Bayesian MLS Best XI: Week 5


This week-by-week MLS Best XI aims to take out some of the reputation and biases that exist in creating a Best XI at the end of the season. Using Bayes’ Theorem based logic (not any Bayesian statistics) I am creating my overall Best XI for the 2017 season from the first week through the last. Players will only be replaced when I determine they have passed a player at their position on the whole of the season. Essentially, this is the 2017 MLS Best XI up to this point. There is no set formation and it can shift as it needs to.

This is the point of the year that I expect season-long data to start infiltrating my decision-making process. Not to override what I’ve seen by any means, but to supplement and check what my eyes are telling me. I also don’t intend this to become a WhoScored list of players with the most actions for their position, though it is information worth considering.

Take defenders as the most stark example of this. The absolute best defender in MLS might have the most tackles, clearances, and interceptions at his position. A whirling dervish of goal-saving blocks and tackles. On the other hand, it’s very possible – and in my opinion likely – that the best defenders are ones that don’t need to fill up the box score.

This is much harder to spot and I’ll do my best. The box score counting stats are not generally a bad thing, except in the case of a great game-saving block after foolishly giving the ball away near the top of your own box. Looking at you Ashley Cole. It’ll just take some patience to connect the counting stats to what I’m seeing and to allow both to steer me to the best possible outcome.

With that out-of-the-way, Week 5 is in the books and we’ve come to the week where MLS teams diverge in games played. Right now the league ranges from three (Orlando and their rained out game) to five games played. For the next six months or so points per game will be the best way to look at team success, not the place in the standings.

The short international break, or international slow down, has us here and weeks 4 and 5 will admittedly be less valuable for me. A lot of the players on this Best XI so far have international responsibilities and didn’t play these two weeks or played as a 30 minute substitute. I very much enjoyed getting to see some players I don’t normally get to evaluate, but it does lessen the importance from a Best XI perspective.

*Note about the stats below – they’re all in per 90 minutes and they have been sorted by position and a minimum of 3 appearances to be included. I’m also flipping between 90 minutes and per game fairly freely. This is very technically wrong, but fits for the most part this far into the season.*


GK: Tim Melia (SKC)

I’m both happy and a bit uneasy that my three top guys for this position are also the top guys in xGA-GA – how often a team scores against the keeper compared to how often they “should” score. This is a good thing because it means I’ve been identifying successful keepers (not exactly difficult to find successful results) but this way of ranking keepers is too simplified.

I’m more encouraged by the results because it’s an objective way of looking at things, which I like, and Bill Reno gave 10 examples of good goalkeepers – to varying degrees – and he had high praise for Melia on both shot stopping and some of the difficult to pin down skills as well. This list did not include my other two players listed below: Tim Howard and Alec Kann.

In the Hunt: Tim Howard (COL), Alec Kann (ATL)

LB: Greg Garza (ATL)

I was enamored with Garza’s varied ways of helping the team offensively in Week 1. Some of that has carried over, but Garza has mostly become a classic overlap, crossing left back. Importantly, Garza is a very good version of that archetype. He’s tied for second with 1.3 accurate crosses/90 for defensive backs and has a low 1.8 inaccurate crosses/90. Finally, the Atlanta defender is tied for first in Key Passes/90 with a man I’ll highlight below, Joevin Jones.

On the defensive side, Garza attempts 2.5 tackles and has yet to be marked with a “Dribbled Past” or failed tackle and is third in the league in passes blocked/90 for all full backs. While he hasn’t lived up to my lofty (and almost certainly unjustified) praise to start the season, Garza has filled the role of overlapping left back to a T.

I just want to note the names below. While “In the Hunt” speaks for itself, there are varying degrees of how secure a player is from the rest. Justin Morrow and Joevin Jones are both hot on the heels of Garza. Jones more so after Morrow missed his Week 5 matchup with a now ironic heel injury.

Jones attempts an astronomical 5.3 total dribbles – highest among backs – and is successful on 2.8 of those attempts per 90 minutes. This high activity offensive style doesn’t result in being dispossessed often by its very nature. Jones is first in defensive backs for fewest dispossessions at just 0.3 per 90.1

In the Hunt: Justin Morrow (TOR), Joevin Jones (SEA), Fabinho (PHI)

CB: Leandro González Pírez (ATL)

You’re looking at my biggest bias for a single player right now. I very much enjoy Leandro González Pírez. His positioning appears good to me on the aggregate but he’s also busy on defensive actions and is one of the better center back passers in this short season.

