Expected Threat (xT) Explained: How Football Teams Create Danger

Expected Threat, usually shortened to xT, measures how much a football action increases the likelihood of a possession eventually producing a goal. It assigns a value to different areas of the pitch and rewards passes or carries that move the ball into more dangerous positions.

Unlike Expected Goals, xT does not wait for a shot to occur. It can identify the pass that breaks a defensive line, the carry that enters the final third or the switch of play that creates an advantageous attacking position. This makes it useful for evaluating progression and creativity earlier in an attacking sequence.

However, xT is not a complete measure of player or team quality. Its meaning depends on the model, the tactical context and what happens around the player in possession. It should therefore be treated as one part of a broader analytical process rather than a standalone answer.

What Does Expected Threat Mean In Football?

Expected Threat estimates the probability that a team will score from a particular location on the pitch, accounting for the possible ways a possession can develop from that position.

The central idea is straightforward: having the ball in some areas is more valuable than having it in others.

A team circulating possession between its centre-backs may control the ball without creating much immediate danger. Moving that possession into the space between the opposition midfield and defence usually increases its attacking potential. Reaching the penalty area increases it further.

An xT model attempts to quantify those differences.

If a player receives the ball in a zone with an xT value of 0.02 and passes it into a zone valued at 0.08, the action has added approximately 0.06 xT. In simplified terms, the pass has increased the possession’s expected threat by six percentage points.

The exact values vary between models. The important point is not the isolated number but the change in attacking potential produced by the action.

Why Was Expected Threat Developed?

Many traditional statistics describe what happened without adequately measuring its importance.

A player can complete 60 passes while mostly moving the ball sideways in safe areas. Another may complete only 25 passes but repeatedly break defensive lines and move possession towards goal. Pass completion alone does not distinguish effectively between those contributions.

Even progressive-pass definitions have limitations. A pass can qualify as progressive because it travels a specified distance towards goal, yet finish in a crowded or strategically unimportant area. Another, shorter pass may find a teammate between the lines and transform the attack.

Expected Threat was developed to attach value to ball progression according to the danger of the starting and finishing locations. It helps analysts move beyond counting actions and towards evaluating their likely attacking impact.

This reflects an important principle covered in our guide to the football statistics that actually matter: a statistic becomes useful when it captures something relevant to performance, survives contextual scrutiny and adds information beyond the scoreline.

How Is Expected Threat Calculated?

There is no single universal xT model. Different analysts and data providers can divide the pitch differently, use different datasets or make different assumptions. Most basic models, however, follow a similar process.

1. Divide the pitch into zones

The pitch is split into a grid. Each cell represents a distinct location from which the team in possession may pass, carry, shoot or lose the ball.

A coarse grid is easier to interpret but can miss subtle differences between nearby positions. A more detailed grid captures greater spatial variation but requires more data to produce reliable estimates.

2. Study what teams do from each zone

Using a large historical event dataset, the model estimates the likelihood of different actions occurring from each cell.

For example:

• How often does a player shoot from the zone?
• How often is the ball passed or carried elsewhere?
• Where do successful moves tend to finish?
• How often is possession lost?

3. Estimate the probability of scoring

If a shot is taken, the model considers the chance that it becomes a goal. If possession moves to another zone, it considers the future scoring potential of the new location.

This process is repeated across the grid until every zone receives an estimated threat value. Areas close to goal, especially central areas inside the penalty box, usually carry the highest values. Deep and wide zones usually carry much lower values.

4. Value each successful action

The value added by a pass or carry can then be expressed in simplified form as:

xT added = destination xT − starting xT

Suppose a midfielder receives possession in a deeper zone valued at 0.015 and completes a pass into a central attacking zone valued at 0.065:

0.065 − 0.015 = 0.050 xT added

If the player instead passes backwards into a zone valued at 0.010, the move reduces the immediate threat of the possession by 0.005.

That does not automatically make the backwards pass a poor decision. It may help the team escape pressure, retain possession or create a better opening later. The calculation describes the change in positional threat, not the total tactical quality of the decision.

A Practical Expected Threat Example

Imagine two teams each produce ten shots.

