Opta Analyst's David Segar analyses whether Premier League teams have become more direct in their play.
Football tends to go in cycles.
It seems like only yesterday that we were talking about playing out from the back, an obsession with passing opponents into submission, and the effectiveness of high pressing.
However, even casual Premier League fans will have noticed a bit of a shift in the early weeks of the 2025/26 season.
It appears as if quite a few teams are adapting their methods this season, however, so we wanted to do some more digging to see how different things really are.
Of course, we’re only 50 games into a 380-game campaign, so this is a fairly small sample size and there is plenty of time for numbers to change. But as things stand, there are some noticeable differences between the football we’re seeing now and what we’ve had in the past year or two. Here are 10 things we spotted:
- Passing at a 15-year low
- Teams becoming more patient
- Teams passing less to get forward quicker
- More teams going long
- High turnovers falling
- Fast breaks slowing down
- Set-piece goals increasing
- Short corners decreasing but threat increasing
- Long goal-kicks on the rise
- More long throws per game
Passing at a 15-year low
In recent years, passing seems to have become more intricate than ever, with teams trying to work their way beyond the opposition’s press or through a low block.
It has generally led to an increase in passes. On record (since 2003/04), the Premier League season with the most attempted passes was 2020/21 (an average of 945.2 passes per game), when almost all games were played behind closed doors due to the COVID-19 pandemic. That was likely due to players feeling a bit less urgency without having fans roaring them on.
The average in the 2023/24 season came close to matching that (941.0), so teams were seemingly evolving in that direction even with fans in attendance.
However, that reduced to 893.4 passes per game last season, and in 2025/26, games are averaging just 849.1 passes. That is the lowest average since 2010/11 (820.1), so it seems teams are now going for a "less is more" approach in the main.
Passing stats per match since 2016/17
| Season | Passes per game | Accuracy % |
| 2016/17 | 881.7 | 78.9 |
|---|---|---|
| 2017/18 | 906.4 | 78.8 |
| 2018/19 | 917.5 | 79.6 |
| 2019/20 | 915.2 | 80.0 |
| 2020/21 | 945.2 | 81.5 |
| 2021/22 | 907.3 | 80.7 |
| 2022/23 | 905.7 | 81.3 |
| 2023/24 | 941.0 | 82.9 |
| 2024/25 | 893.5 | 83.7 |
| 2025/26 | 849.1 | 82.6 |
Teams becoming more patient
It doesn’t mean teams are necessarily playing more risky passes, though. Passing accuracy across the league is still at its third highest on record (82.6 per cent), perhaps suggesting teams are just being a bit more patient and slower in their build-up. We have seen quite a few examples of centre-backs happy to keep the ball for long periods until a passing option opens up or they’re pressed, which could partly explain these numbers.
Either way, teams are passing the ball less but not yet sacrificing overall accuracy, so does that mean passing sequences are getting shorter and more direct?
Teams passing less to get forward quicker
Taking data from the 17 ever-present teams in the Premier League since the start of last season, we can see that 11 are averaging fewer passes per sequence in the early stages of 2025/26 than they were in 2024/25.
Eleven of teams are also attacking upfield at a greater speed, while seven are doing both. These teams are getting the ball forward quicker than last season, and they are doing so in fewer passes, including Manchester United, Brighton & Hove Albion, Newcastle United, Brentford, Liverpool and Arsenal.
Premier League team style comparison
As you can see, Manchester City are one of the more interesting movers, going from 5.12 passes per sequence last season to 3.97 in the current campaign, while also increasing their direct speed upfield from 1.43 metres per second (m/s) to 1.61. A lot of that will have been impacted by their more direct approach against Arsenal on Sunday.
Not everyone is going more direct, though. Nottingham Forest, firstly under Nuno Espirito Santo and now with Ange Postecoglou, have increased their passes per sequence from the lowest in the league (2.88) to the second-most (4.20).
They have also gone from the fastest team in the league in terms of direct speed upfield (2.04m/s) to the joint-sixth slowest.
