The Creator of xG: The Sam Green Story

In partnership with

Attacking Football

·29 de abril de 2025

The Creator of xG: The Sam Green Story

Nowadays, you hear it everywhere. On Match of the Day, in WhatsApp groups, from pundits trying to explain how a team came away with nothing despite battering the opposition. xG is part of the football conversation now. But for all the talk, most people couldn’t tell you where it came from, or who came up with it. Football has gone from saying that a player should have scored it, to saying that the shot had an xG of 0.56 within the matter of years. But where did xG come from?

That credit belongs to Sam Green. He’s not a household name, and he never tried to be. But his quiet work behind the scenes reshaped how we understand football. His story isn’t about chasing headlines. It’s about curiosity, data, and a desire to make sense of the game’s randomness.

OneFootball Videos

How It All Started

Back in April 2012, while working as an analyst at Opta, Sam Green published a blog post titled “Assessing the performance of Premier League goalscorers.” On the surface, it looked like any standard deep dive into striker efficiency. But buried in that post was a concept that would quietly reshape football analytics.

Green began with a simple question. If some players need twice as many shots to score the same number of goals as others, why is that? The answer, he suggested, lies in the quality of the chances, not just the quantity.

“There is an obvious correlation between the number of shots a player takes and the number of goals they will score,” Green wrote.

“However… shooting from a great distance or from a narrow angle may occasionally result in a goal, but missing the target or putting the ball into the opposition keeper’s arms will hand possession back.”

It wasn’t a new idea that shot location mattered. What Green did was build a model to quantify it. By analysing thousands of shots using Opta data, he assigned each attempt a probability of resulting in a goal. That value became what we now know as Expected Goals, or xG.

From Scattergun to Science

Green’s 2012 post explained the xG model in its earliest form. It wasn’t flashy, but it was clever. It considered not just distance and angle, but also whether it was a header, a through ball, a set-piece, or part of a counterattack.

“A header struck at the penalty spot from a corner will not have the same chance of being a goal as a shot from the same location on the counter attack,”

he wrote. That distinction, obvious to footballers and fans alike, was finally being translated into numbers.

He used the model to compare players like Darren Bent and Adel Taarabt. Bent, with fewer shots, had far higher-quality chances and a better goals-per-shot ratio. Taarabt, with a love for long-range efforts and set-pieces, often shot from poor positions. The difference was visible not just in output, but in the numbers. Bent had an xG of 9.9 from 43 shots. Taarabt had just 3.0 from 73.

Darren Bent Shot Map

Adel Taarabt Shot map

Green didn’t stop there. He broke xG into layers: the quality of the chance (location, context), the quality of the shot (trajectory, placement), and the impact of the goalkeeper. This is where his work was genuinely ahead of its time.

He introduced new metrics like Shooting Goals Added (SGA) and Keeping Goals Added (KGA), showing how players like Rafael van der Vaart improved their chances through elite shot execution, while goalkeepers like Tim Krul consistently saved attempts that would beat others.

Why It Took Time to Catch On

Despite the clarity of the concept, xG wasn’t an instant hit. Traditionalists scoffed. Even some data people weren’t sure how to use it. But slowly, the tide turned.

Betting firms adopted it first, seeing how it could reveal the truth beneath misleading scorelines. Then came the data-savvy clubs. Sam Green moved from Opta to Aston Villa being employed as their Head of Research. He arrived on the recommendation of Ian Graham, who works in the same position at Liverpool. This move made him one of the early data scientists employed by a Premier League team, at a time when clubs were just beginning to recognize the value of analytics. (For context, Arsenal went so far as to acquire the entire analytics company StatDNA in 2014, integrating its staff and research into the club’s operations, a level of commitment to data that few other clubs matched at the time.)

“The plan was to work out the value of every action the player carries out in terms of scoring or preventing goals,” Green told The Athletic. “Say we’re looking at a midfielder… If he completes a pass sideways, that wouldn’t particularly affect his rating but winning a header from a set-piece, or making an interception in the final third, would count for more. We would then combine everything the player does in the context of the game to get a ‘goal value’. “I’d then work in collaboration with the scouts to cross-reference the players they filed positive reports on and also suggest other similar players as alternatives, based on my findings.”