González Pírez is great at completing short passes, a common quality of center backs. More importantly – based on difficulty of the task, not necessarily importance – González Pírez has the most accurate long balls/90 for center backs with 10.5 and is inaccurate with only 4.5 per 90. The first player to have less inaccurate long balls is Lawrence Olum and he attempts roughly 5 fewer long balls per game.

Flipping to defense, the Argentinian is tied with Drew Moor for most shots blocked/90 with 1.8. I prefer a moderate to low clearance rate for my center backs. There’s always a need for a quality clearance, but good positioning and quick thinking can often lead to a tackle/interception leading to possession of the ball instead of an inaccurate clearance that will likely cede possession. González Pírez is a bit high with 5.9 clearances per 90, especially compared to his compatriot here, Ike Opara.

In the Hunt: Jonathan Spector (ORL), Drew Moor (TOR), Alexander Callens (NYC)

CB: Ike Opara (SKC)

I’ve described Ike Opara’s free ranging role in previous Best XI’s. The defensive structure and style of Sporting KC allows Opara to make plays in the open field and he’s excelling. Opara is fifth in attempted tackles for center backs (3.3/90) and is tied for the second-fewest times dribbled past per game at 0.3.2

Opara has had a weakness in the air, winning just 2.8 of 3.3 aerial duels per 90 minutes. I noted above my preference for clearances and despite his defensive action counts, Opara has just 3 clearances per 90 – enough to release the danger but not so much that he’s booting it longer than necessary. The positioning of Opara higher up the field than most center backs during the run of play likely contributes to a bit of this and his high tackle rates eliminate some need for clearances.

The clip below shows the good and the bad. Sebastian Giovinco turns Opara around and gets open in the box. That’s a common sentence for any center back tasked 1-on-1 with Giovinco. However, Opara makes a good recovery to put pressure on Giovinco before he takes the shot.

In the Hunt: Jonathan Spector (ORL), Drew Moor (TOR), Alexander Callens (NYC)

RB: Nick Lima (SJ)

The shine is coming off Nick Lima a bit, especially on the offensive side of the field. He was a focal point of the attack in their win over the Vancouver Whitecaps in Week 2. Lima does lead the full backs with 3.8 total dribbles per 90 and 1.8 successful dribbles. He could be more selective, but for a rookie, that’s more than acceptable. The right back position hasn’t impressed me too much this season and I’m keeping an eye out for someone to step up.

Lima bottled up Rodney Wallace fairly well in Week 5. That’s high praise in my opinion as Wallace has been one of the best players this season for me. The rookie right back is a very selective tackler and hasn’t been dribbled past yet despite playing every minute this season.

In the Hunt: Alvas Powell (POR), Keegan Rosenberry (PHI), Steven Beitashour (TOR), Tyrone Mears (ATL)

CM: Aníbal Godoy (SJ)

I started this write-up with Godoy falling out and Miguel Almirón staying in. They were fairly neck and neck for me and my memory of Godoy’s quiet Week 3 overtook the asterisks surrounding Almirón’s success thus far. Let me clear one thing up, I think Almirón has been fantastic, but diving into the stats this week lead me to believe that Godoy deserves to keep the spot for now.

Godoy has been the most successful dribbler for central midfielders in the league behind only Armando Cooper in volume with 4.3 total dribbles per 90 (Cooper has 6.1). However, Godoy is successful on 3.7 attempts and only fails on 0.7 attempts compared to his international teammate’s 2 successful and 4.1 unsuccessful.

Once he opens up space, Godoy also has a healthy 2 key passes per 90 minutes and is tied for the most accurate long balls (10.3) with a miniscule 1.3 inaccurate long balls for center mids. Godoy is pulling the strings from the midfield well while making late runs for…ambitious shots. The biggest thing Godoy can work on for me is taking higher quality shots.

In true number 8 form, Godoy isn’t in the top-tier of total tackles attempted, but is tied for fifth with 4 successful tackles per 90 and more importantly gets dribbled past just 0.3 times per 90 minutes.

