Classify the present state of football for London clubs as uncertain, and you’re unlikely to get reasonable pushback. If it ended today, we’d have the first Premier League season since 1995/96 with Chelsea, Arsenal and Tottenham all outside of the top three. The pushback might come in the form of a supporter for another London club noting you’ve left them out of it. That may very well be a Brentford supporter claiming you’ve passed over, in this moment, at once the city’s most fascinating and uncertain team.
The level of paused uncertainty surrounding Brentford is befitting of a club in disarray, not a club on the verge of completing its most important season since the return of football after World War II had them relegated a top flight to which they haven’t returned. For a financially stable and unapologetically forward-thinking club that’s made smart inbound and outbound transfer moves in recent seasons, there’s a lot going on here:
- Because of the coronavirus lockdown, Brentford don’t know which ground they’ll be playing at next season. Last Saturday was supposed to mark the final scheduled match at Griffin Park (though a likely play-off tie would have changed that) with a move to the under-construction new stadium Brentford Community Stadium scheduled for July. That’s now on hold.
- Once they sort out their address, they still don’t know which division that address will host in 2020/21.
- From there, they don’t even know if the current season will be considered a season and whether they’ll have the opportunity to play for a Premier League spot. Sitting in fourth place is fine with a promotion play-off looming. Sitting in fourth place is not fine in a season uncertain to resume.
What they do know is regardless of whether this season is completed, it’s already been a good season. It’s certainly their best since their first back in the Championship in 2014/15, and a lot has happened with the club since then, even if it doesn’t quite show by a quick scan of Championship tables dating to their promotion.
It’s well known the club under chairman Matthew Benham has shown a willingness to do things differently. They in 2016 did away with their youth academy in favour of a B team that competes in no structured league. It was to some an anti-establishment move. To Brentford, it was yet another way for a smaller club to prosper in an evolving football world. Then, in the October of the 2018/19 campaign, manager Dean Smith left for Aston Villa. Brentford promoted Thomas Frank to the lead role. The irony here is Frank spent nearly two decades refining his craft as a youth coach, which, given the recent history of the club at youth level, sounds on the surface not unlike eliminating art from the curriculum and hiring Picasso to teach.
Frank’s methods have worked, and Brentford’s discovery process has proven lucrative.
The first full campaign under Frank started with two wins from seven, but it has evolved into one in which the (calcifying) table shows Brentford with the best goal difference in the Championship. The Bees’ +31 mark is more than twice that of neighbours Fulham (14), who sit third.
This comes after four seasons of mediocrity finishing between ninth and 11th and follows the sale of 25-goal scorer Neal Maupay to Brighton, Ezri Konsa to Aston Villa, and Romaine Sawyers to West Brom after sending Chris Mepham to Bournemouth in the January 2019 window. They also sold Scott Hogan mid-season in 2017.
Though it was a comparatively lavish summer for Brentford in terms of spending, the pounds spent on replacing those names – often in the Danish market Frank knows well – hardly matched the pounds coming in, which can be a frustrating realisation for a supporter of a recurrently mid-table club. Any frustration directed at club management last summer should now be met with trust.
It’s sometimes difficult to find, but the intersection of good business and good football is a beautiful place. So let’s get to the numbers before a decision is made with the season and someone at the EFL tells us to delete them.
There’s Possession, and Then There’s Consequential Possession
At a glance, Brentford’s style hasn’t changed a great deal, which isn’t surprising since Frank’s been at the club for since 2016 as an assistant and took over as manager in October of the 2018/19 season. They’re still a possession-based side – about 58% last season and this – yet their attack has taken them from 10th in xG per game last season (1.4) to third (1.6).
While they’re countering more, and their 6.4 counter attacks per match lead the Championship and are 40% more than league average, they’re not necessarily converting these into goals. They’re creating 1.9 shots per match from counter compared to the league average of 1.3, but it’s resulting in league average scoring involving counter attack (0.2 goals per match).
However, their scoring from build up, sustained threat and crossing is up. This makes sense because there’s a slight change in their build up and sustained threat frequencies. Last season, they were playing more build up than sustained threat, which is common – it’s easier to knock the ball around at midfield than it is on the edge of your opponent’s penalty area. This season, they’re playing with more sustained threat than build up, which is uncommon. Essentially, they’re playing with more consequential possession.