A bit too Avant-Garde for Aston Villa

Green’s stint at Aston Villa (2014–2016) is often regarded as unsuccessful, though not through lack of effort or expertise on his part. The period when Green worked at Villa coincided with a turbulent time for the club. Villa were struggling in the Premier League, undergoing managerial changes and battling poor results. Despite adding a bright analytics mind to their staff, the club did not see a turnaround in performance. In fact, Aston Villa endured a dreadful 2015–16 season, finishing bottom of the league and suffering relegation from the Premier League.

This outcome inevitably cast a shadow on the impact a data scientist could have in the short term. It highlighted a reality in football: analytics is only as effective as the club’s willingness and ability to act on it. Villa’s use of data under Green was, in hindsight, too little and too late to arrest years of decline. As one analysis noted, Villa were in a “five-year slide towards relegation” due to a decade of mismanagement, a trajectory not easily reversed by a single analyst amid an unstable leadership environment.

Several credible commentators have reflected on why analytics didn’t immediately rescue Aston Villa. One issue was cultural. Football clubs in that era often had a “Proper Football Men” mentality that was resistant to data-driven insights. An innovative analyst like Sam Green might have found his advice overlooked by coaches who trusted traditional scouting and intuition over models and statistics. Indeed, data was sometimes met with open skepticism; for example, when other clubs employing analytics struggled, the media narrative often blamed the reliance on numbers.

At Villa, manager turnover was high during his time there, Paul Lambert, Tim Sherwood, and Rémi Garde all took charge during Green’s stint at the club, ergo making it difficult to establish a consistent analytic philosophy.

It’s likely that any recommendations Green made (such as recruitment targets or tactical tweaks suggested by xG trends) were not fully implemented amid the chaos that was unfolding. As analytics expert Ted Knutson observed generally of that period, “if you stop one transfer mistake a season it pays for your analysis team for years”, but Villa continued to make mis-steps in the transfer market despite having a data scientist, implying that his input may not have been heeded in key decisions.

For Aston Villa at that time, this forward-looking, learning-oriented use of data never fully materialised. Instead, the club’s decision-makers largely operated in fire-fighting mode, and Green’s tenure ended without fanfare as the club was relegated and underwent a clear-out of staff.

The issue with analysis is that it can take years to see the impact of the change, and in football you rarely get that time. While his time would not be seen as a success, it would be unfair due to the length of time he was at the club.

The Lasting Impact of Green’s xG Model

Today, Expected Goals is part of every serious football conversation. It’s displayed on TV graphics, mentioned in commentary, and scrutinised in tactical briefings. But the ideas underpinning it haven’t changed much from what Green laid out in that 2012 blog.

“The work that Green and the team did back then has helped create new roles at clubs and a thirst for more data to try to gain competitive advantage,” Rob Bateman

Managers use xG to evaluate performance beyond the scoreline. Analysts use it to spot trends, like whether a striker is underperforming or simply unlucky. Recruitment teams rely on it to find players who get into good positions consistently, even if the goals haven’t flowed yet.

Even fans have adapted. Where once a striker missing two sitters would be slaughtered, now there’s nuance. At least he’s getting into the right areas.

And while modern models are more sophisticated, incorporating defensive pressure, passing sequences, and player tracking data, they are still built on the same foundation Sam Green laid over a decade ago.

The Quiet Legacy

Sam Green’s name doesn’t trend on Twitter. He’s not doing book tours or punditry. But there’s something poetic about that. Like the stat he created, he’s never been about the headlines. He was interested in the why behind the what, and he built a framework that helped football understand itself better.

“Green does not regard himself as the inventor of Expected Goals. He did not know, while he was building his system, of the work being done by StatDNA and Decision Technology, but he recognises now that they were all thinking along similar lines. There was a reason he said, that when he was out pitching Opta’s services to clubs, he did not spend much time in north London. Among his peers at clubs and consultancies, there were a handful of people developing metrics that did much the same thing, after all. His version, though, would be the one that caught the imagination, the one that popularised the idea, the one that would be presented to the Royal Statistical Society, the one that would, in time, germinate so broadly that it appeared on Match of the Day. Green was in the right place at the right time to give football what would turn out to be its breakthrough metric, the one that would take analytics if not into the mainstream, then certainly into one of its tributaries. Football has Green to thank for the – belated, eventual – arrival of xG.” Rory Smith on Sam Green’s impact.

In his own words, “The data explored above only touches on the ability we now possess to explore statistical trends in football.” He ended that blog by pointing out how data could explain storylines that might otherwise be chalked up to luck.

Over a decade later, that idea still rings true. And football, from Premier League benches to amateur scouts on laptops, is better for it.