In the Hunt: Kellyn Acosta (DAL), Maxi Moralez (NYC), Alex (HOU), Miguel Almirón (ATL), Dax McCarty (CHI)

CM: Diego Valeri (POR)

I’ll keep this one short. I think Valeri has the most secure spot in this lineup right now. He continues to score in various ways, has the second-highest xG+xA/96 with 0.92 (behind only Lee Nguyen who has 150 less minutes). I mean, the below clip is quite nice.

In the Hunt: Kellyn Acosta (DAL), Maxi Moralez (NYC), Alex (HOU), Miguel Almirón (ATL), Dax McCarty (CHI)


LM/LW: Justin Meram (CLB)

This week put my previous starting left wing and starting right back against each other in Rodney Wallace and Nick Lima. Neither stood out in this matchup, though I suspect some of that is due to Lima taking a defensive approach. Wallace was not silenced by any means, but Justin Meram was hot on his heels already and has continued to perform.

I really wanted to, and came close to, cheating this week and moving Meram over to the right and keeping Wallace installed here on the left. This might be what I’d do in a real life situation, but alas, this isn’t real life and I’m going to be a bit position strict. This may become moot with Columbus acquiring Kekuta Manneh – generally a left midfield/winger. Meram could slide to the right if Columbus favors Manneh over Finlay. That’ll have to play out before I can slide Meram and Wallace into this team together though.

Meram has more creativity both on the ball and in his passing. Wallace is more of the type to find the open space and be there when the ball arrives. I’m not knocking that style at all, but Meram’s play is just a touch higher at the moment for me. Wallace actually leads the position with 1.8 Key Passes/90 to Meram’s 1.43 which I didn’t expect.

The Columbus star has a more well-rounded offensive game and they’re both good enough defensively. I see Wallace as a bit better though Meram a bit safer in case you’re wondering. Meram boasts the fifth-highest xG+xA total at 3.25 while Wallace has an xG heavy xG+xA of 1.81.

This wasn’t meant to be a Wallace vs. Meram comparison section, but they’re neck and neck for the best left-sided attacker.

In the Hunt: Romell Quioto (HOU), Rodney Wallace (NYC), Darlington Nagbe (POR)

CM: Haris Medunjanin (PHI)

Key passes aren’t a perfect statistic. They just measure how often someone shoots off of a given player’s pass. It’s heavily dependent on another player’s individual decision process. But a top ten that involves Valeri, Sacha Kljestan, and Benny Feilhaber can’t be all bad…right? Medunjanin is second for central midfielders (attacking, defending, and otherwise) in key passes with 3.3/90 minutes.

I’ve used string puller before, but Medunjanin is one of the best examples of this. Someone lying slightly deeper, though often not at a number 6 position like Dax McCarty and Michael Bradley, and spraying balls around the field. Usually this results in the pass before the pass and very few shots. Both are fairly true, but Medunjanin is also getting the pass as shown by his key pass numbers.

Part of the above is bolstered by high volume tactics like crossing (4.6 per 90) and taking corner kicks. He’s converting well on these types of attempts, but they’re not quite ideal in my opinion. My bias tells me someone like Almirón, McCarty, Alex, or Acosta will rise to this position as soon as next week. Players like Michael Bradley, Feilhaber, and Kljestan may start to sneak in here as well.

I can feel the apprehension while typing this and selecting Medunjanin, but I’m going with it for now.

In the Hunt: Kellyn Acosta (DAL), Maxi Moralez (NYC), Miguel Almirón (ATL), Alex (HOU), Dax McCarty (CHI)

RM/RW: Jack Harrison (NYC)

I considered dropping in Nicolas Lodeiro here but his high dispossession numbers and few appearances drove me off…for now. Alphonso Davies might have a coming claim despite his age and I’ll attempt to check my bias for prospects at the door.

