# Expected Assists (xA) · Football Level 8: Expert

Football · Level 8: Expert · Insporta Book (https://book.insporta.app/football/level-8/expected-assists-xa/)

Expected Assists (xA) in Football (Level 8: Expert) on Insporta: 100 sourced facts behind 10 tests. Covers: xa against assists; and more.

Expected assists and the numbers around them: key passes, chances created, shot- and goal-creating actions, crosses and cut-backs, and the assist records they are measured against.

## Test 1: xA against assists

- Expected assists value each chance a pass sets up, so the total reflects chance quality rather than whether team-mates scored. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- An assist needs a goal, so a creator with poor finishers around him looks worse than he is; xA credits the chance itself. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Assists above xA usually mean the shots after his passes went in more often than their quality predicts, which tends not to last. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- A shot-based model hands the xG of the resulting shot to the passer, so the pass is worth exactly what the shot is worth. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Assists swing on single finishes; xA moves with the chances he creates, which is steadier from one run of matches to the next. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- A 30-metre screamer is a low-xG shot, so the pass before it earns little xA even though it goes down as an assist. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Expected goal involvement is the sum of a player's own expected goals and the expected assists of his passes. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- A pass in his own half rarely leads to a shot at all, crosses from deep produce poor headers, and cut-backs give shots facing goal. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The pass itself is valued through the shot it leads to, which is why a pass that produces no shot earns no xA in such a model. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- No goal means no assist, but the chance was created, so the pass still adds its value to 8's xA. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))

## Test 2: Key passes and chances created

- Opta logs a key pass when the recipient shoots without scoring; if the shot goes in, the pass is an assist instead. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Every final pass to a shot is a chance created: an assist if it is scored, a key pass if it is not. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Key passes are final passes to shots that miss, are saved or are blocked, so none of them became an assist. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- A pass that sets up a 30-metre speculative shot and one that sets up a tap-in are one key pass each; xA tells them apart. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Big chances are judged by the data collectors, typically one-on-ones or close-range shots with little pressure. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- A delivery that finds a head for an attempt at goal is a key pass, so a regular corner taker collects many of them. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- More key passes from far out may only mean more low-value shots; the location and quality of those shots show whether he is creating more. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The closer the ball arrives to goal, the more often the move ends in an attempt. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Chances created are key passes plus assists, and a creator's passes that lead to shots are only a small share of all his passes. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The pass led directly to a shot that was not scored, which is exactly a key pass. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))

## Test 3: Shot- and goal-creating actions

- FBref credits the two offensive actions that directly lead to a shot, whoever makes them. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Take-ons, fouls drawn, shots that lead to another shot and defensive actions that lead to a shot all count. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The same two actions are counted, but only when the shot they lead to goes in. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- A successful take-on that leads to a shot is credited, so a player who beats defenders creates even when someone else makes the final pass. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- A shot that leads to another shot counts as a creating action, and here the second shot was scored. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The two actions before the header are the foul drawn and the free-kick delivery, and only the first is the winger's. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- It captures the different ways a player sets up shots without needing the shot to be scored. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The move runs forward in sequence: the left-back feeds midfield, the midfielder beats his man and then releases the striker, who scores. Read backwards from the goal, that pass and the take-on are the two goal-creating actions credited. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Up to two actions are credited per shot, key passes are one per unscored shot, and goals are rarer than shots. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The two offensive actions before 9's shot are the passes from 6 and 8; the shot itself is not one of them. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))

## Test 4: Assist records

- De Bruyne's 20 for Manchester City in 2019-20 matched Thierry Henry's 2002-03 total. (Source: Premier League, Bundesliga and La Liga records (Opta))
- Giggs made 162 Premier League assists for Manchester United between 1992 and 2014. (Source: Premier League, Bundesliga and La Liga records (Opta))
- De Bruyne had set 20 at Wolfsburg in 2014-15; Müller passed it on the last day of 2019-20. (Source: Premier League, Bundesliga and La Liga records (Opta))
- Assists are a statistic, not part of the Laws; systematic collection by data companies only came in the 1990s. (Source: Premier League, Bundesliga and La Liga records (Opta))
- Data providers credit the assist if the touch did not significantly alter the ball's direction or speed. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The fantasy game credits the player who won the penalty; Opta's assist needs a pass, cross or similar touch to the scorer. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Messi's assists for Barcelona between 2004 and 2021 are the most in La Liga history. (Source: Premier League, Bundesliga and La Liga records (Opta))
- Henry set his in 2002-03, De Bruyne his in 2014-15 and Müller his in 2019-20. (Source: Premier League, Bundesliga and La Liga records (Opta))
- Giggs made 162, Fàbregas 111, Rooney 103 and Lampard 102. (Source: Premier League, Bundesliga and La Liga records (Opta))
- Özil's 19 for Arsenal in 2015-16 fell one short of Henry's 20; De Bruyne equalled Henry in 2019-20. (Source: Premier League, Bundesliga and La Liga records (Opta))

