xG be havin’ you! Do we Really Value the Process?
“We had the feeling we did many things right. Everything was fine. And then we concede a very strange goal, doing many mistakes in the build up to the(ir) goal…The decisions we take are the right decisions, it's about the executions…” These were the words of now England Football Manager Thomas Tuchel when his then Chelsea team lost to 4-2 Arsenal back in 2022. Reading that transcript suggests that Tuchel was happy with the process and performance and only due to some bad luck, they lost the match.
Hearing these comments, you would have expected he would have fielded a very similar team for the next game. Removing the outcome and emotion and defaulting to rationale and data. His reaction for the next game? Changing 40% of his outfield players.
I recently spoke at the Association of Sporting Directors about the realities of outcome and process. Football is a microcosm of sport, much of the modern discourse is centred around being “process-led.” Coaches and clubs alike now talk about identity, performance models and playing philosophies. The belief is that if the process is right—how a team trains, prepares and performs—then results will eventually take care of themselves.
But when the final whistle blows and the score shows a loss, that ideology is quickly tested. Because in sport, results still matter more than anything else.
To test this hypothesis, I dug a bit deeper to examine whether that are coaches are more process driven or do they use the outcome to dictate how they react in the next game.
I analysed 34 Premier League matches where a team won by a single goal despite being significantly outperformed in terms of Expected Goals (xG)—specifically, games where the losing team created at least around 1.0 xG more than the winner. In other words, did the team that was victorious but “got away with one” make more changes in the subsequent game in comparison to the team that had more and better chances to win but couldn’t convert and came away with the loss.
The outcome? On average:
Winning teams made 1.8 (16%) changes to their starting XI in the next match.
Losing teams made 3.4 (31%) changes.
Even when the losing team created better chances, arguably performed better and should have won the game, they were almost twice as likely to make personnel or tactical adjustments in their next outing. In comparison, the losing team, despite an arguable “unjust” victory made half as many changes.
This tells us something important about the way decisions are made in football (and probably sport): despite all the talk about process, outcomes still dictate behaviour.
What is xG and Why Use It?
Expected Goals (xG) is a statistical model that assigns a value to every shot based on the likelihood it results in a goal from 0.00 to 1.00. The model accounts for factors such as the distance from goal, angle of the shot, whether it was a header or a footed attempt, and the type of chance (open play, set piece, rebound etc.).
While xG doesn’t capture everything. It doesn’t factor in aspects such as defensive or offensive pressure, the quality of the pass before the shot or the skill level of the player. However, it remains the best publicly available metric to objectively assess the quality of chances created and conceded.
In my opinion, it doesn’t replace watching the game, but it offers an important layer of context. And crucially, over time, xG correlates strongly with results. That’s why it’s widely used across recruitment, performance analysis, and tactical planning.
The Case for Outcome-Based Decisions
Despite the growing emphasis on process, there are legitimate reasons why results continue to dominate decision-making:
1. Football is judged on results
League tables, contracts, job security—they all hinge on outcomes. No one gets rewarded for good xG if the team is losing games.
2. Emotional Investment and Control
Coaches are emotionally invested where a loss means more pressure, more questions and a perceived loss of control. To regain that sense of control, coaches can make changes to either feel like they are doing something or show people they are doing something.
3. Margins matter
Football is a low-scoring sport where games are often decided by fine margins. A side that consistently underperforms on the scoreboard may have deeper issues—finishing ability, game management, or psychological resilience—that xG doesn’t capture.
The Case for Process-Based Decisions
While outcomes are important, making decisions solely based on results can be equally dangerous:
1. Good processes produce results over time
Teams that consistently create better chances than their opponents usually win more often in the long run. Knee-jerk reactions to losses can disrupt tactical cohesion and player development.
2. xG helps isolate performance from luck
Sometimes the better team loses because of a deflection, a missed sitter, or an opposition goalkeeper having a brilliant game. xG helps coaches see through the chaos and identify whether a plan is working.
3. Avoiding overcorrection
If a team loses a game they should have won, making several changes might worsen performance. It can signal panic to the squad and undermine belief in the system.
What the Data Revealed
In the matches I analysed, a consistent pattern emerged:
Teams that lost despite clearly performing better than the opposite (i.e. should have won but lost) made significant changes to the line up in the next game.
Teams that won, who essentially “got away with one”, tended to stick with the same line-up.
This reflects a behavioural truth: the result—not the performance—still dictates the majority of short-term decisions, even at the highest level.
That’s not always wrong. But it raises the question: are teams truly process-led or are they simply using that language when results align?
Should You Prioritise Process or Outcome?
The answer depends on context. But one truth holds:
Outcome is the metric that everyone sees.
It defines public perception, internal pressure and professional consequences. Clubs, managers, and players are ultimately judged on results.
However, if you’re serious about being a process-driven organisation, you must start every review with performance. The result should shape the tone of the discussion, but not override the quality of what actually happened on the pitch.
Ask:
Did the performance justify the result?
If we played this match 10 times, would the outcome likely be the same?
If the process was poor, change is warranted. If the process was strong, resist overreacting—even in defeat.
Final Thought
Tuchel finished the interview saying “…there are a lot of positive things in there and…we don't add a note to it that overshadows that”. Clearly, the changing 40% of your starting line-up for the next game suggests the maybe we’re not as process-led as we think and inevitably the outcome is king.
In the short term, the scoreline shouts the loudest. But over time, it’s the quality of performance probably determines who stays at the top…Just hope that you have enough time to see the score match the performance!