Brentford have become the most direct team in terms of passes per sequence (2.84), while Chelsea have been the most intricate (4.52).
AFC Bournemouth, Everton and Wolverhampton Wanderers have significantly slowed down compared with last season, with mixed results to say the least.
Wolves are the slowest, with direct speed upfield of just 1.46m/s, and are bottom of the table, while high-flying Crystal Palace are the fastest attackers (2.02m/s).
One of the more surprising things is that Spurs haven’t really changed much either in terms of directness, despite going from Postecoglou to Thomas Frank, formerly of the more direct Brentford. The only difference really is that they’re moving upfield a little quicker.
The only team who have basically maintained the same numbers as last season are Aston Villa. But with just three points, one goal and no victories to their name in the league, they are perhaps not the best advert for the status quo.
More teams going long
Moving back to passing, we wanted to see if teams are attempting more long balls this season. A long ball/pass is a pass that travels at least 32 metres.
We have certainly seen some sensational kick-offs where Premier League sides have launched their kick-off straight out for a throw-in deep inside the opponents' half.
Fewer passes this season may also mean less passing out from the back. After all, there have been an average of 99.7 long balls per game this season in the Premier League, up from 93.4 across the whole of last season.
In the 2023/24 campaign the average was 101.3, and it was 104.4 the season before that
For those interested, excluding goalkeepers, Bournemouth defender Marcos Senesi has played the most long balls this season in the Premier League, with 73 of his 279 attempted passes being classified as "long".
More long balls would likely lead to shorter passing sequences, so let’s look at that next.
In the 2024/25 season, passing sequences lasted for an average of 10.4 seconds, with an average progress upfield of 12.6 metres.
Passing sequences in the Premier League this season have lasted almost a full second less (9.6secs), while the average progress upfield has reduced by half a metre (12.1m).
In fact, average sequence progress is currently as short as it has been in any of the last 10 seasons. With sequences not getting teams quite as far, it could be because they are trying more direct routes of advancement.
High turnovers falling
Something that can disrupt passing sequences are high turnovers, which is when a team win possession in open play 40 metres or less from the opponents' goal.
We have certainly seen plenty of them in the era of high pressing, so could that be why we’re getting shorter passing sequences?
Possibly, but not because teams are successfully turning the ball over. It could be that the threat of them is playing a part, though, as we’re actually seeing significantly fewer high turnovers this season.
In 2023/24, there were an average of 16.7 high turnovers per game, the most in the last 10 Premier League seasons. That dipped to 14.6 last season, and is down at just 11.5 in 2025/26, the lowest in the last 10 seasons.
That’s not to say teams aren’t pressing any more or that they’re not doing it as effectively, but is likely a result of opponents being more willing or inclined to kick it long from the back, which would also explain the rise in long balls.
Man City made the most high turnovers last season (9.4 per game) but have only made the seventh most this season (6.4 per game), while Liverpool have gone from the seventh-most (8.1 per game) to the third-fewest (4.4 per game) in 2025/26. Again, though, that is likely due to opponents not giving them a chance to win the ball high and playing more long balls out from the back.
It’s also notable that, across the league, the percentage of shot-ending high turnovers has dipped too, going from 17 per cent in 2023/24, to 16.4 per cent last season, to 14.9 per cent now; so not only are there fewer high turnovers, but those we’re seeing are leading to shots less often too. It’s clearly less of a weapon than it was in the last two seasons at this early stage.
Premier League high turnovers
Fast breaks slowing down
There were two goals from fast breaks in the opening game of the season between Liverpool and Bournemouth. In fact, Antoine Semenyo's equaliser at Anfield was one of the most direct pieces of play you’re likely to see, as the winger ran from deep inside his own half to score. Would that set the tone for the new campaign?
Not especially. If anything, it was a continuation of last season, which had the highest total of fast breaks on record in the Premier League (since 2006/07) with 2.04 per game.
This season, that’s down to 1.56, though that’s still more than all but three Premier League campaigns. So, while it shows a reduction from 2024/25, it’s still a feature of the modern league. It could increase as the season develops, teams with new signings get more used to one another, and tactics evolve further.
