file under: tragicomedy

Twenty-seven percent possession is no problem if you do it right

Yesterday, a friend sent me these quotes from Philadelphia Union coach Jim Curtin, pulled from an interview he gave to the Inquirer after Philly’s 2-0 win over NYCFC on Saturday:

“Here in Philadelphia, we don’t care about possession. Possession without purpose is meaningless, and we’re OK with teams — I think in probably every game we’ve played so far, they’ve led in the possession numbers, but possession doesn’t win games.”

The article by Union beat writer Jonathan Tannenwald sets Curtin’s Union up as a counterpressing side in the Klopp / Rangnick model: a high press, quick attacks, fairly direct passing. It’s probably a little early, after four matches played, to really test whether that description fits this season’s Union, but hey, might as well try. Here are some takeaways from that match and from the Union’s season so far.

It’s a dynamite start to 2022

Jim Curtin has been around longer than fbref have been publishing MLS data, so I went as far back as I could. Here’s a rolling average of the Philadelphia Union’s expected goals for and against, going back to the 2018 season:

I use a ten-game rolling average to account for the kinds of radical swings that can happen game to game; if you have, let’s say, a hapless D.C. United side at home one week and MLS Cup-winning Columbus Crew away the next, the lines will jump all over the place. Here, you can see that the Union played their best football under Curtin in early 2019 (when Brenden Aaronson came into the team) and then in a 2020-2021 stretch that ends in July (Jamiro Monteiro’s last appearance before his summer leave of absence was 08 July vs. NY Red Bulls, and, neatly, everything trends down after that point).

As you can see, the Union ended 2021 hot and started 2022 on a tear. The attack is clicking as well as it has since we got good in 2019, and the defense is allowing less than one expected goal a game. We’ll see how sustainable it is, but right now, the Union look great.

Possession numbers explain nothing here

I wanted to test Curtin’s specific claim — that possession numbers mean nothing to us. Curtin is going against accepted wisdom here; a 2021 study in Portuguese journal Motricidade found that, in the UEFA Champions League at least, teams with more ball possession won 49.2% of their matches, drew 23% of the time, and lost 26.5% of the time.

However, in this limited case at least, the numbers do seem to bear out what Curtin’s saying. Here, I’ve charted a ten-game rolling average of the Union’s possession percentage against their expected goals difference over the same period:

Overall, these variables have next to no correlation (r = .0675), meaning that the Union’s possession numbers don’t explain their performance over time.*

Funnily enough, the one game this year when Philly “won” the possession battle was the 1-1 draw against Minnesota in week one (53% possession); in the following three games, all wins, Philly had 36% at Montreal, 31% against San Jose, and a surprisingly low 28% at NYCFC. Each of those wins was, as Tannenwald notes, “deserved” according to StatsBomb’s xG models (via fbref) — the Union actually created the least threatening chances in the Minnesota match (1.0 xG from 19 shots) when they had the majority of the ball, and they created much better chances (2.2 xG from 13 shots) in the NYCFC match.

Game state matters

When I say “game state”, I mean the scoreline at a particular moment in the match. Game-state effects are obvious when you’re watching a match live (“they grabbed the lead early and parked the bus to protect it!”) but can kind of disappear when you’re scrolling through match reports (“how the hell did they score if they never had the ball?”). Studies have shown, for example, that Leicester City’s 2015/16 title can be at least partially explained by game state; Leicester got lucky in that they often scored first, forcing teams to open up against them in search of an equalizer, which was really dangerous to do against Jamie Vardy and Riyad Mahrez.

mlssoccer.com has (I don’t know how I never saw this before) possession stats for five-minute intervals as part of their match reports. Here’s the chart for NYCFC vs. Philly:

Having watched the match live on TV (admittedly on my second screen, I had Middlesbrough v. Chelsea on the first), it didn’t seem to me like we had 27 or 28% of the ball, and this match timeline explains why: in the buildup to the first goal (scored in the twelfth minute) and the second goal (scored in the 33rd), Philly genuinely controlled the match. After the second goal, they had almost none of the ball, and they created almost nothing, because, well, why bother? You’re up 2-0 away, none of the shots New York are taking are particularly threatening, you’ve got an established back four and the best goalkeeper in the league. There are, of course, plenty of teams who’d handle the situation differently (Guardiola teams, yeah, but Swansea famously played for “defensive possession” during their last stay in the Premier League — they sort of just kept the ball in their own half all the time so that nothing happened), but, to turn Curtin’s phrase on its head: if you don’t need a goal, then you don’t need the ball.

That’s not to say the Union sat back and did nothing for the whole second half. One of the best measures of defensive engagement we have is passes per defensive action, or the number of passes attempted by the attacking team divided by the number of tackles attempted, interceptions made, & fouls given away (in short, the number of attempts, regardless of their success, to win the ball back) — a low PPDA is active defending, a high PPDA is passive defending. Philly’s PPDA for Saturday’s match was 12.37, which is pretty low — the league average in 2021 was 10.21 by my calculations. The Union’s rolling average has hit twelve plenty of times over the past few years, though it’s generally been trending downward (read: we’re pressing more intensely) since the start of 2019:

Game state and defensive positioning

There really aren’t a ton of second-half highlights in the video posted by the MLS YouTube channel after the match. It’s a 7:51 video, and the second half starts at 5:34. This feels about right; I don’t remember much happening in the second half, and the xG chart shows a few low-value NYCFC chances and nothing else. The fact that the Union had a comfortable lead mostly explains why. Not rocket science, but here’s an example: the ball gets turned over in midfield, and, because he always does, Union midfielder Jose Martinez moves up to put pressure on the ballcarrier:

The awkward Yankee Stadium camera angles briefly make this look like a dangerous situation. The Union’s diamond midfield is spread way out, with Alejandro Bedoya out on the right wing and Leon Flach caught upfield. If fullbacks Kai Wagner and Nathan Harriel were in their usual advanced positions, this’d be trouble. A quick pan of the camera reveals a very compact back four with everything in front of them:

Here’s a shot from the Minnesota game where the Union, chasing a go-ahead goal, basically have seven players up in the attack; that’s Wagner at the top left calling for the ball as he makes a late run into the penalty area.