Ultimately, with the international break altering things a bit this week, I settled on Jack Harrison. I consistently take note of Harrison in every NYCFC game. He may not be as good as his wing teammate Rodney Wallace, but this position doesn’t have someone playing on the level of Meram. Hence why I almost moved Wallace or Meram to the right.4

An inverted attacking winger is one of my favorite types of wingers and Harrison fits perfectly. He’s a selective but good dribbler with pace and a nice shot after he cuts into the heart of the defense from outside. A cut back across the box to a curler to the far post is a common site from Harrison, but that shot opens up the ability to cut even deeper sometimes for a shot like the one below. This one is a bit lucky and the curler from right side of the box is something close to a 0.07 – 0.09 xG shot, but it’s useful in the grand scheme of the offense.

I can feel myself back tracking on this already. We’ll see what happens in Week 6 with a full complement of players (mostly).

In the Hunt: Kevin Molino (MIN), Alberth Elis (HOU), Sebastian Blanco (POR), Nicolas Lodeiro (SEA), Alphonso Davies (VAN)

ST: Fanendo Adi (POR)

Strikers score goals. There are plenty of ways that a striker can contribute outside of scoring goals, possibly just as much in the aggregate if a teammate is finishing. My initial example is early season Jozy Altidore in 2016. He could not find the back of the net, but his defensive work rate, hold up play, and threat opened up space and chances for his teammates.

Unfortunately, Josef Martínez is out for some time, though I considered leaving him in this slot in the short term. I love that he leads the league in shots within the penalty area. Alas, it’s time to move on until Martínez gets back on the field. You can’t be the best player at your position if you’re not on the field (for this activity). Christian Ramirez is playing great now that he’s getting his MLS legs underneath him. Ramirez averages about a shot in the 6-yard box per game.

Close behind though, is Fanendo Adi with 0.7 shots in the 6-yard box per 90 minutes. Adi is second in xG according to ASA with 3.01, behind only Ola Kamara’s 3.79. I’m a bit skeptical of Kamara’s total and want to see that play out a bit. I’m feeling low rent (very low rent) Edinson Cavani on Kamara at the moment. Gets into great positions so often that his xG totals increase while he finishes at an average rate. This is a good thing, I just want to see it play out a bit before I make a decision on Kamara.

I don’t think about Adi as a dribbler, but with 2.8 per 90, Adi is just ahead of Giovinco on the season. The Nigerian plays with a very smooth power when he’s on the ball. You’re not going to see a ton of speed or ambitious dribble moves, but more nuanced positioning and using his size and power to hold off defenders while he heads toward goal.

The defender actually does well to push Adi further toward the end line here, even if going into a slide tackle was a risky proposition. But this is the sort of dribbling I see with Adi, he doesn’t showcase his strength in holding the ball with a defender on his hip here, but it’s a good example from this week.

Aerial volume is heavily related to the style of the Portland Timbers, but with relatively (and I stress relatively) low volume of 4.8 aerial duels per 90, Adi is winning 2.8 of them. He’s not quite elite despite his size, but he actually doesn’t need to be with the style and quality of passers surrounding him in Valeri, Sebastian Blanco, and Darlington Nagbe.

The success of a striker can be heavily reliant on the success of a team. Martínez will stay on his trail once he returns. Ramirez is showing why he is such a successful poacher. David Villa continues to be a steady presence passing and scoring. And Erick Torres continues to score goals, though with a lower xG total – which is both a good and bad thing.5

In the Hunt: David Villa (NYC), Fanendo Adi (POR), Cyle Larin (ORL), Josef Martínez (ATL), Christian Ramirez (MIN), Erick Torres (HOU), Ola Kamara (CLB)

  1. This can be misleading and needs a bit more clarification. A high attempted dribbles in itself keeps dispossessions low because a dispossession requires that there is not a dribble attempt. It does imply that Jones is not being idle with the ball but is being aggressive in attacking his defender. 

  2. Johan Kappelhoff, Jelle Van Damme, and Drew Moor have more than 1.5 tackles per 90 and zero times dribbled past 

  3. This is a fairly worthless difference at this point in the season 

  4. For my own sake I hope Meram moves to the right with Manneh coming in but that may be unlikely with Finlay still playing well for Columbus 

  5. He’s finishing tough chances but can he be expected to continue it? 

About the author: Colby Rogers

Colby is the Editor-in-Chief, Founder and Lead Contributor to Other League. Also a law student focusing on Labor & Employment law and intersections with law and sports. You can find him on Twitter via @Colby_OL.