So what does this actually mean in terms of the specifics of their attack? On a team level, a lot. On an individual level, even more.
Brentford’s average possession is moving 23.9 metres up the pitch from start to finish, which ranks third, but that’s actually decreased half a metre from last season. Their average sequence length is 13.8 metres this season, which is second, but that’s only up from 13.7. So it’s not that they’re advancing the ball significantly farther up the pitch when they gain possession or begin a sequence. There are still meaningful discussion points from their sequences, but we’ll return to those in a defensive context.
The efficacy of how some of these possessions and sequences end, however, is definitely worth more time. As stated earlier, Brentford are producing more goals from crossing this season. In 2018/19, their 18.3% completion rate was 44th of 44 in the top two divisions. This season, it’s up to 22.3%, which is 22nd and 13th among Championship sides.
The quality of their 1-v-1 play has also improved. Brentford’s take-ons are down slightly this year from 18.6 per match to 17.6. But their take-on percentage is up from 48.8% to 54.5. In the Premier League and Championship combined, that’s up from 30th to 20th.
Their pass completion in the final third has increased modestly, their touches in the opposition’s penalty area have jumped (24.5 per 90 to 28.3), their expected assists have gone up (1 to 1.2), and ultimately their possessions and sequences resulting in a goal are up (1.5 to 1.7). To summarise: Brentford have seen a spike in creativity, which is resulting in more sustained threat and ultimately more goals coming from the possession-based style we’ve come to know them by.
Here’s the who of the why.
New Players, and Players Made New
Out: Maupay, Konsa, Sawyers, and Yoann Barbet, as well as Mepham six months before. Five players that had significant roles last season, though to varying degrees of success when considering their Possession Value, were sold over the summer.
It’s a bit intricate to state they’ve been replaced by better players across the starting XI. Rather, they’ve been replaced by a better team – and one that makes the most of the existing core of players Frank and his staff retained. But first, a few newcomers to the starting XI.
We’ll start modestly: From sustained threat to quality crossing to better 1-v-1 play, Mathias Jensen’s move to London after essentially a lost year in Spain has helped in all of the categories discussed in the team style section. The 24-year-old Danish central midfielder has succeeded in 73.8% of his 42 take-ons and completed 28.6% of his 140 crosses. Sawyers averaged 1.5 key passes per 90, while Jensen is at 2. Sawyers’ per-90 averages of involvements in sequences ending in a shot (5.2) or goal (0.4) were lower than Jensen’s (5.8 and 0.6). And Jensen’s sequences started that end in a goal (0.34) ranks second in the division among those with at least 900 minutes to some guy named Rooney.
Sawyers last season was and Jensen this season is a positive influence when measured with PV+, but they’re not quite at the level of one of Jensen’s teammates. That’s fellow midfielder and Arsenal youth product Josh Dasilva, who joined the club in the summer of 2018 but spent most of his first season on the bench.
In addition to being a first-team regular and a measurably positive influence on Brentford’s possession, the 21-year-old is completing 23.1% of his 78 crosses and winning two-thirds of his 66 take-ons to help improve those team numbers mentioned above. In front of goal, he’s been a dream. Dasilva has eight goals on a meagre 3.6 xG, so he’s drastically overperforming against what’s expected from the chances he’s had. If he can stretch that kind of overperformance into a second season and larger sample size, he’s going to be taken very seriously as a finisher.
About serious finishers: This has all been building up to someone else, and it’s not Ollie Watkins or Saïd Benrahma.
The addition of Bryan Mbeumo as a left-footed attacking player operating on the right side has been significant. At surface level, his productivity has been noticeable in the context of identifying an attacker with an above-average eye for goal. Mbeumo has scored 14 goals (none from the spot), and he’s done so with a 7.2 xG. His performance versus xG (goals minus expected goals) of +6.8 is easily the best in the division and ranks second among the big-five European leagues and the Championship behind only the current European Golden Boot leader:
For a 20-year-old, we might call that starting fast. But he did it last year in Ligue 2 for Troyes with 10 goals and a 5.9 xG mark. Add in his two games there at the start of this season with another goal, and we’ve got a 72-match span (read: legit sample size) with 25 goals and a 13.4 xG. This is prolific finishing. He’s almost doubling his xG. Granted, a two-season combination of Ligue 2 football and most of another in the Championship isn’t the Premier League or La Liga, but it’s something to pay attention to.