Wagner is a very useful attacking fullback, hitting 3.71 crosses per 90 in 2021. He’s been in the upper echelon of attacking fullbacks in MLS for a couple years now & was recognized as such with an MLS All-Star game spot in 2021. The Union are at their best when he’s up providing width on the left flank, drawing the opposition fullback out so the left striker, normally Sergio Santos, can run into the channel. If we’d needed a goal against New York City, Kai probably would’ve been up here:

but we didn’t need a goal, so he wasn’t up there, and the move ended with Taty Castellanos trying to rip the ball out of Andre Blake’s hands for some reason. And that’s how you win a game with 27% possession: get a few goals early and then hang out and enjoy the spring day.

* if you’d prefer, the Union’s game-by-game expected goals difference has an r = .2642 correlation with their game-by-game possession stats — this is a very very weak positive correlation but worth mentioning because it’s a different number from the rolling averages one

On the Ricketts family, Chelsea FC, and American oligarchy

Let me tell you a quick story. It ends with the execution of Carey Dean Moore, a Nebraska man who confessed to killing two cab drivers in 1980. I guess it technically begins on 01 May 1975, when the SEC deregulated the stock market by eliminating fixed brokerage commissions, a move that allowed everyday people like you and me to make or lose a couple hundred bucks in the stock market. Discount trader First Omaha Securities “broke ranks” (their words, not mine) with major brokerage firms, offering discounted rates to “make Wall Street more accessible to the individual investor”; the company, now known as TD Ameritrade, manages $1.3 trillion in client assets, and its founder, Joe Ricketts, is worth, according to Forbes, $4.5 billion.

This might not sound like it has much to do with Chelsea FC, but bear with me.

The Ricketts family have featured prominently in Republican party politics since Joe stepped down from the TD Ameritrade board in 2011. Father Joe, described by CNN as “the patriarch of one of the most influential (and wealthy) families in conservative politics” and by Politico as a “megadonor”, spent $5.5 million trying to stop Donald Trump from winning the 2016 nomination through the Our Principles PAC; once Trump became inevitable, Joe pledged over $1 million in 2016, and another $2.5 million in 2020, to pro-Trump super PACs. (Trump, apparently willing to forgive and forget, nominated his son Todd for deputy secretary of commerce.)

Todd’s older brother Pete long harbored political ambitions of his own. Back in 2006, he ran one of the most embarrassing Senate campaigns in recent memory, spending over $11 million of his own personal money on his campaign and losing by thirty points to Ben Nelson — a Democrat — in Nebraska. (No Democrat has won a statewide race since.) Undeterred, Pete spent another million running for governor in 2014. He outspent his opponent almost three-to-one and won. He has served as Governor of Nebraska since.

Ernie Chambers is, depending on who you ask, a harmless kook, a reprehensible and irresponsible rabble-rouser, a legend of Nebraska politics. He was first elected to the Nebraska state legislature in 1970, served until 2008 when he was term-limited by a new constitutional amendment, ran again in 2012, won handily, and served another two terms as the rep for majority-black North Omaha. In 1980, he introduced and passed a resolution calling for divestment from apartheid South Africa; Nebraska was the first state to do so, and other states followed. In 2007, he sued God to make a point about frivolous lawsuits; his point was mostly lost in national news coverage. And in 2015, on the thirty-seventh attempt (once a year since 1973), Chambers passed a bill out of the state legislature to eliminate the death penalty in Nebraska. Governor Ricketts vetoed the bill, but, as reported here by Mother Jones, the bill survived a dramatic floor fight and was passed into law:

As senators continued to rise to speak for another hour, the lobbying of (State Senator Robert) Hilkemann by both sides grew so intense that he left the floor. Two citizens who were standing in the hallway reported seeing representatives from the governor’s office shouting at him before the final vote. After Hilkemann finally decided to continue his support for the repeal, sealing the 30-19 victory, death penalty activists rushed to the chamber door to embrace him and shake his hand. He smiled broadly but insisted that he didn’t deserve any special recognition. “We were all number 30,” he said.

The death penalty was suspended in the UK in 1965 and formally abolished in 2003. Most of Europe has done away with it. “Perpetrators of crimes must be held accountable and punished,” says a 2020 statement from the Council of Europe. “However, the experience of abolitionist countries has shown that the death penalty does not deter violent crime nor contribute to a safer society. On the contrary, killing as a punishment perpetuates a cycle of senseless violence.” In the United States, twenty-seven states have the death penalty, but thirteen of those haven’t used it in over a decade. Temporarily, Nebraska reduced it to twenty-six. Almost immediately after the repeal vote, though, an organization called “Nebraskans for the Death Penalty” formed with the goal of reinstating executions through public referendum. Unfortunately, their website is down, but here’s a description from a 2016 article in The Intercept:

In November, voters will have an opportunity to overturn LB268, thanks largely to the efforts — and considerable family fortune — of Gov. Ricketts. Denouncing the law as proof that “the legislature has lost touch with the citizens of Nebraska,” Ricketts poured his own money into repealing it, funding a petition drive launched under the banner Nebraskans for the Death Penalty. Last summer, the group attracted enough signatures to put the issue on the November 2016 ballot; in September, the Lincoln Journal Star reported that the group had raised more than $913,000 — “a third of it from Gov. Pete Ricketts and his father, Joe Ricketts,” the founder of TD Ameritrade. Last month the Nebraska Supreme Court heard oral arguments in a lawsuit alleging that the governor improperly downplayed his role driving the referendum in order to mask the “violation of his duty” to enforce state laws, even those with which he disagrees.

*nebraskans for the death penalty dot com, 14 March 2022*

The lawsuit failed, the referendum passed, capital punishment was reinstated, and Carey Dean Moore, who’d been on death row for almost forty years, who’d had 2007 and 2011 executions halted, who’d seen the repeal effort succeed from his cell, was executed via a lethal injection of illegally obtained fentanyl. “Capital punishment,” said Governor Ricketts, “remains the will of the people and the law of the state of Nebraska. It is an important tool to protect our corrections officers and public safety. The state continues to carry out the sentences ordered by the court.”

In between all this, the Ricketts family bought the MLB’s Chicago Cubs for $845 million. Reportedly, they’re interested in buying Chelsea. In articles about this interest, current Chelsea owner Roman Abramovich is described each time as a Russian oligarch. The Guardian describes the Ricketts family only as owners of the Chicago Cubs. The Athletic notes that Tom Ricketts is a soccer fan.