Team A reaches shooting positions through slow circulation followed by speculative attempts from outside the penalty area. Team B repeatedly progresses the ball behind the opposing midfield but sees several promising moves stopped before the final shot.

Shot totals treat the teams equally. Expected Goals should distinguish between the quality of the attempts they actually take, but it still says little about Team B’s dangerous possessions that did not end in shots.

Expected Threat can capture more of that earlier attacking work.

Consider one Team B possession:

1. A centre-back passes into the defensive midfielder.
2. The midfielder turns and finds a number ten between the lines.
3. The number ten carries the ball towards the penalty area.
4. A defender intercepts the attempted final pass.

No shot occurred, so the sequence generated no xG. Nevertheless, the line-breaking pass and central carry materially increased the danger of the possession. An xT model can recognise those contributions.

If similar progressions appear repeatedly, they may reveal attacking quality that a review based only on goals, shots or xG would overlook.

Expected Threat Versus Expected Goals

Expected Threat and Expected Goals measure related but different parts of attacking performance.

Our guide to Expected Goals explains how xG evaluates shot quality using factors such as location, angle and chance type. That makes it valuable for analysing the opportunities a team ultimately creates and concedes.

xT works further back in the attacking process. It asks how the ball reached dangerous territory and which players increased the likelihood of the possession producing a goal.

Neither metric replaces the other.

A team may accumulate high xT through repeated entries into threatening areas but fail to turn them into good shots. That could indicate poor final decisions, ineffective penalty-area movement or strong opposition defending. Conversely, a team may generate reasonable xG from relatively little territorial threat through counterattacks, set pieces or individual moments.

This is one reason xG is not enough on its own. Combining shot-based and possession-based metrics gives analysts a more complete view of how attacking danger is created.

What Actions Can xT Measure?

Most xT analysis focuses on successful passes and carries because they move possession from one pitch location to another.

Passes

A pass earns positive xT when it moves the ball into a zone with greater scoring potential. Line-breaking passes, cutbacks, switches into open space and passes into the penalty area can all add threat.

Long passes are not automatically more valuable. A short central pass into space between the lines may add more xT than a 40-metre pass towards the touchline.

Carries

A carry adds xT when a player advances the ball into a more threatening location while retaining possession. This can highlight dribblers, attacking full-backs and midfielders who move through pressure rather than relying exclusively on passing.

Carry value is especially helpful when comparing players whose progression styles differ. Two players may advance possession by similar amounts, but one does so primarily through passes and the other through ball-carrying.

Crosses and cutbacks

Crosses can add threat when they move the ball into a dangerous penalty-area zone. However, a location-based model may struggle to distinguish between a hopeful high cross and a controlled cutback to an unmarked attacker.

More advanced possession-value models can incorporate the pass type, defensive pressure and positions of teammates and opponents. A basic xT grid usually cannot.

Unsuccessful actions

Some models record only the positive value of successful actions. Others penalise turnovers according to where possession was lost and the threat surrendered.

This difference matters. A player attempting ambitious passes may generate substantial positive xT when they succeed but also lose possession frequently. Ranking players only by successful xT added may reward aggression without fully accounting for its cost.

What Does xT Reveal About Players?

Expected Threat can identify players who consistently move their teams into more dangerous positions, including contributors who receive limited credit from goals and assists.

It may reveal:

• Midfielders who break opposition lines.
• Full-backs who progress possession into advanced wide areas.
• Wingers who carry the ball towards the penalty area.
• Creative players who find dangerous central spaces.
• Defenders whose distribution starts valuable attacks.

Assists depend on a teammate taking a shot and scoring. xT can credit the progression even when the move does not produce a goal or immediate assist.

For example, a midfielder may regularly pass through pressure to release a winger. The winger then supplies the final pass. Conventional statistics credit the winger if a goal follows, while the midfielder’s earlier contribution may disappear from the record. xT can assign value to both actions.

However, total xT should not be used as a simple league table of creativity.

A player on a dominant team will usually have more possession, more opportunities to act and more teammates occupying dangerous spaces. A set-piece taker or primary ball-progressor may accumulate xT partly because of their assigned role. Minutes, possession share, territory, tactical responsibility and team strength all affect the result.