OK fine, it’s also about Ollie Watkins and Saïd Benrahma. Actually, it’s mostly about Ollie Watkins and Saïd Benrahma, but the second layer to Mbeumo coming into the side is how he might be impacting those players by balancing the pitch and allowing Watkins to leave the right side for good.
Saïd Benrahma 2018/19 v 2019/20 per 90
|Season||Touches||Passes||Final 3rd Passes|
|Passes Entering Final 3rd||Goals|
|Take-Ons||Open-Play Crosses||Chances Created|
On the surface, it may seem they’re getting a bit less on the left side from Benrahma this season than they did last. His goals and assists lag his xG and xA, which means he’s not finishing as expected and the opportunities he’s creating for others aren’t being finished as would be expected. Fault aside, that’s not ideal. This is a season after he outperformed his xG and those he provided for outperformed his xA. So it was the opposite. This 180 deserves more of a look, and what we’re going to find is Benrahma’s underlying attacking numbers are actually stronger this season.
Benrahma’s touches per 90 are similar to last season. His take-ons per match are consistent season to season (and obnoxiously high as the only Championship player at seven per match). His crosses from open play are nearly identical.
Oh but look here: His pass attempts have increased by more than six per match, his pass attempts in the final third have increased, his passing percentage in the final third is up, and his passes that enter the final third are up drastically. His chances created have dipped, yet his big chances created have actually slightly increased. What does this mean? It means there’s more quality here even if there’s less quantity.
In 2018/19, the origin of his passes weren’t saturating the final third the way they are now, and their destination is now more often in the penalty area.
His possessions lost per 90 are almost identical season to season, but his possessions lost in the final third are down, which goes back to the original point of the team sustaining threat at a higher level this season.
Refer back to the 2018/19 PV+ graphic above, and you’ll see Benrahma’s influence was positive. Refer back to this season’s graphic, and he’s the best on a better team. So while his own shots and primary assists might be resulting in less than ideal efficiency, his overall impact remains.
This is a tough numerical act for Ollie Watkins to follow.
With Maupay in the team last season, Watkins played more on the right. He and Sergi Canós weren’t great taking on defenders or linking up with teammates on crosses (though Canós was an important player in terms of PV+).
Canós has been out since October due to injury, but that follows a season after succeeding with 43.5% of take-ons, completing 14.9% of crosses and 61.5% of passes in the attacking third. He was giving the ball away more than anyone in the team with 39% of his touches resulting in lost possession.
Watkins is playing more of a centre forward role and has attempted 18 crosses all season from open play after trying 54 last season with limited success (11.1%). His take-ons per 90 are down from 2.8 to 1.4 with his percentage increasing from 39.2 to 52.
Mbeumo, too, isn’t taking on defenders or crossing regularly. Again, this team is more about quality than quantity (which in football ends up working out to plenty of quantity).
Less is more with Watkins. His touches are down from 55.1 to 38.6. It’s leaving him to do what he’s proven he’s good at – scoring goals. He’s doing it at a better rate than Maupay did a season ago. Maupay’s 25 came on 143 shots, which supplied him a 26.6 xG, which is to say most Championship strikers could have scored 25 from those chances. Watkins thus far has 22 off 96 shots with a 19.5 xG.
Roles seem to be better defined in this squad, and efficiencies are coming from it. From the left, crosses from open play have increased in frequency (4.8 to 6.5) and efficacy (11.1% to 17.2), and big chances created have improved (0.2 to 0.3). On the right, crosses from open play have dipped (7 to 5.9) along with take-ons (6.5 to 5.8), but the efficacy of each has increased and big chances created from that side have spiked (0.4 to 0.6).