What is an oligarch? Joel Samuels, professor of law at the University of South Carolina, describes oligarchs as “private individuals who benefitted from government connections to amass huge wealth”, largely from the privatization of state enterprises.

By owning the major industries – from oil and natural gas to steel, nickel and other basic industries – Russian oligarchs held sway over all aspects of daily life. With time, they expanded to other areas, such as media and sports. But, at their core, the oligarchs gained their wealth from industry and wielded their power openly and with impunity.
Over the past 20 years, the number of oligarchs has grown, but the base of their power remains the same: a relationship with the president that leads to personal financial gain.

Samuels of course uses post-Soviet Russia as his descriptive example, and none of this is to deny that Roman Abramovich is a rich and powerful man from Russia. There is such a thing as an oligarch. Oligarchs are men so rich and so well-connected that, whether or not they personally hold office, the political system warps around them. Oligarchs have gravitational pull. Maybe the President owes them a favor. Maybe their son wants to be governor and maybe their son gets mad when the legislature overrides his veto and he needs some money to run a ballot initiative to override the override. Maybe their other son wants to win the World Series so that people will like him. And why stop there? Maybe he’s a soccer fan, too, and the European champions are suddenly for sale at a knockdown price.

Maybe there’s not a “Good” person who could buy Chelsea. Maybe it’s impossible to be a good person in any sense and have $3 billion or £2.5 billion or however you want to write out the reported purchase price. Maybe Tory property developer Nick Candy or medical device tycoon Hansjörg Wyss or the “no government links” Saudi Media Group would be worse. And maybe mentioning airstrikes in Yemen or discrimination against trans people in Nebraska or the whole thing about spending millions to get elected to state office and then losing one vote in the legislature and spending another million to get that vote overturned so you can kill a guy is whataboutism or whatever. I’m all for applying scrutiny to the owners of huge community institutions like football clubs. Let’s just make sure to apply the same scrutiny to the American oligarchs, too.

New data viz for the blog

In the past, when doing scouting reports or player comparisons on here, I’ve mostly used this github site to make StatsBomb-style radars with fbref data. It’s relatively easy and works well enough. There are a few reasons, though, why I wanted to make something new.

1. Percentiles

The github site linked above has you input minimum and maximum values for each stat you want it to plot. Ideally, instead of raw values, you’d have the player’s percentile rank compared to their positional peers. The median number of progressive passes per 90 for forwards in the top five European leagues in 2020/21 is 1.99; the average number of progressive passes per 90 is 2.25. (This is largely because, if you look at the data, the really creative forwards play between four and seven progressive passes per 90. This would already fuck up the average of a dataset where most guys are under 2.0, but then, at the very top Josip Iličić plays 8.24 and Lionel Messi plays 9.43.) If a guy plays two progressive passes per 90, he’s very slightly better than 50% of his positional peers, so I want the data point to be just past, not just below, halfway. fbref does give percentile stats on their player scouting reports, but if I use those with the radar generator, I can’t show the per/90 numbers without making a whole separate chart. Ultimately, I decided to download a bunch of data from fbref, sort it by positional groups, and use a percentile rank function to plot data points on the chart.

2. Simplicity

StatsBomb use different templates for different position groups, meaning they can convey more detailed and relevant information for a given player. But they’re not the easiest to read if you’re totally new to soccer data.

I therefore see the appeal of something like The Athletic‘s pizza charts, a clean, universal data visualization that works for every position group.

*Three colors for three areas of the game makes sense, too, I’m stealing that*

As I talked a bit about in my last post, though, the smarterscout numbers are sort of opaque. Attributes from one to 99 like FIFA stats makes conceptual sense, but which numbers are being combined to create “Defending impact”? Are they baseline stats to which the average fbref user has access, or is it based on, like, tracking data that I don’t have? How can I compare the numbers I can find to the numbers smarterscout uses?

3. Time

The way I was doing things before — get a bunch of csv data from fbref, put it in Excel, filter, compare, type in github generator, save, copy, paste — was not a particularly efficient use of time. (Using Excel now probably isn’t the most efficient, either, so sure, I’ll learn to code someday.) This blog is an after-work hobby; the tedious parts should be cut out so I can spend more time doing the fun parts.

To that end, I spent a lot of time today setting up a sheet that, in the future, can spit out graphics really quickly and easily. It uses simplified terms on a pie / radar chart, adjusts for possession, and gives both per/90 stats and percentiles compared to positional peers in the top five European leagues. I can customize it more easily because I know how it works, and I can really quickly turn player data into something presentable and, more importantly, intelligible:

Everything means more or less what you’d expect it to.

Attacking (purple)

Shot volume is shots per 90. Basic box score stat. How often do they shoot compared to other players?

Shot quality is non-penalty expected goals per shot. How dangerous are the shots they’re taking? These two numbers combined would give you xG/90, which StatsBomb puts between these two points on the chart — but, since it’s one number times the other, do we really have to include it separately?

Shot-creating actions is purple and with the shooting attributes for now. It includes passes, dribbles, fouls won, saved shots, etc. that lead directly to a shot attempt. It’s the closest we can get to a universal “attacking” attribute: how often is the player on the end of an attack?

Passing (blue)

Expected assists is the number of assists the player would have with totally average finishing and goalkeeping. It’s “chances created” but better because xG is a more precise measure than “chances” or “half-chances” created.

Pass completion percentage. Sort of a stylistic measure, but it’s gotta be on there.

Progressive passes per 90. These are defined on fbref as passes that move the ball at least ten yards towards goal or passes into the opposition penalty area. A necessary companion to assists; the Jorginhos of the world insist.

Dribbling (turquoise)

Progressive carries per 90. Progressive carries move the ball at least five yards forward or into the penalty area. Taken together, these two points show how involved the player is in moving the ball up the pitch.

Dribbles completed per 90. I went with completed dribbles rather than attempted or success percentage because attempted isn’t necessarily a marker of skill and success percentage isn’t a useful number if we don’t know how many are attempted. You’d want all three for full context, but I’m counting here on the player’s coaches and teammates to tell him, if he’s losing the ball constantly while attempting dribbles, to cut it out.

