A throw-in is the restart awarded when the ball crosses a touchline, and it is one of the most frequent events in any football match. Teams take dozens of them per game, yet for years most were treated as a routine way to put the ball back into play. Match data on RubiScore shows how often throw-ins occur and where, and a growing number of coaches now see them as a tactical opportunity rather than a pause.
The change has been driven by analysis. Because throw-ins happen so often, a small gain in how well a team uses them adds up over a season. Event data for the competitions covered can be explored at https://rubiscore.com. Below are six throw-in strategies seen in the modern game, with the data signature each one leaves.
Two features of the Laws of the Game make the throw-in unusual. There is no offside offence directly from a throw-in, so attackers can stand anywhere. And a goal cannot be scored directly from one, so a throw-in always needs a second touch to become a goal.
At the same time, a throw-in is a weak way to pass. The ball travels slowly, from the hands, and opponents can stand just two metres away. Analysts have long noted that teams lose possession surprisingly often in the first few seconds after a throw-in. That tension, a restart with special freedoms but poor accuracy, is why teams approach it in such different ways.
The long throw treats a throw-in near the opponent's penalty area as a set piece, similar to a corner. A player with a strong, flat throw sends the ball into the six-yard box, where tall team-mates attack it.
The tactic had faded from the top leagues for years, but it returned strongly in the Premier League in 2025-26, with clubs such as Brentford and Arsenal widely cited as leading examples. Opta analysts counted more goals from long throws into the box in 2024-25 than in any Premier League season since their records on this began in 2015-16, and the trend spread further the following season.
Why does it work? There is no offside, so attackers can crowd the goalkeeper. The ball arrives lower and flatter than many corners, and defenders get less time to organise.
Data signature: throws into the penalty area, first contacts won, shots and expected goals within a few seconds of the throw, and goals from throw-in situations.
The quick throw is the opposite approach. Instead of waiting for team-mates to take position, a player takes the throw immediately, catching the opponents before they reorganise.
It is most useful in transition. If a team wins a throw-in after a counter-attack or a defensive error, a fast restart can find a runner in space before the defence recovers. Ball boys and the placing of spare balls around the pitch can influence how often this is possible.
Data signature: a short time between the ball going out and the throw being taken, followed by forward passes or carries, and a high share of throws taken by whichever player is closest to the ball.
Many teams use the throw-in to keep the ball rather than to attack directly. A common pattern is throw, return and switch: the ball goes to a nearby player, comes straight back to the thrower or to a supporting full-back, and is then moved away from the crowded touchline.
Retention routines rely on numbers. Teams bring extra players towards the ball, including a centre-back or a midfielder, so the thrower has several short options. Some clubs have hired specialist throw-in coaches to drill these patterns. Liverpool's appointment of Thomas Grønnemark in 2018 is often cited as an early example in elite football.
Data signature: the share of throw-ins that lead to possession kept for a set time, the length of the sequences that follow, and the number of passes before the ball leaves the wide area.
Because offside does not apply directly from a throw-in, teams can use movement freely to create a free receiver. Players start close together and then split, or one player runs towards the thrower to drag a marker while another drops into the space behind.
These rotations are often rehearsed for throws in the middle third, where the aim is to escape pressure and turn quickly. A successful rotation lets a midfielder receive facing forward instead of with his back to goal.
Data signature: throws received by players in central positions, progressive passes or carries straight after the throw, and fewer duels on the first touch.
In their own half, some teams prefer to throw the ball forward along the touchline towards a striker or winger, accepting a contest for the ball. The aim is territory rather than control. Even if the first header is lost, the team can press the second ball higher up the pitch.
This approach suits direct teams and is common when a side is under pressure deep in its own half. It moves the game away from danger, at the cost of a lower chance of keeping possession.
Data signature: throw-ins that travel forward a long distance, aerial duels on the receiving end, and second-ball recoveries in the following seconds.
The simplest option is to throw the ball backwards to a defender and start the attack again from a calmer position. Teams that build patiently from the back often use this when the wide area is crowded.
There is one restriction worth knowing. A goalkeeper may not handle a ball thrown directly to him by a team-mate. If he does, the opponents are awarded an indirect free kick. So a reset throw usually goes to a centre-back or full-back, who passes on to the goalkeeper with his feet if needed.
Data signature: throws that travel backwards, low-risk passes in the following phase, and a switch of play to the other flank.
Consider two hypothetical teams that each take around 25 throw-ins per match.
Team A uses retention routines almost everywhere. Most of its throws go short, and it keeps the ball for several passes after a large share of them. It rarely creates shots directly from throw-ins, but it rarely gives the ball away near its own goal either.
Team B uses throws down the line in its own half and long throws in the final third. It loses the first contact more often, so its retention figure looks poor. Yet it wins many second balls high up the pitch and creates a steady number of shots from long throws into the box.
On a simple "possession kept" measure, Team A looks far better. On "shots created from throw-ins", Team B leads. Neither measure alone tells the full story, because the two teams are trying to do different things.
A few habits help make sense of the numbers:
The match timeline on RubiScore records the goals and key events that follow restarts, which helps connect a throw-in routine to the chance it produced.
The throw-in has moved from an afterthought to a planned part of the game. Teams now choose between attacking the box, restarting fast, keeping the ball, rotating to free a receiver, gaining territory or resetting. Each choice leaves a different trace in the data, and the most useful way to judge any of them is to follow the few seconds after the ball leaves the thrower's hands.