What Is Expected Threat (xT) in Football?

Expected threat, usually shortened to xT, is a football metric that puts a value on every location on the pitch according to how likely a team is to score soon after having the ball there. Instead of measuring shots, it measures danger — the threat a side generates simply by moving the ball into better positions.

Why Shots Alone Miss Most of the Game

Expected goals, or xG, changed how football is analysed by scoring the quality of every shot. But a shot is the final act of a possession, and most of what a team does never becomes one. The pass that splits a defence, the carry that drives forty yards upfield, the switch that shifts the point of attack — each creates the conditions for a shot without ever appearing in a shot-based metric.

Expected threat exists to value that build-up. It asks not how good a chance was, but how much danger every earlier action added on the way to it. In doing so it makes visible the part of the game that xG is blind to: the progression work that happens before, and often instead of, a shot.

The gap this fills is large. Across a single match a team might take somewhere around a dozen shots but complete several hundred passes and carries, which means the overwhelming majority of its work on the ball is progression rather than finishing. A metric that scores only shots therefore leaves most of what distinguishes one attack from another unmeasured. Expected threat was designed to put a number on that neglected majority — to reward a team for patiently working the ball into dangerous territory even on the possessions that break down before anyone shoots.

How Expected Threat Works

The idea behind xT is to treat the pitch as a grid of zones and give each zone a value. That value answers one question: if a team has the ball in this zone, how likely is it to lead to a goal within the next few actions? Zones near the opponent's goal carry high values, and zones near a team's own goal carry very low ones.

From there, every pass or carry is scored by the change in zone value it produces. Moving the ball from a low-value zone into a higher-value one adds expected threat, and the credit for the action is simply the difference between where the ball ended up and where it began. A backward pass into a safer area can even carry a small negative value.

Each zone's value is built from historical data on what actually tends to happen next from that area — a blend of two possibilities:

the chance of scoring directly from that zone by shooting, and
the chance of moving the ball to an even more dangerous zone and scoring from there.

Chained together across the whole grid, those probabilities produce a map of the pitch in which every location has a threat value, and every ball progression can be measured by how much threat it adds.

A simple way to picture it: a pass that carries the ball from just inside the opponent's half to the edge of the penalty area moves possession from a modestly threatening zone into a highly threatening one, so it earns a clear share of expected threat. A square pass between two centre-backs, by contrast, keeps the ball in a near-zero zone and adds almost nothing. The metric rewards the first and quietly overlooks the second, which is close to how a coach watching the same two passes would rank them. What xT does is turn that intuition into a consistent number that can be applied to every action in every match, rather than leaving it to the eye.

What the Metric Actually Rewards

Because xT is about moving the ball into danger, it rewards the players and actions that progress possession rather than only those that finish it. The kinds of actions that score well include:

Line-breaking passes — balls played through or past a line of defenders into more dangerous space.
Ball-carrying — driving runs that move possession upfield through open ground.
Switches of play — long diagonals that relocate the attack into space on the far side.
Passes into the box — deliveries that move the ball into the highest-value zones near goal.

The totals can be summed per player, to show who most advances the ball into threatening areas, or per team, to show which sides manufacture the most danger through open play. This is the real information gain of the metric: it separates a player who merely completes passes from one who completes passes that matter. Two midfielders can post near-identical pass-completion figures while contributing wildly different amounts of threat, because one recycles the ball sideways and the other keeps moving it forward into space. Pass-completion percentage, read on its own, would rate them as near equals; expected threat exposes the difference, and often it is the deep-lying player with the lower completion rate — the one attempting the harder, more valuable forward balls — who turns out to be doing the more important work.

How xT Differs From xG and xA

It helps to place expected threat next to the two metrics it is most often confused with:

xG (expected goals) values the quality of a shot — the chance that a given attempt is scored.
xA (expected assists) values a pass by the quality of the shot it directly sets up.
xT (expected threat) values any ball progression by how much it raises the chance of scoring soon, whether or not a shot ever follows.

Both xG and xA depend on a shot happening. Expected threat does not, which is exactly why it can credit the quiet, early build-up play that the other two never register. The three are complementary rather than competing: xG and xA describe the end of a move, and xT describes the journey. A complete picture of an attack uses all three, because a team can be strong in one and weak in another — full of threatening build-up that never produces good shots, or clinical from the few chances it manufactures without ever controlling territory.

The Limits of Expected Threat

Expected threat is a model, and like any model it simplifies. A few caveats are worth carrying whenever the number is read:

It is positional, not tactical: it values reaching a dangerous zone, but it cannot see the off-ball run that dragged a defender away to open that zone up.
Implementations differ: grid resolution, the underlying action model, and the data used all vary between versions, so two numbers both labelled "xT" are not always measuring quite the same thing.
It can flatter low-risk progression: safe forward passes into moderately better areas accumulate value steadily, which may overstate a cautious ball-mover.
It is descriptive, not predictive: a high xT total describes danger created over a period, not a result guaranteed on any given day.

Read carelessly, xT can be mistaken for a final verdict on a player. Read carefully, it is a lens on one specific thing — how effectively a team or player moves the ball toward goal.

How to Read an xT Number

The practical value of expected threat lies in comparison. A single xT figure means little in isolation and a great deal when set against peers in the same role: full-backs against full-backs, deep midfielders against deep midfielders. It is best read alongside xG rather than in place of it — xG for how good the chances were, xT for how the team arrived at them. A side can dominate expected threat and still lose, if it works the ball into dangerous areas all afternoon but finishes poorly; reading the two metrics together is what tells that story, where either one alone would mislead. The most useful question xT answers over a run of matches is a repeatable one: which players reliably move this team up the pitch, and which simply keep the ball moving on the spot.

Platforms such as RubiScore surface these possession-value contributions next to the more familiar shot-based numbers, so a follower can see both the finish and the build-up behind it. Used that way, xT answers a question shot-based metrics cannot: not who took the chances, but who created the danger in the first place. It is one of the clearest tools football analytics has for valuing the part of the game that happens before the shot — and the reason a quiet, progressive pass is no longer invisible in the data published on rubiscore.com.