Why Expected Threat (xT) is the smartest metric you aren't using

What is expected threat football exactly?

Expected threat, or xT, is a way of putting a number on how much a specific action—like a pass or a dribble—increases a team's chances of scoring a goal in the near future. It looks at where the ball starts and where it ends up, assigning a value to that movement based on how much more 'threatening' the new position is compared to the old one.

Think about a standard game. We’ve all seen that one midfielder who absolutely runs the show but never seems to get the goal or even the final assist. Traditional stats like xG or basic assist counts make them look invisible. This is where xT comes in. It recognizes that moving the ball from the halfway line to the edge of the box is a massive contribution, even if the person who receives it then fumbles the cross.

The model usually works by dividing the pitch into a grid, sort of like a chessboard. Every square on that grid has a 'threat' value based on historical data from thousands of matches. If you move the ball from a low-value square (say, your own half) to a high-value square (the 'Zone 14' area just outside the opponent's box), you've generated xT. It’s about the journey, not just the destination.

Why do we need xT if we already have xG?

We need xT because expected goals (xG) only tells us about the very end of an attack, whereas xT explains how we got there in the first place. If xG is the grade you get on your final exam, xT is the credit you get for all the heavy lifting you did during the semester to prepare for it.

The problem with relying solely on xG is that it’s incredibly narrow. It only cares about the moment the trigger is pulled. But what about the winger who beats three men and fires a ball into the six-yard box? Or the deep-lying playmaker who hits a forty-yard diagonal to open up the defense? If the striker misses the ball entirely, the xG for that move is zero. That feels wrong, doesn't it? The winger did their job perfectly. Expected threat football fixes this by rewarding the action of getting the ball into that dangerous spot, regardless of whether a shot actually happens.

By using data from providers like Opta or Stats Perform, analysts can see who is consistently putting the opposition under pressure. It moves the conversation away from just 'who scored' to 'who is making us dangerous'.

How is the pitch-grid actually calculated?

The grid system calculates the probability of a goal being scored from any given square within the next few actions. Analysts look at two main things for every zone: the chance of scoring directly from there and the chance of moving the ball to an even better zone.

Imagine a square right in the center circle. The chance of scoring a goal directly from there is almost zero. However, the chance of passing the ball from there into the penalty area is decent. Now imagine a square on the corner of the six-yard box. The chance of scoring is high, and the chance of a successful pass into an even better spot is also high. The 'threat' value of the second square is much higher.

• Passing xT: The value gained by successfully passing the ball into a more dangerous zone.
• Dribbling xT: The value gained by carrying the ball forward into a higher-probability square.
• Negative xT: Yes, you can lose threat by passing backwards or sideways into 'safer', less dangerous areas.

This allows us to see who the real progressors are. You'll often find that full-backs or wing-backs rank surprisingly high in xT because they spend their whole afternoon moving the ball from the middle third into the final third. They are the engines of the attack, even if they aren't the ones finishing the move.

What are the limitations of expected threat?

The main limitation of xT is that it generally doesn't account for the position of defenders or the specific context of a match. It treats every pass from Square A to Square B the same, whether there are ten defenders in the way or none at all.

It’s a 'location-based' model. It assumes that if you get the ball into the box, that's a good thing. Most of the time, it is. But we've all seen those desperate, hopeful crosses into a crowded area that have zero chance of finding a teammate. A basic xT model might still give that player a little 'threat' credit just because the ball landed in a high-value zone. It’s not perfect, but it’s a lot better than just counting successful passes.

Another thing to keep in mind is that xT favors players who take risks. A player who constantly tries difficult through-balls might have a lower pass completion rate on sites like FBref or Understat, but their xT will be much higher than a 'safe' midfielder who only passes sideways. It’s a metric that loves ambition. If you're looking to find the next creative gem, this is usually where you start looking.