Cracks In Philadelphia Union’s Defense Revealed By (Un)Expected Goalkeeping Data

By James Nalton.

For the previous five seasons, Philadelphia Union has been one of the more consistent teams in Major League Soccer, regularly finishing in the top three of the Eastern Conference and among the best teams in the overall MLS standings.

Some might say the Union under Jim Curtin has performed above expectations, but such consistency across a five-year period has been built on something substantial, not through luck or circumstance.

In turn, this has raised expectations in Philadelphia to a new level on the back of an obviously solid, successful game plan, and some more than solid players within that — a good number of them from the club’s own academy.

The Supporters’ Shield standings, which combine the Eastern and Western conferences to produce an overall MLS table, are one of the best judges of a team’s quality and consistency. In the five years between 2019 and 2023, the Union won one Supporters’ Shield title in 2020 and was unlucky not to have a second in 2022, when it was better than LAFC but lost out because MLS uses total wins rather than goal difference as its tiebreaker.

2024 has been a different story, though, and has seen a drop in performance levels from Philadelphia Union that will see its worst league finish for nine years.

The highest the Union can finish in the overall table in 2024 is 17th. The only season in team history in which it has finished lower in the Supporters’ Shield standings is 2015 when it finished 18th. Meanwhile, the Union has finished with fewer than 40 points — the maximum it can achieve in 2024 — on just three occasions (2010, 2012, 2015).

In some ways, it feels like Union has not played too badly in 2024. Daniel Gazdag is top ten in the league for goals scored and Kai Wagner is top ten for assists (per Opta). Only MVP candidate Luciano Acosta of FC Cincinnati has created more chances (107) in MLS in 2024 than Wagner (98).

At times, familiar traits seen in the previous five seasons of consistency and success have been evident, but most of the time, there have been glaring holes in this Union team that show an obvious need for improvements. It is also fair to question why the team waited for a season like this to happen, rather than building from a position of strength.

One of the reasons the Union is often said to be overperforming is its low salary spend compared to other top teams, so in theory, there has always been plenty of room for improvement, even when the team was already playing at a high level.

One of the Union’s biggest strengths during its successful period has been goalkeeper Andre Blake. Has his absence in 2024 been the reason for the team’s decline, or has it merely served to highlight underlying problems that needed highlighting?

Andre Blake and goalkeeper xG expectations

From the season in which the Union won the Supporters’ Shield in 2020, to 2023, Blake has been one of the best goalkeepers in MLS. It’s not a stretch to say the Jamaican is the best goalkeeper in the Concacaf region since the career of Costa Rican star Keylor Navas has begun to wind down.

The transfer of expected goals (xG) data from a statistic that measures the quality of a chance for the attacker, to one that can also effectively judge the level of difficulty a shot is for a goalkeeper to save, has been a very useful addition to the advanced stats landscape.

The xG-on-target stat (sometimes called post-shot xG) goes one further than xG and looks not at where the shot is taken from, but at where in the goal it is placed. From a goalkeeper’s perspective, it can be used to measure how difficult a shot is to save.

If the number of goals conceded is subtracted from the total xG-on-target a goalkeeper has faced, then you get a decent indication of how many goals they have prevented — how many goals were the shots faced expected to produce, versus how many they actually did.

For example, a shot from outside the area with an xG-on-target of 0.10 which results in a goal would be 0.10 minus 1, resulting in a negative of -0.90 against the goalkeeper.

A shot from closer to goal with an xG-on-target of 0.60 which is saved would add +0.60 in the goalkeeper’s favour. Across the season, the difference can then be used as a measure of goalkeeper shot-stopping quality.

There are numerous ways to measure a goalkeeper’s contribution in the modern game where, in many teams, they are required to be an additional outfield player in attacking build-up from the back and sweeping up in defence, but preventing goals is still a key measure of their contribution.

Statistically, 2021 and 2022 were Blake’s best seasons for the Union. In 2021 his xG-on-target minus goals conceded was +6.9 with an average of +0.27 per 90 minutes. In 2022 it was at +10 and +0.30 per 90.

For context and comparison, the best goalkeeper in MLS in 2024, Charlotte FC’s Kristijan Kahlina, has an xG-on-target minus goals conceded of +10 at an average of +0.30 per game, identical to Blake’s best season. In the six seasons FBref has been providing this data, only Đorđe Petrović (+13.1 for New England in 2022) and Stefan Frei (+15.2 for Seattle in 2018) have boasted better totals.

In the games Blake has played in 2024, his per 90 stat comes in at +0.12, with Oliver Semmle at +0.02. Semmle has been average but not bad while Blake has been decent, but not outstanding.

What it means

For any team, having a goalkeeper that saves shots they might not be expected to is a bonus. At the top level of the game and at the top level of goalkeeping you could say it is a requirement, but this should not be something a team is relying on for defensive success, and other areas of the team need to be functioning at a high level.

In 2024, it is clear that the Union’s defence has not been functioning at a high level or even at an average one.

With one game left to play, the Union has conceded 53 goals. The only seasons in which it has conceded more were 2015 and 2016, conceding 55 on both occasions.

At the other end of the field, 2024 has been one of the best seasons in Union history when it comes to creating and scoring chances.

The only year in which the Union scored more goals was that 2022 season when it finished level on points with LAFC at the top of the Supporters’ Shield standings. The Union had the better goal difference having scored 72 goals and conceded just 26, but the winner was decided on total wins giving LAFC the edge and the shield.

This season, the team has scored 61 goals, and despite its poor defensive record still has a goal difference of +8.

Returning to xG, but this time in an attacking sense, only LA Galaxy have a higher xG than the Union in 2024, and no team has created more “big chances” per the data.

Philadelphia Union second in MLS in 2024 for expected goals (xG), with 60.1. Their attack has created and scored plenty this season. pic.twitter.com/e2gtQ13FX7

— James Nalton (@JDNalton) October 14, 2024

So it’s clear that the attack has not been a problem and, in fact, it has been one of the best seasons the Union has experienced in this regard.

It is difficult to make assessments on what might have been due to varying game states producing varying returns in terms of xG and goal scoring, but if the Union had an outstanding goalkeeping performance similar to that of Charlotte this season or Blake’s best seasons, it would be pushing Columbus Crew and FC Cincinnati for those spots behind Inter Miami at the top of the table.

This is not to put any blame for the Union’s blip of a season on Semmle. The data shows his performance has not necessarily been bad, and it would be unreasonable to suggest all-time MLS goalkeeping performances every season.

If anything, the lack of an outstanding goalkeeping performance has highlighted the need for defensive improvement across the team.

There have been lapses in general play and at set pieces, as highlighted when Cucho Hernandez scored two headers from corners in the Union’s 3-2 defeat against Columbus Crew last week.

As Jonathan Tannenwald recently wrote in the Philadelphia Inquirer: “When you get beat twice in a game on textbook corner kicks, it hammers home that this season’s collapse can’t just be the backup goalkeepers’ fault.”

The positive is that the weaknesses in the team are fairly obvious. Having the services of one of the best goalkeepers in the league more often would help, but Blake’s absence has meant the areas that need strengthening are clearer.

It is good for the future of the team that these non-goalkeeper defensive issues have been highlighted, even though some might say they could have been predicted and addressed sooner.