Defensive (aqua)

Pressures. Success rate could be good to know, too, but knowing whether a player is active in defense is probably enough.

Tackles plus interceptions. Possession-adjusted.

True tackle win rate. Explained by Tom Worville here. I use tackles/(tackles+dribbled past+fouls committed).

Aerial (teal)

Aerials won. I could see putting this in the defensive category (The Athletic does), but it’s not really a defensive stat, is it? Kacper isn’t in the 80th percentile for aerials won because he defends a lot; he’s in the 80th percentile because he’s 6’4″. Some defensive headers, sure, but probably some shots, probably a lot of clearances or out-balls. So it’s got its own color.

Anyway, I know there are a million different scripts or sheets or programs or whatever that can output soccer data visualizations, and I’m not sure mine is better than any of them, but it’s mine, and I like it, so I’m gonna use it. At the very least, judging from the Kacper Przybyłko radar, it’s producing graphs that pass the eye test. I’ll try to see what’s available on the Union’s new striker, but no promises — some stuff might not be available for the Danish league.

Yeah but Tammy Abraham could do it just as well

Basically everyone is using the Romelu Lukaku interview from a couple weeks ago to reinforce their prior assumptions about the big man, so I’m going to do the same. Here’s strikers from Europe’s top five leagues on a scatter chart, goals per 90 on the Y, non-penalty xG plus assists per 90 on the X:

Here’s a couple radars from their fbref numbers this year:

And a stylistic comparison via smarterscout:*

Finally, a couple GIFs:

I dunno, maybe letting him go was a bit premature?

Real quick: I’m not gonna pretend to know how smarterscout’s algorithm measures all of this stuff. They explain it here and I read it but it basically comes down to whether you think North Yard Analytics have a good model or not. What you need to know is that the numbers on this chart represent stylistic frequency: that is, how often does the player attempt a dribble, how often do they engage in an aerial, how often do they disrupt an opposition move, etc. Tammy’s in the red, obviously, and Lukaku is in blue. The numbers bear out what many, including Tifo’s Alex Stewart, have pointed out since The Interview dropped: Tuchel is asking Lukaku to play as a central target man (Receptions in box: 94; Dribbling: 56; Progressive passing: 25; Aerial duels: 48) for Chelsea, which is NOT how he played last season for Inter (Receptions in box: 91; DRIBBLING: 78; PROGRESSIVE PASSING: 55; AERIAL DUELS: 25). Or, for you visual learners:

Basically, there’s something to the idea that Lukaku, if not being “misused” by Tuchel this season, is being used differently than how he was used at Inter last year. But that whole argument is totally irrelevant to my thesis, which is that Tammy Abraham could basically do anything that Tuchel has asked his center forward to do this season, so it was silly to sell him and then pay double for Lukaku, who is probably a better ball-carrier than Tammy and maybe a better passer but who isn’t better at doing any of the things that Tuchel apparently wants Lukaku to do as a lone striker. Add in the fact that Tammy is pressing more for Mourinho than Lukaku is for Tuchel (and is more involved in the build-up, too), and it seems pretty clear that Tammy is as good a stylistic fit for this Chelsea side, maybe a better fit, than the man who replaced him.

How to win the league playing dogshit soccer

Watching the New England Revolution’s playoff loss against NYCFC last night (2-2 [3-5 pens]), I was struck by, well, just how bad they looked.

*Most of the second half looked more or less like this*

The final shot count (NE 15 [4 on target] — 16 [4] NYC) gives the impression of a fairly even game, and StatsBomb’s xG model (via fbref) concurs: New England accumulated 1.2 xG to NYC’s 1.6, which is the kind of gap where a loss is fair but, if things go your way, a draw is fair also.

I can’t find an xG match summary for this one (though I did find a tutorial on how to create an xG match summary in R… maybe that’s something worth learning over winter break?), but we can pretty easily tell what happened just by looking at the shot timeline.

A whole seven of New England’s fifteen shots came in extra time. Five were after Taty Castellanos scored in the 109th minute, four were after Castellanos was sent off in the 113th, and three, including the equalizing goal, were with NYC right back Tayvon Gray hobbling around holding his obviously pulled hamstring.

This was not an even match. This was a home playoff game where the runaway Supporters’ Shield winners went down early, equalized off a set piece, and then played a full ninety minutes of absolute dogshit soccer. And yet, they drew; and yet, they had chances to win in extra time; and yet, they would have won had the game gone on another five minutes; and yet, anyone can win on penalties.

Of course, I started wondering: have New England actually been this bad all year?

There are a few common ways to determine whether a soccer team is as good as its record, and I’ll hit the big ones. The first is expected goal difference, and even if you’re not a data guy, it’s fairly intuitive: how do the chances you’re getting compare to the ones you’re giving up? If a team’s expected goal difference (xG minus xG allowed) diverges wildly from its goal difference (goals minus goals allowed), then they’re, whatever you want to call it, “hot”, clinical, overperforming, lucky, et cetera. Teams or players can and do overperform compared to xG for long stretches of time (think Leicester City 2015/16, who came fifth in the xG table according to understat but slightly overperformed while all the traditional big sides underperformed) before eventually reverting to the mean. Strikers (except the very best ones) will outpace xG one year and fall behind the next, and most teams end up scoring mostly in line with expectation over time.

For New England, xGD is our first major red flag. These numbers come from the incredibly useful data app on American Soccer Analysis, which has New England at a relatively disappointing eighth in xGD (9.75). The Revs had the second-best goal difference in MLS (22), but they overperformed in attack (G: 62, xG: 56.59) and in defense (GA: 40, xGA: 46.84), and a twelve-goal swing is probably enough to send a good-but-not-great side top.

Again, overperformance of this sort generally isn’t long-term sustainable, unless you’re dealing with remarkably talented players like Harry Kane or exceptionally unique managers like Lucien Favre. By his own account, Bruce Arena is not one for that fancy stuff (“Analytics and statistics,” he said in 2016, “are used for people who don’t know how to analyze the game”), but his side has, since he was appointed in May 2019 (yellow line on the graph), scored mostly at the rate we’d expect them to; the Goals line peaks and troughs with the xG line just how you’d expect, and this year’s overperformance shows up in those two huge spikes toward the end.