And of course we’ve buried Brentford’s reworked defence at the bottom of this, but the reality is that’s the end of the pitch that’s seen the greatest improvement.
Getting Their Own House in Order
Most of what we’ve discussed so far has been in an offensive context, which ultimately led to modest increases in scoring and would matter a lot less if Frank hadn’t transformed Brentford into the best defensive side in the Championship.
Brentford Defence 2018/19 v 2019/20 per Game
|Season||Goals Against||xG Against||Opp. Touches in|
Ending in Final 3rd
Ending in Shot
Brentford’s xG against per game is a Championship-low with an actual rate second to Leeds (0.8). The similarity in their actual and expected goals against means it’s likely sustainable in the long term. The Bees were also sustainable last season, but sustainably bad. Going back to 2013/14, no Championship side has posted a per-game xG against rate under 0.9, so their defensive performances this season have been outstanding.
There are a lot of little things that have also improved. Opponent sequences ending in the final third are down, which is notable but not drastic. Opponent touches in their box are down, and sequences ending in a shot have dropped, which is second to Leeds (8.9).
In terms of their opponents’ style, they on average aren’t playing possession-based football, which isn’t surprising if you look at general possession percentages, so it’s worth looking at the transition phases. Their opponents’ counter attack percentage compared to the league average has increased from +1% to +10 this season, but they’ve gone from allowing a third of a goal per game on the counter to 0.1. And their opponents are pressing less. Last season, Brentford opponents pressed them at 41% above the league average. This season, despite Brentford having similar possession, that number is down to 18%. That’s supported by Brentford opponents seeing a drop in sequences starting in the attacking half with a defensive action (5.6 per match to 4.2).
In terms of winning the ball back or pressuring, Brentford’s own sequences starting in the attacking half with a defensive action are almost identical, which is in line with their own high press remaining below league average marks. But their total sequences starting in the middle third have increased from 20th (60) to sixth (63.3) while their sequences starting in the defending third have dropped from 68.8 to 63.6.
These are nice improvements to see, but it probably starts with something simpler. In 2018/19, Brentford’s duel percentage in the defensive half was a Championship-low 55.1%. This season, it’s 59%, which is fourth. Their aerial percentage in the defensive half has gone from last (49.6) to fifth (59.7). So in addition to starting more sequences in the middle of the pitch rather than their defensive third, they’re winning the ball more in their defensive half. Combine those two and you’ve got a lot less threat occurring near your own goal.
Aside from Mbeumo and Jensen, Frank also brought in 6-foot plus central defenders Pontus Jansson and Ethan Pinnock along with 6-foot-1 central midfielder Christian Nørgaard just in front of them. Jansson and Pinnock were joining Julian Jeanvier between Rico Henry on the left and Henrik Dalsgaard on the right while replacing the collective of Konsa, Barbet and for half of 2018/19 Mepham. The new duo have done admirably with duel percentages both over 60 and aerial percentages both over 64, which are significant numbers given the team-wide 2018/19 defensive half averages.
Out wide, the percentage of passes going forward has increased for both Henry (36.5% to 42.5) and Dalsgaard (37.4 to 44.3). With new players to go forward with all around Dalsgaard – Mbeumo in front and Nørgaard, Jensen and Dasilva in the middle – he has not only directed his attention forward, but while doing so is attempting more passes per match (42.3 to 52.2) while increasing his overall pass completion percentage (73.5% to 76.2) and his pass completion percentage in the attacking half (71.8 to 75). This has resulted in a step in the right direction with his PV+.
Nørgaard is most responsible for filling Kamohelo Mokotjo’s previous midfield role, which sits in behind Jensen and Dasilva. Within those defensive responsibilities, Nørgaard and Mokotjo are nearly identical with interceptions per 90, but Nørgaard is a more active and combative defensive player. He’s averaging 2.9 tackles, 11.1 duels and 2.8 aerials per 90, which are all significant increases from Mokotjo’s 1.7, 7.6 and 1.1 in 2018/19.
This all amounts to opponents’ big chances created dropping from 1.3 per match to 0.7. Their opponent xG per shot is 0.09. That’s the best mark in the Championship and the Premier League – as certain as it gets.