Beyond the Scoreboard
We’ve all been there, watching our team dominate a game—more shots, more possession, more everything—only to lose 1-0 on a fluke counterattack. In those moments, the scoreboard feels like a liar. The most common reaction is to blame the coach for not
getting the result. But what if the coach's game plan was perfect, and the team was just unlucky? For decades, judging a coach was a notoriously tricky business, often boiling down to the crude metric of wins and losses. That record, however, is a noisy signal. It gets skewed by individual player errors, moments of individual brilliance, questionable refereeing, and pure, dumb luck. A better question isn't "Did they win?" but "Did the coach's system put the team in the best position to win?" To answer that, we need to look deeper.
Introducing Expected Goals (xG)
Enter Expected Goals, or xG. This isn't just another obscure number for stat-heads; it's a powerful lens for understanding what’s really happening on the field. In soccer, where it has become a staple of analysis, xG assigns a value to every single shot, representing the probability of it becoming a goal. This probability, on a scale from 0.0 to 1.0, is calculated by analyzing hundreds of thousands of historical shots. It considers factors like the shot's distance from the goal, the angle, the body part used (foot or head), and the type of play leading to it. A tap-in from the six-yard box might have an xG of 0.7 (a 70% chance of scoring), while a speculative 30-yard strike might be a 0.02 (a 2% chance). By adding up the xG value of every shot a team takes in a game, you get a total that represents the number of goals they should have scored based on the quality of the chances they created.
The Process vs. The Payoff
This is where xG reveals the true impact of a coach. A coach’s primary job is to create a tactical system that consistently generates high-quality scoring opportunities while simultaneously preventing the opponent from doing the same. In other words, a good coach builds a machine that churns out high xG for their team and allows low xG against. Whether the players actually convert those chances (the payoff) is often down to individual finishing skill or sheer luck. When you see a team with a high xG total but a low number of actual goals, it suggests the coach's process is working. They are tactically outmaneuvering opponents and creating great chances, but the players are failing to finish them. Over a long season, you can spot coaches whose teams consistently underperform their xG and might be better than their record suggests—they're just snake-bitten. Conversely, a team that constantly scores more than its xG suggests might have elite finishers, or they might be riding a lucky streak that is bound to end. This helps separate a genuinely good tactical system from a team that's just getting fortunate bounces.
A Tool, Not an Oracle
Of course, xG isn't a magic bullet. It has its limitations. Most models don't account for the specific player taking the shot—a chance has the same xG whether it falls to a clinical striker or a clumsy defender. It also doesn't capture the context of a game. A team sitting on a two-goal lead might concede low-quality chances willingly, skewing the numbers. And a brilliant dribble that beats three players but doesn't end in a shot records an xG of zero. For this reason, xG is most useful over a larger sample size, like half a season or more, where the noise of individual games and lucky bounces tends to even out. It's not meant to be the final word on any single game, but rather a guide to underlying trends. It complements watching the game; it doesn't replace it.