That’s not to say Arena didn’t have a positive effect on New England when appointed. He definitely did, with the caveat that New England look absolutely awful here under Friedel, and they haven’t been consistent world-beaters under Arena. This is the profile of a good side, not a historically great one.

That more or less eliminates the “Managerial Warlock” explanation, leaving us with two: either New England have some really good players, or their record-setting 2021 season was the product of, depending on how you think of it, statistical variance or dumb luck.

In front of goal, Gustavo Bou, Tajon Buchanan, and Emmanuel Boateng were hot this year (distance above the 1:1 trendline), and leading scorer Adam Buksa got a respectable 16 goals from 16.3 xG.

Buchanan in particular is one to watch going forward. Bou and Buksa seem to have largely found their level in MLS, but where Buchanan will end up after Club Brugge is anyone’s guess.

That just leaves one potential answer: goalkeeper Matt Turner. John Muller, whose work I like a lot, called him “the most valuable soccer player in America” in a column for FiveThirtyEight this summer (Muller’s since moved to The Athletic), and his argument, which I’ll let you read for yourself, is pretty convincing. Turner is good, and as Sam Stejskal details here, the Revs found him through pure happenstance. According to StatsBomb’s post-shot expected goals model, he saved six goals that an average keeper would’ve been expected to let in (well, 5.9), good for third in the league in 2021. There’s half your overperformance, and a hot year from Bou and Boateng gets you the other half.

*Source: https://fivethirtyeight.com/features/the-most-valuable-soccer-player-in-america-is-a-goalkeeper/*

There are some loose ends left to tie up here.

Do the 2021 New England Revolution stick out in any other obvious ways? Well, they led the league in scoring with 62 goals, for which they deserve credit, but the leading scorers from 2019 (the last full 34-game season), LAFC, scored eighty-five (85), 2018’s leading scorers (Atlanta) got 70, 2017’s (Toronto)74… It’s not like New England blew people away, at least by recent MLS standards. Their lower expected goals total suggests some good fortune was involved, and they ran up the score in a 5-0 win at Inter Miami (it was 4-0 at halftime) and a couple 4-1s (both were 3-0 before halftime). Of course, a goal’s a goal, but I’d love to look at their schedule game-by-game and see how often they were playing from in front versus how often they had to come from behind; game state can influence a lot of these numbers in ways that don’t show up when you aggregate them. They took 14.62 shots per 90, good for fourth in the league, and those shots were of about average quality. They had average possession stats (50.2%, as average as you can get), dribbled an average amount, pressed an average amount, and scored six goals from set plays (league median: five). Hell, their total and average guaranteed salaries, per American Soccer Analysis again, are sort of low by league standards. Apart from the points, there’s not all that much worth remarking on.

So, how do you win the league without being particularly good? Have one or two guys get hot in front of goal; get a good keeper if possible, and a coach who’s at least average. Being decent to start with is helpful but isn’t strictly necessary. And, of course, it always helps to play in MLS.

The Philadelphia Union need more guys

Last year, no team registered more defensive pressures than the Supporters’ Shield winners, and only three teams (LAFC, NYCFC, and RBNY) registered more pressures in the attacking third. Philadelphia will miss Brenden Aaronson’s energy and attacking output along with Mark McKenzie’s defensive solidity, but the club’s tactical identity will not change in 2021.
Joseph Lowery, “MLS tactics at a glance” — The Athletic, 14 April 2021

On Sunday, the Philadelphia Union drew 1-1 with Chicago Fire. The Union could rightly claim to have been unlucky; they outshot Chicago 25(!)-4 (xG: 2.0 — 0.2, per StatsBomb via fbref), with six of those shots coming in an encouraging first seven minutes. Chicago scored from their first real attack via some slack far-post marking by Olivier Mbaizo:

*Two attackers there, and he’s not really picking up either of them*

but Philly quickly reestablished control and, especially after Wyatt Omsberg’s 34th-minute red card (and the subsequent free kick, converted by Kai Wagner from maybe two inches outside the box), looked like the only possible winners.

Whether by coincidence or by design, the Union started the second half equally strongly. Cory Burke, in particular, was lively throughout, and he came close to putting Philly ahead in the 47th and again in the 48th:

*The look Cory’d be wearing for the rest of the match*

After the initial flurry of chances, though, the Union quieted down, not registering a shot between the 54th and 71st minutes. Ilsinho came on in the 63rd for Leon Flach but didn’t create much. Every Union attack kind of looked a bit like this:

Slow passes behind a static forward line followed by a low-percentage shot from distance. I was at the match with my sister, a former player, who immediately picked up on our lack of movement. “If you’re marked,” she said, “you’re not going to get open by standing still,” and there was a lot of standing still going on in the dying minutes.

If one of the Union’s many clear-cut chances had gone in, the whole match would look very different in retrospect; we’d be talking about grit, determination, champions winning when they’re not at their best, etc.

But, on the whole, this was a tired performance from the Union, one that exposed what I’d say is this team’s biggest problem: the Philadelphia Union don’t have enough guys.

I’d argue that guy-deficiency is the underlying cause of all of Philly’s issues this season. Guy deficiency explains why last season’s Supporters’ Shield winners are hanging around in fifth with 25 points from 17 matches, as close to the Eastern Conference basement (Inter Miami, 12 pts.) to the top (New England, 36 pts.). Guy deficiency explains the seven draws (fourth-most in MLS, behind Houston and Nashville [9] and Atlanta [8]) and the recent string of disappointing results. Frustratingly, guy deficiency was a foreseeable problem; qualifying for continental competition means you’ll need more guys, and the Union lost more regularly contributing guys (five; Ray Gaddis, Warren Creavalle, Andrew Wooten, Mark McKenzie, and Brenden Aaronson) than they brought in (two, by my count; Daniel Gazdag and Leon Flach). Stuart Findlay is a guy, but he’s played 113 minutes total, and Alvas Powell is a guy, but he’s only been here a couple weeks. I have no idea how good he is, but midfielder Jesus Bueno, signed from Venezuelan club Deportivo Lara last week for an undisclosed fee, is a guy, which goes some way to addressing the guy problem, but it’s hard to bed in mid-season MLS acquisitions, and not all your signings work out (e.g. Oravec), but, credit where credit is due: he’s a guy. If the Union want to defend the Shield, make a playoff run, and give themselves a shot in the CCL semifinal, they’ll probably need a couple more guys, and quick, because guy deficiency is blunting their attack, slowing down their midfield, and leaving their defense exposed, and as the season goes on, it’s only going to get worse.