Useful comparisons may therefore include:

• xT per 90 minutes.
• xT per possession or per touch.
• xT from passes versus carries.
• xT by pitch zone.
• xT under pressure.
• xT retained after accounting for turnovers.

Even those adjustments require context. Per-touch metrics can favour high-risk specialists, while per-90 figures can favour players in possession-dominant systems.

What Does xT Reveal About Teams?

At team level, Expected Threat can help explain how a side creates attacking danger rather than simply how often it shoots.

Analysts can examine:

• Which areas generate the most progression.
• Whether attacks develop centrally or through wide areas.
• How effectively the team enters the final third.
• Which players connect build-up play to the attack.
• Whether dangerous possession becomes quality shots.
• Which defensive zones opponents exploit.

A team with strong xT but modest xG may be progressing well without converting territory into clear chances. That gap invites further investigation rather than an automatic conclusion.

Possible explanations include poor final passes, weak movement in the box, an overreliance on crosses, opponents defending the penalty area effectively or a temporary run of failed execution.

A team with limited possession but efficient xT generation may be highly dangerous in transition. Another may dominate territory and accumulate xT through sustained pressure. The totals may look similar even though the attacking mechanisms are entirely different.

This is why xT should be incorporated into a structured football match analysis framework rather than interpreted without tactical context.

How Tactical Roles Affect Expected Threat

A player’s xT output partly reflects the opportunities created by the team’s structure.

Consider two full-backs:

The first plays for a team that builds with three defenders and allows the full-back to receive high and wide. The second remains deeper to protect against counterattacks. The first will probably record more xT, even if both execute their roles equally well.

The same issue applies to midfielders. A deep midfielder may make the pass before the pass that produces the largest increase in threat. A number ten receiving between the lines may benefit from the difficult progression completed by a teammate behind them.

Football actions are interdependent. Space is created through movement, positioning and collective structure. An xT model generally assigns value to the player moving the ball, but not necessarily to the teammate whose run dragged a defender away.

Good analysis therefore asks two separate questions:

• Which player recorded the valuable action?
• What tactical conditions made that action possible?

This distinction prevents analysts from treating a model output as a complete account of individual performance.

Expected Threat And Defensive Analysis

xT can also be used to examine how teams prevent danger.

Rather than judging a defence only by goals or shots conceded, analysts can study where opponents are allowed to progress the ball and how much threat those actions generate.

A defence may suppress shots while still allowing repeated entries into valuable zones. That could suggest the current results depend on last-ditch interventions and may be difficult to sustain. Another team may concede possession deliberately but force opponents into low-value wide areas.

Potential defensive applications include:

• xT conceded per match.
• xT conceded per opposition possession.
• Threat allowed through central and wide areas.
• Threat conceded after turnovers.
• Threat prevented by interceptions or pressing.
• The relationship between xT conceded and xG conceded.

These measures can complement analysis of xGA, pressing, field tilt and penalty-area entries. They do not remove the need to assess opponent quality, game state or tactical intention.

How Game State Influences xT

Expected Threat totals are shaped by the score and match situation.

A team trailing late in a match often takes more risks, commits additional players forward and generates more attacking possession. Its xT may rise because the tactical incentives have changed, not necessarily because its underlying quality has improved.

The leading team may defend deeper and accept territorial pressure while protecting central space. It may produce little sustained xT but remain dangerous on the counterattack.

Red cards can distort the numbers further by changing space, possession and tactical responsibilities. Opponent quality, venue, fatigue and fixture congestion can also affect how easily a team progresses the ball.

This mirrors the broader challenge of analysing team form properly. Raw recent totals become more informative only after accounting for the conditions in which they were produced.

Whenever possible, analysts should split xT performance by:

• Level, winning and losing game states.
• Home and away matches.
• Equal-strength and red-card periods.
• Opponent quality.
• Open play and set pieces.

The Main Limitations Of Expected Threat

xT is useful because it simplifies a complicated attacking process. That simplification also creates limitations.

Different models produce different results

There is no definitive xT value for a pass. Results depend on pitch-grid dimensions, training data, action definitions and modelling choices.

Two providers can evaluate the same action differently without either calculation being obviously wrong. Comparisons should use a consistent source and methodology.

