expect this to persist).Set Pieces.PNG

In terms of shots output from open play, both teams are sort of in the middle of the MLS pack (Atlanta at 9.3/game and Houston at 8.6) with Atlanta converting these into goals at a lead-leading 18% rate, and Houston a below average 8%.

Ben Baer at MLSSoccer.com wrote a piece laying out a gameplan for slowing down the Dynamo. It involves playing a low block and forcing Houston to have the ball — I just can’t see Tata allowing for this. So, if we assume that the game will look sort of like the DC game in terms of possession and geography, when the inevitable turnovers happen in midfield it will be vital for Atlanta to thwart counters before they start.

On the possession side of things, if aggressively playing short out of the back isn’t working early (you’ve seen them, where they’ve got the centre backs on either side of the 18 yard box), I would call an audible just to minimize the chance of having to chase a 1 goal deficit for most of the game. I would think that the longer the game remains level, the more solid Atlanta’s home field advantage will be. Rule #1 will need to be avoid getting in upside down game states against a great bunkering/countering team.

Last week, Atlanta looked much improved with winning second balls in midfield, which was important since Portland played so directly when pressed. I suspect this will be important again, and might be the most noticeable thing for the viewer. On Sunday Saturday, if you notice Atlanta picking up those loose balls in midfield, you’ll probably be having a good time. If not, it may be agonizing.

Houston’s attacking weapons are hard to ignore at this point with 4 chance creators over 0.6 xG+A per game in Torres, Elis, Manotas, and Quioto (data from ASA through last weekend’s results).

I’m thinking this one’s 2-1 Atlanta with a penalty 🙂 but like I said, it could go in any direction, really.

Saturday 7pm Fox Sports Southeast


3 thoughts on “Houston Dynamo Stats Scouting Report”

  1. I’m worried that this is going to be the match that Houston gets their act together on the road and runs all over our back line. Or that we have another DC situation where we probably should have won but Houston gets a couple of lucky/poorly defended goals to take the points. It will definitely be a good one.

    “On Sunday if you notice Atlanta picking up those loose balls in midfield, you’ll probably be having a good time. If not, it may be agonizing.”

    I will be watching for this. But, I hope you don’t show up on Sunday because you’ll miss the match as it’s on Saturday. 😉


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s