Last year, the Union were able to rely on a consistent core of guys who basically played every available minute. Andre Blake in goal; Mark McKenzie in defense; Jamiro Monteiro, Ale Bedoya, and Brenden Aaronson in midfield; and Kacper Przybylko in attack all clocked over 1,800 of 2,070 available league minutes in 2020. That wasn’t a problem last year; even including MLS is Back, the season was significantly shorter than usual, and travel distances were reduced when the league abandoned inter-conference games due to COVID travel restrictions. This season, however, the Union are set to play at least 40 matches (up from 23), and as of today, a full nine players have played more than 80% of available minutes.

Five Union players are in the top 25 for minutes played in MLS this year, more than any other club (the closest is Seattle with three):

And it looks worse when you take the CONCACAF Champions League into account. Obviously, CCL clubs in general dominate this chart (more available minutes means more minutes played), but again, the Union are particularly averse to rotating; nine of the top 25, and eight of the top eleven, minutes-getters in America this season are Union players, and Andre Blake would be up there if you include the three matches he played at the Gold Cup — that’d put him at 1800 minutes, even with Michael Bradley for fifth overall.

That’s a lot of soccer, in a short period of time, for a small number of guys.

The effects show up not only in the table but in the data. Last year, when basically every team in world football relaxed their press, the Union made 161.8 pressures per 90 and 36.8 attacking third pressures per 90, good for second and fifth in the league, respectively. Those numbers were basically level with 2019 on a per-90 basis; 161.7 pressures per 90 was only seventh in MLS that year, but with 39.1 attacking third pressures per 90 (fourth in MLS), the Union were still among the highest-pressing sides in the league.

This year, the Union’s numbers are, funnily enough, about the same (168.7 pressures per 90; 40.0 attacking third pressures per 90), but, compared with the rest of the league, they’re being handily outrun.

It’s possible that, contra Lowery’s (and my own!) assumption, the Union are not all that aggressive a pressing team in general; they simply kept playing their usual game in a year when everyone else was sitting back, and having a small squad didn’t hurt them because the season was shortened. (That’s a slightly different argument, a different post, maybe, for a different time.) This year, the squad is smaller, the guys are all a year older, and the league is, as a whole, pressing harder. And, given the number of minutes in the Union players’ legs, it’s no surprise they aren’t keeping up.

There are a couple solutions available to coach Jim Curtin. He could rotate the senior players more frequently, giving more minutes to guys like Matt Real, Anthony Fontana, and Stuart Findlay. He could send a bunch of seventeen-year-olds out there at once; Quinn Sullivan seems to have surpassed Jack McGlynn in recent weeks, and Paxten Aaronson is a regular on the bench. Or, he could sign more guys. So far, he’s chosen the latter; in this window, Philly have picked up Powell, Gazdag, and Bueno (also known as Chuchu), and who knows if there’s money for more. Signing more guys only works if you play the new guys, though; the cases of Matej Oravec and Findlay (and arguably Andrew Wooten) speak to a possible disconnect between recruitment and coaching staff at Subaru Park that’ll have to be addressed at some point. Hopefully, the Union get a couple more guys in, like, today, because the CCL semis are coming up, and they’ll need all their guys for that one.

Matej Oravec — Philadelphia Union Statistical Review

Been a bit since I posted here (other work to do), but the Union have made a couple interesting moves in the summer window, and I thought I’d take a look at one today: the outgoing loan of defensive midfielder Matej Oravec to Slovakian club Železiarne Podbrezová. Below the jump, I’ll be taking a look at the stats from Oravec’s year and a half in Philly, examining a few clips, and analyzing his overall game at the usual level of detail you’ve come to expect from this blog. So, without further ado:

¯\_(ツ)_/¯

Too-soon Leon Flach scouting report

I’ve wanted to do a quick radar-and-GIF post on this guy ever since we signed him, but fbref doesn’t track 2. Bundesliga or CONCACAF Champions League advanced stats, meaning the sample size was, until this weekend, too small to justify it. (Hell, even if fbref had 2. Bundesliga advanced stats, he only played 120 minutes for St. Pauli anyway!)

But, with the MLS season chugging along nicely, we’ve now seen just about enough to start making judgments about potentially another Ernst Tanner gem: the totally unheralded German-American central midfielder Leon Flach.

Flach’s defensive numbers should immediately jump out at you. The twenty-year-old is leading MLS in basically every relevant per-90 defensive stat. He’s first for pressures, first for successful pressures, first for successful pressure %, first for tackles, first for tackles plus interceptions, near the top for blocks, clearances, and % of dribblers tackled; over 468 minutes thus far, Flach is here, there, everywhere.

https://widgets.sports-reference.com/wg.fcgi?css=1&site=fb&url=%2Fen%2Fplayers%2F39859fa1%2FLeon-Flach&div=div_scout_summary_MF

Some of this is, of course, team style — Philly are fourth in the league for pressures per 90, marking them out as one of the more active teams without the ball — but even compared to 2020 Jose Martinez, whom I would have considered to be an active defensive midfielder, Flach is nuts.

One caveat: Flach played pretty much exclusively as the left-sided eight in the Union’s 4-4-2 diamond until Martinez was suspended for elbowing Valentin Castellanos in the face (something we’ve all wanted to do for a while, let’s be real); since then, he’s played at the six, which according to this 2020 interview with American Soccer Now is where he sees himself long-term. Presumably he’ll slide back out to the left side when Martinez returns to the eleven, playing essentially the position N’Golo Kante so controversially filled for two and a half years at Chelsea: a ball-winning eight in a midfield three.

On that question, I was a Sarri / Lampard sympathizer; if you’re wedded to a midfield three (or a diamond) in a high-pressing system, it doesn’t make sense to pin someone with Kante’s energy down in front of the back four. When you’ve got a player who’s equally adept at hassling defenders outside his own box…

and tearing down the wing in the counter…

you have to, as a coach, find a way to use that.