Set-piece goals increasing
It hasn’t exactly been a season dominated by set-pieces, but they have certainly become a bigger feature than we’ve been used to in recent years.
Of the 124 goals so far in 2025/26, 31 of them have come from set-pieces (excluding penalties). That means exactly 25 per cent have been from corners, free-kicks or throw-ins.
Over the last 10 Premier League seasons, the highest percentage of goals coming from non-penalty set-pieces was 23.9 per cent in 2016/17. It was 20.6 per cent in 2024/25 and 19.8 per cent in 2023/24. Again, it is quite possibly something that will come down to a similar percentage as more games are played in 2025/26, but as things stand, we are seeing a greater proportion of goals coming from set-pieces.
Percentage of goals coming from non-penalty set-pieces
| Season | Set-piece goals % of all goals* |
| 2016/17 | 23.9 |
|---|---|
| 2017/18 | 21.9 |
| 2018/19 | 22.9 |
| 2019/20 | 22.0 |
| 2020/21 | 19.2 |
| 2021/22 | 21.8 |
| 2022/23 | 21.2 |
| 2023/24 | 19.8 |
| 2024/25 | 20.6 |
| 2025/26 | 25.0 |
*Excludes penalties
Short corners decreasing but threat increasing
Speaking of corner kicks, another slight change you may expect when teams are playing more directly is a reduction in short corners.
Of the 482 corners taken in the Premier League this season, only 65 have been taken short (13.5 per cent). That is a significant reduction from last season’s 18 per cent. In fact, it was steady around that figure for three seasons (17.7 per cent in 2022/23 and 17.8 per cent in 2023/24), with the last time it was anywhere near as low as this being in 2021/22, when just 14 per cent of corners were short.
It seems to be working, as we’re seeing on average 0.44 goals per game from corners this season, compared to 0.35 last season, though it was 0.45 in the campaign before that, so perhaps it’s simply due to our old friend, variance.
Long goal-kicks on the rise
At the other end of the pitch, as we recently touched upon, there have been more goal-kicks sent long this season. Premier League fans have become very familiar with the sight of goalkeepers doing everything they can to pass sideways to a centre-back from a goal-kick, but perhaps that too is a trend on the way out.
As recently as 2017/18, the ratio of goal-kicks that ended in the attacking half was over 75 per cent, but that fell all the way down to just over 40 per cent last season.
In 2025/26, for the first time in 10 years, the percentage has gone up, with 48.4 per cent of goal-kicks ending in the opposition’s half.
We should mention that it comes down to 44.4 per cent if you discount Burnley. Martin Dubravka’s 61 goal-kicks taken as long balls into the opposition’s half are more than twice as many as any other team.
More long throws per match
Last season, there were an average of 1.52 long throws per match (long throws being those that travel at least 20 metres and end in the opposition’s box), and that has more than doubled in 2025/26 with an average of 3.44 per match so far. Indeed, that’s more than double the rate of any other season in the previous 10 years, with 1.67 per match in 2018/19 being the highest rate before this campaign.
Long throws per match since 2020/21
| Season | Total |
| 2020/21 | 0.89 |
|---|---|
| 2021/22 | 1.27 |
| 2022/23 | 1.44 |
| 2023/24 | 1.47 |
| 2024/25 | 1.52 |
| 2025/26 | 3.44 |
West Ham United are the only team in the Premier League not to have attempted a long throw into the box yet this season, while Brentford lead the way with 23.
There have been four goals from long throws across this season's 50 games so far, an average of one every 12.5 games. That’s quite a way up on last season, in which a goal from a long throw was scored once every 27 games on average. That too was a huge increase on the season before (one in 76 games).
Long throw-ins are also producing more than twice as many Expected Goals (xG) this season (0.11xG per game) as any other campaign on record.
Watch teams using long throw-ins
Again, it is far too early in the campaign to be making sweeping statements or drawing conclusions, but we’ll see how things develop and likely revisit much of this later in the campaign.