Basic models value locations, not the full situation

Two passes can start and finish in the same zones but have very different qualities.

One may find an unmarked teammate facing goal. The other may reach a player surrounded by defenders with their back to goal. A simple grid-based model may award both actions the same xT.

It may not know:

• How many defenders were nearby.
• Whether the receiver could control the ball cleanly.
• Which direction the receiver was facing.
• Whether teammates were making useful runs.
• How quickly the defence could recover.

Off-ball movement receives little credit

A run that creates space can be essential to an attack without touching the ball. Event-data xT models generally cannot reward it directly.

Tracking-data models can capture more of this information, but they are more complex and less widely available.

Volume can be mistaken for efficiency

Players on dominant teams naturally receive more opportunities to accumulate xT. High totals may reflect quality, role, possession volume or some combination of all three.

Negative actions may be handled inconsistently

A model that rewards successful progression without penalising failed attempts can overrate players who repeatedly surrender possession through ambitious actions.

xT does not guarantee a good shot

Moving the ball into a dangerous zone increases attacking potential. It does not ensure that the team will select the right final action or produce a high-quality attempt.

The correct response to these limitations is not to reject xT. It is to define what the metric measures, understand what it omits and combine it with other evidence.

How Expected Threat Can Inform Market Analysis

Betting markets ultimately express expectations through prices. Advanced statistics are useful only when they help an analyst assess whether those expectations are well supported.

xT can contribute by revealing attacking or defensive patterns that may not yet be obvious in goals and results. For example:

• A team may be progressing into dangerous areas consistently despite a poor scoring run.
• A side with strong results may be creating little territorial threat and relying on unusually efficient finishing.
• A defence may be conceding repeated high-threat entries despite allowing few recent goals.
• A key player’s absence may remove the team’s main source of ball progression.
• A tactical matchup may expose the area through which an opponent generates most of its threat.

These are analytical signals, not automatic betting conclusions.

The market may already account for them. The xT sample may be small. The opponent or game state may explain the apparent pattern. A team can also create territorial danger without possessing the final-third quality required to convert it.

An analyst must still translate the evidence into probabilities, compare those probabilities with the market and recognise uncertainty. Our guides to implied probability and value betting explain why identifying an interesting statistic is not the same as establishing that a price is wrong.

How To Use xT Responsibly

A disciplined approach begins with a question rather than a league table of xT totals.

For team analysis, ask:

• Where is the threat being created?
• Which actions and players generate it?
• Does it lead to good shots?
• Is the pattern stable across opponents and game states?
• How might the next opponent restrict or expose it?

For player analysis, ask:

• How much of the output comes from passes and carries?
• Is the player efficient or simply heavily involved?
• What tactical role gives the player these opportunities?
• How often does the player lose possession?
• Would the contribution transfer to a different system?

For market analysis, ask:

• Does xT add information not already visible in xG, results and prices?
• Is the difference large and persistent enough to matter?
• Could team news or tactical changes disrupt the historical pattern?
• Has the market already adjusted?

This approach treats Expected Threat as evidence to investigate rather than a shortcut to certainty.

Key Takeaways

• Expected Threat measures how much a pass or carry changes the future scoring potential of a possession.
• xT values the process before the shot, while xG evaluates the quality of shots that are actually taken.
• The metric can identify line-breaking passers, progressive carriers and teams that consistently reach dangerous areas.
• High xT does not guarantee high-quality chances, goals or future results.
• Model design, tactical role, possession volume, opponent strength and game state all affect xT output.
• Basic xT models may overlook defensive pressure, off-ball movement and the quality of the receiving situation.
• xT is most informative when combined with xG, tactical analysis, team news and market-implied probabilities.
• The purpose is to improve understanding of how danger is created—not to manufacture certainty from a single statistic.

Related Guides

• What Football Statistics Actually Matter?
• What Is Expected Goals (xG)?
• xG Is Not Enough
• How To Analyse Team Form Properly

Think Like An Analyst

Join the GoalIQAI newsletter for evidence-based football analysis, explanations of advanced metrics and a clearer view of how data, tactics and market expectations fit together.

Subscribe to GoalIQAI