Honestly, there’s something of a “poor man’s Kante” about Flach, both in his abilities and his deficiencies. Here’s Kante’s 2018/19 Sarri season for comparison:

Apart from Kante being a much better (criminally underrated throughout his career) passer, you’re looking at a similar shape here: an aggressive ball-winner who’s decent at moving the ball forward while maybe lacking a bit of quality in the final third. It’s noteworthy that 2018/19 Kante and 2021 Flach are both in the 92nd percentile of midfielders in their respective leagues for carries into the final third per 90 (Kante: 2.79; Flach: 2.12); Flach could probably carry the ball through midfield more often than he’s currently doing, and if I’m Jim Curtin, I’d be encouraging him to drive forward whenever possible.

I missed the New York Red Bulls match (band practice) but would otherwise, I guess, caution that Flach’s last couple performances haven’t matched the standards he set in the early going. He’s looked a bit tired at times:

*& caught out in an area where he really shouldn’t be*

and he doesn’t seem particularly likely to break his goalscoring duck on current evidence:

*I’d give this one 0.02 xG maybe*, *given where the defender is & that it’s on his left foot*

but, look, this is a free transfer who’d played barely one senior match before going straight into the Union eleven, and he’s hardly looked out of place. I’ll check back in on him in ten games or so — can he do a bit more in possession, maybe put one of those counters in the back of the net? — but so far, Union fans can be pleased with Leon Flach.

The Super League is about wages, too

Most of the posts about the newly-announced European Super League have correctly noted that this is “about money” (New York Times), that it’s being driven by “capitalist greed” (Jacobin), et cetera, and there’s been plenty of discussion about how the Super League will accelerate the already-stark inequalities in global football by essentially giving all the Champions League money plus some (according to FT, maybe €4 billion, double what the UCL currently gets) to fifteen clubs instead of letting UEFA spread it around a bit.

All of these articles focus mainly on revenues, and not without reason; the numbers are astoundingly big. I’d argue it’s less about the size of the check, though, than about its consistency, and that even those who immediately recognize the American-style financialization the Super League would bring to European football are (this post from the Double Pivot’s Michael Caley aside) mostly missing the other angle: the Super League is about player wages.

https://twitter.com/MC_of_A/status/1383948286487187468?s=20

Top-level European soccer is a corrupt and unequal industry, yes, but there’s one sense in which it’s “fairer” than others: wages-to-turnover ratios, which any good Marxist should recognize as how capitalists talk about surplus value.

In most industries, owners are under significant pressure from competitors in the market to keep wages as low as possible, ideally at or below 15-30% of gross revenue — in other words, to let the workers keep 15-30% of the value that they one hundred percent create. Football clubs, whether owned by their members or by a locally-based businessman, are under a different kind of pressure: they want to win games. They want to bring home trophies and avoid being relegated. That means spending as much as the owners can afford on player wages. Top football clubs are not and have never really been money-making entities, and their wages-to-turnover ratios reflect this:

Deloitte, as a financial institution, unsurprisingly end their 2020 review of football finance by advocating for the simplest and most American of solutions to the problem, as they see it, of increasing wages in football: a hard salary cap.

With football currently on pause due to the impact of COVID-19, and losses set to spike due to the financial impact of this disruption, now is the opportune time to create more effective cost control regulations for the future and an overall regulatory framework to facilitate, monitor and enforce a more financially sustainable environment for Football League clubs.
Deloitte Annual Review of Football Finance 2020; “Put a cap on it”

Now, I’m not trying to make some contrarian point about how doing a Leeds is good, actually. Having football clubs be financially sustainable within some kind of overall regulatory framework is something I could get behind. But, regulated by whom, and in whose interest? If what we’re saying is, it’s bad when clubs drop down two divisions in two years and have to make a bunch of normal people redundant to cut costs, yes, I’d agree that’s bad. When Deloitte, JPMorgan, et al talk about “financial sustainability,” though, it’s clear they’re talking about something else entirely.

I alluded to this in my previous post about why the Philadelphia Union aren’t going to spend any money on an attacking midfielder, but it should go without saying that salary caps exist to benefit owners and, if structured favorably enough, guarantee profitability. MLS has a salary cap because the organizers didn’t want the league to go bust too quickly after the 1994 World Cup, which seems like an understandable concern on its face; the demise of the NASL was too recent to ignore.

Today, MLS is as secure as any other professional league, but its single-entity structure (see Fraser v. MLS) and incredibly restrictive and asinine salary cap rules let it maintain a wages-to-turnover ratio of just 28% in 2017, far lower than other American leagues and miles behind the 70% wages-to-turnover ratio UEFA deems to be sustainable for a well-run club. If the MLSPA has negotiators worth their salt, they’d have been hammering this point home at the last round of collective bargaining; not only is the salary cap effectively limiting the quality of the American league in a world without one, but also, MLS players are getting fucked, and unlike their counterparts in Europe, their club owners are starting to turn a profit.

The European Super League statement doesn’t mention a salary cap explicitly, but it does contain a telling reference to a “sustainable financial foundation with all Founding Clubs signing up to a spending framework,” and I would challenge anyone who doesn’t think this means a salary cap to answer: how could such a spending framework leave 70% of club spending untouched? Clubs can’t cut commercial costs, they can’t cut infrastructure, they can’t stop buying equipment, keeping up the stadium, paying the nutritionist and the physios; in business, the easiest thing to cut is labor costs, provided you can still attract and keep your most able workers, and hey, if all the other rich teams in the world sign up for a salary-capped league along with you, getting Messi to lower his wage demands won’t be such an issue, right?

It seems pretty clear that, despite the short-term focus on revenues, broadcast deals, and that big JPMorgan advance, the long game here is going to be (potentially in contravention of EU antitrust regulations?) to tamp down on wage spending as a percentage of revenue. Maybe player wages will stay static for a while and revenue will increase. Maybe the COVID salary cuts were the start of a downturn. At levels below the Super League, player wages will fall as advertisers and broadcasters leave the domestic leagues behind — besides, why should e.g. Leicester City or Olympique Lyon bother paying competitive wages to the James Maddisons or Memphis Depays of the world if they legally can’t make the Champions League?