## Test 5: Crosses and xA

- Many crosses meet a defender first, and those that reach an attacker usually give a contested header rather than a clean shot. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Every xG model rates a header below a shot with the foot from the same place, which is one reason crosses carry little xA. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The defence is retreating while the shooter arrives on the ball with his whole body facing the goal. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Volume is not quality: many crosses from deep produce few shots, and the shots they do produce are poor. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Alexander-Arnold beat his own mark of 12 from the season before; much of his output came from crosses and set pieces. (Source: Premier League, Bundesliga and La Liga records (Opta))
- A corner or wide free kick arrives into a box both teams have organised, so its value behaves differently from a cross in open play. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Teams that cannot get through the middle go round the outside, and the low-value crosses that follow show it. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The closer the delivery is played from the goal line, and the lower it is, the better the shot it tends to produce. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Getting to the line pulls the defence back, and the pull-back finds the space they have left in front of them. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The cut-back finds a runner facing goal near the penalty spot; the cross from deep usually ends in a contested header. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))

## Test 6: xA from set pieces

- A regular corner and free-kick taker collects xA from dead balls, which can flatter his creativity when the ball is moving. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- A direct free kick is a shot, not a pass, so there is no assisting pass for xA to value. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Many corners are cleared first, and the headers that follow are contested, so the average chance per corner is small. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Penalties are worth about three-quarters of a goal each and go to whoever takes them, so they are stripped out to compare open creativity. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- xA goes to the last pass before the shot, and here that is the cross after the short routine. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Event data records throw-ins as a type of pass, so one that finds a head for an attempt carries xA like any other. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Set-piece xA is the value of the shots dead-ball deliveries create, so a rise means those deliveries lead to better chances. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- A long-range free kick rarely goes in, a header from a corner more often, and a penalty is scored about three times in four. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Dividing by the number of corners turns a total into a rate, which is fair between takers who get different numbers of corners. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The corner is the pass that led to the shot, so its value as a chance goes to the taker. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))

## Test 7: Through balls and passes behind the line

- A pass that takes the receiver past the last line usually leaves him alone with the goalkeeper, one of the best chances in football. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Data providers tag a through ball when the pass goes between or past the defenders into space behind them. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- They are aimed at small gaps behind a defence, so defenders and goalkeepers claim a large share of them. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Offside stops play before any shot, so a shot-based xA model has nothing to credit. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- A line holding near halfway leaves forty metres behind it, which a well-timed pass and run can exploit. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Messi's chipped passes to Alba's runs from left-back became one of the most repeated patterns of Barcelona's attack. (Source: Match reports (BBC Sport, ESPN, club reports))
- The further the run, the more time defenders and the goalkeeper have to recover, so the shot that follows is usually worse. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The run has to be timed with the pass: moving too early is offside, too late and the space has closed. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Short passes are completed most often; long switches less so, and passes into the space behind a defence least of all. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The pass splits the away back four and sends 9 into the space behind them, which is a through ball. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))

## Test 8: xA and the finishing of team-mates

- When his passes led to good chances that were not taken, the assists stay low while the xA shows the chances were there. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- One season holds few chances per creator; over hundreds of them, lucky and unlucky finishes cancel out more and more. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The chances he creates are converted more often, so the same quality of passing produces more assists. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The passer decides the pass; whether the chance becomes a goal depends on the team-mate's shot and the goalkeeper. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The better side has better finishers and more of the ball, so assists reward the team around the creator as much as him. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Every missed big chance is also a lost assist for the player who made the pass. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- A gap that lasts for years is less likely to be luck; the usual explanation is the finishing quality around him. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The creator's gap is explained by the finishers' gap, which is where the check has to go. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The passer controls his own pass fully, affects the first touch partly, and has almost no say in the finish. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- A made 12 assists from 7 xA; that kind of overperformance usually shrinks, while B is due more and C is level. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))

## Test 9: xGChain and xGBuildup

- Every player who touches the ball in a possession that ends in a shot is credited with that shot's full xG. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Removing the key pass and the shot leaves the credit for the players who built the move earlier. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Lawrence proposed both measures to credit players whose passes start attacks rather than finish them. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Players who make the passes before the final ball earn nothing in xA; xGBuildup gives them a number. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Understat lists both measures for players in the big European leagues next to xG and xA. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- A ball-playing centre-back touches the ball in many possessions, but rarely makes the final pass. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- A simple sideways pass and the pass that broke the defence earn the same share, however different their value. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The per-90 step makes players with different minutes comparable. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- xA credits the one final pass, shot-creating actions the last two actions, and xGChain everybody in the possession. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- xGBuildup leaves out the shot by 9 and the key pass by 10, so the credit goes to the three players before them. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))

## Test 10: Reading an xA table

- A regular who plays every minute will pile up more xA than a substitute; per 90 puts them on the same footing. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- A few good passes in a handful of matches can produce an extreme rate that longer spells rarely sustain. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- A winger or number 10 is expected to set up shots; a holding midfielder's job gives him far fewer chances to. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Set pieces can dominate a taker's total, so an open-play figure shows how much he creates when the ball is moving. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Penalties, direct free kicks, shots after a dribble or a loose ball have xG but no pass to credit. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- The same total from fewer passes means each of the second player's passes set up a better shot. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Each provider trains its own model on its own data, so the same pass can be worth different amounts. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- Filtering and scaling come first; only then do positional comparisons and luck checks mean something. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- He completes dozens of passes a match, makes a couple of key passes, and those add up to a few tenths of an expected assist. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))
- D's rate comes from little more than three full matches, far too few to separate skill from a lucky run. (Source: Football analytics literature (Opta, StatsBomb, FBref definitions))

Sources: IFAB Laws of the Game / FIFA & UEFA technical reports / Opta, StatsBomb and FBref definitions / club and analyst publications / peer-reviewed research

Take the tests: https://insporta.app/try?utm_source=book&utm_medium=referral