Ultimately, even the highest-paid players stand to lose from any arrangement where their fifteen bosses get to decide between themselves how much money they can make, and that’s what their employers signed up for this weekend: a closed-shop league with guaranteed revenues and a limit on labor costs. The Super League says This far, but no further not to the rent-seekers in football, but to the players, and opponents of the plan can only hope the players recognize this fact before it’s too late.

How do you replace a player like Brenden Aaronson?

It’s a big question for Ernst Tanner, Jim Curtin, and the rest of the Philly front office going into 2021, especially as Aaronson has, by all accounts, been excellent since moving to Austria:

Reports out of the Union camp have fellow homegrown Anthony Fontana taking the reins as the number 10 in Curtin’s 4-4-2 diamond. Fontana hasn’t really played enough minutes for it to be worth putting up a radar (he had 1.09 non-penalty goals per 90 in 2020, which is chart-breakingly good but also came in only 513 minutes from 17 appearances, 12 of which were as a sub), but the eye test has him as less of a dribber / creator than Aaronson and more of an out-and-out goalscorer, an assessment that Curtin seems to agree with, having bumped Fontana up to striker for the Concacaf Champions League games against Saprissa.

I said this in a previous post, but I really don’t expect the Union to sign a 10 for big money in this window. The move they made instead sees German-American center mid Leon Flach join from St. Pauli in the 2. Bundesliga for a reported $300,000. The twenty-year-old looked solid in the CCL and will allow Jamiro Monteiro to, if needed, play further forward.

Still, it’s worth speculating (especially with the Union seemingly involved in the Frankie Amaya thing): if you were gonna bring anyone in before the start of the season, what would you be looking for?

First, here’s Aaronson’s 2020 stats, per fbref, with everything adjusted for MLS standard:

Three things stick out on his radar. First, and this is kind of a system thing, but Aaronson had the most pressures per 90 of any attacking midfielder in MLS last season (26.52) and was fittingly in the 97th percentile for successful pressures (6.85). Aaronson was also good for 2.69 tackles plus interceptions per 90 and 0.61 attacking third tackles per 90, both great levels. The Union will not be fitting in a Wayne-Rooney-at-DC-United type.

Second, Aaronson, while his expected assist numbers aren’t great, ranks highly for touches in the box (4.77/90), carries into penalty area (0.90/90), and goal-creating actions (0.43/90). He also takes good shots (0.14 xG/shot, 82nd percentile in MLS), which is something with which young wingers can struggle. This all speaks to an intelligent player who is good at finding space in the box.

Finally, while his usage stats are pretty low in general, Aaronson profiles as more of a dribbler than a passer; his dribbling numbers aren’t as high as I’d expected, but only 40.84 touches and 26.85 passes per 90 do track with what we saw from Aaronson last year — a player with a tendency to drift in and out of games, seeing relatively little of the ball for huge stretches before popping up with something magical.

So, to summarize: a like-for-like Aaronson replacement needs to be a defensively active creator who can get into the box and make things happen. Now (& this post is already longer than intended), I’ll briefly look at a couple guys who can do that. Under the age of 23, obviously.

From abroad:

Sorting a 2020/21 database from the big five European leagues for young attacking midfielders who can press already narrows down the list quite a bit; Aaronson would have been third in Europe for pressures per 90, for whatever fbref defines as a “MFFW” or “FWMF” born in or after 1998 with at least 450 minutes played (first: Jeong Woo-yeong, SC Freiburg, 29.6; second, Mikkel Damsgaard, Sampdoria, 29.5). Thirty-one players in that category registered at least 18 pressures per 90, shown here on a scatterplot against expected assists:

The catch here, though, is affordability. It’s easy to look at a chart like this and see that Mason Mount, Dani Olmo, Dejan Kuluzevski, and Jens Petter Hauge are good players. They collectively have a zero percent chance of wearing Union blue this season. But, there are a couple names here whose transfermarkt valuation is under €5 million: Emanuel Vignato (Bologna, € 4.5), Mehdi Zerkane (Bordeaux, €2m), Jeremy Le Douaron (Brest, €1m), Nassim Boujellab (Schalke, €1.2m), Dan Ndoye (Nice, €4m), and the aforementioned Jeong Woo-yeong, valued at €3m.

Jeong hasn’t played a ton of minutes in the Bundesliga this season (761 at time of writing), and a buy-back clause between Freiburg and former club Bayern Munich means Philly could be easily and immediately priced out of a move, but in a vacuum (read: in a FIFA / FM sense), he makes about as much sense as an Aaronson replacement as anyone in world football. Hell, I’d say, based on his statistical profile, that SC Freiburg — and soon-to-be former manager Adi Hutter’s new club Borussia Monchengladbach — could do a lot worse than snaffling Aaronson from RB Salzburg in a future window.

The opportunity buy: Matías Pellegrini

This move seems both a bit more out-of-left-field and also more possible. It seems Inter Miami cheated last year by signing four designated players and then saying ha ha no they hadn’t before eventually admitting that yes they had. My opinion on the DP rule and basically all of MLS’s other weird roster quirks is that they’re stupid and have outlived their usefulness. MLS teams won’t be properly good unless they get rid of the salary cap and DP rules and institute promotion & relegation — that is, until they give teams both the ability to spend money on better players and a good reason to do so. The league’s goal isn’t good soccer, though; it’s passable soccer and increasing club valuations, and the current model keeps costs down, so it’ll stay unless someone makes the league change it.

That said, 2020 expansion team Inter Miami are now, because of the DP rule and either their inability or unwillingness to count to three, in a position where they have to sell or cut one of their best players by tomorrow. Blaise Matuidi and Gonzalo Higuain are big names, and Rodolfo Pizarro is their best player, so Miami are apparently shopping twenty-year-old Argentinian winger Matias Pellegrini around the league for a trade. It’s worth mentioning that, with transfer windows closed in Europe and South America, there really aren’t a ton of potential destinations for Pellegrini. He’d have to come inside and play as a ten instead of a left winger, but I can’t recall any player ever complaining about that. He presses well, creates a lot of shots (particularly from throughballs), is a capable (72.2% successful) if not high-volume (1.69 attempted per 90) dribbler, and the best decade of his career is ahead of him. At the right price, he’s worth a punt.