Sep 17, 2020
Missing your Ex(pected Goals): Investigating Brighton’s Big Chance Enigma
Twenty-Eight. That’s the number of big chances Brighton have missed this calendar year, having failed to even test a statistically horrific Kepa on Monday night; Lewis Dunk headed wide in-front of an empty North Stand, whilst Neal Maupay and Aaron Connolly both failed to connect convincingly to crosses which, combined, generated 0.5 Expected Goals (xG).
Unsurprisingly, no side converted a lower proportion of their big chances last season than Graham Potter’s Brighton. An ugly conversion rate of marginally over 23% was undoubtedly responsible for the 7.22 fewer goals scored than ‘Expected’. Of non-relegated sides, only Everton (7.92) underachieved their xG by a larger amount.
Collectively, the Seagulls have scored a criminally low four big chances this calendar year. Crimes are always followed by an investigation, so, lets forensically analyse the theoretically easy-to-score chances which Brighton appear incapable of capitalising on.
Identifying the Culprits
Neal Maupay missed as many big chances last season as he scored goals (10), whilst Leandro Trossard (8 big chances missed), Aaron Connolly (5 big chances missed), Glenn Murray and Lewis Dunk (both missed 4 big chances) missed a cumulative 9 big chances more than the goals they produced as a group.
The Key Evidence
Now then, its time to retrace steps. Whilst all big chances are alike in their unanimous high xG values, most vary in what sort of big chance they are. This becomes clear when Brighton’s 2020 big chances are broken down into the method of shot assist:
· 8 aerial crosses
· 5 second balls from set pieces
· 4 low crosses
· 4 aerial through passes
· 2 ground through passes
· 2 cutbacks/passes across goal
· 2 set piece deliveries
What naturally stands out is Brighton’s modus operandi of creating big chances — aerial crosses. Now, neither Maupay nor Connolly are as incompetent from aerial deliveries as some would lead you to believe. At Brighton, Neal Maupay has actually been closer to clinical than wasteful as an aerial threat — his three goals from 11 headed shots is a better return than his 2.26xG, boasting an impressive 0.21xG per headed shot. Last season, he actually scored more headed goals than right-footed finishes.
Likewise, Aaron Connolly is capable of converting from these crosses, he just is yet to do so in the Premier League. In his final U23 season, he scored three times in four games from aerial crosses, heading home against Luton and Everton, whilst going airborne to score with his instep and complete a hat-trick against MK Dons.
Crosses themselves actually proved a primary source of goals for Brighton last season, so they themselves produce an enigma, as the risk of possession loss is high, but accurate crosses lead to high quality chances.
There’s Turbulence with Aerial Shots
That said, aerial shots are innately harder to score. Though the data from this article is 7 years old, it still holds analytical weight; the wonderful people at StatsBomb found that although headed shots were harder for goalkeepers to save than foot shots — 12% versus a 9% conversion rate — they were also harder to hit the target with. Nearly 57% of headers were off target shots, with that figure over 16% lower for foot shots, at 40.5%.
Brighton unfortunately support this trend, with just one-third (2/6) headed big chances hitting the target — Glenn Murray versus Everton, Neal Maupay versus Sheffield United (goal) — whereas 65% of foot shots (13/20) were goal-bound. They lead to more goals, too: Trossard versus Liverpool, Trossard versus Norwich, Maupay versus Arsenal.
Furthermore, StatsBomb found a key underlying reason for the higher rate of goal conversion from headed finishes — a higher proportion of these occurred closer to goal. The number of headed shots was over four-fold that of foot shots, with shots closer to goal naturally more likely to lead to goals, given the reduced distance giving goalkeepers less time to react and requiring less precise finishes, since the ball has less distance to travel.
Recency biases are definitely a thing, so Brighton missing their last 8 consecutive big chances has certainly accentuated the issue. Of the last 8, five were aerial crosses. It could well be that Brighton are not necessarily creating the wrong big chances, but creating in manners which do not create the scenarios whereby their forwards are naturally clinical.
For instance, of the ten big chances Neal Maupay has been involved in this year, just three have been struck with his left foot, with two of those three being second (and third) ball pop shots in injury time at Bournemouth. He scored half his goals with this ‘weaker’ foot last season, converting 5 times from 37 shots and, alike with headers, marginally outscoring his xG of 4.3. Maupay’s dexterity in his finishing is certainly a dangerous attribute of his, though Brighton seem to struggle to create chances for him this way.
Three of his big chances were created from the left-hand side of the pitch, worth noting again that two of those are actually from a Solly March corner from the left. The other, a waist-height Leandro Trossard cross against Liverpool, where Allison saved Maupay’s relatively tame effort, with the Frenchman only capable of connecting with his thigh. Instead, Brighton have primarily found Maupay with either aerial crosses from the right, or through defensive line-breaking passes from central areas. Again, the pattern remains: three through passes have created big chances for Maupay, the two which fell onto his right-foot (Crystal Palace at home, Bournemouth away) were both saved by Guaita and Ramsdale respectively, with the solitary slotted pass onto his left foot — from Aaron Connolly at home to Arsenal — being dispatched by Maupay into the North Stand goal.
Back then, to the acrophobia of Expected Goals. StatsBomb released a thoroughly informative article in July of this year, demonstrating the effect that shot height has on xG. What they found, was that despite height having a negligible effect on most shots, it did affect the probability of some. Now, contextually, InfoGol — whom the big chance data is taken from — use Opta statistics, whom do not include shot height in their Expected Goals calculations. StatsBomb determined that the height of the shot could reduce the xG by up to, or even more than, HALF. A big disclaimer; we do not know if the xG would change considerably for the Brighton big chances, though there a significantly clear volume of symmetry between their findings in 2013 and 2020, that aerial chances simply are harder to score.
Exemplary of this, Brighton scored twice from crosses post-lockdown. Leandro Trossard netted both, sweeping Aaron Mooy’s low cross at Carrow Road into the far corner, whilst striking Tariq Lamptey’s dipping cross against Liverpool just as it bounced up from the floor. This highlights the importance of striking a ball as close to the floor as possible; it is easier for a player to get up and over the ball, keeping their strike down.
Avoiding paralysis by analysis, key contributors to this would be that the trajectories of crosses can vary more, due to the combination of ball spin and air resistance. Furthermore, the friction of football on grass could well slow a ground cross down slightly moreso than an aerial one, with fine margins often the difference between off and on target, and on target and in target. Third and finally, the head is obviously a much more rounded surface than the foot, with less muscle groups able to be activated and less power able to be generated on a shot. To overcompensate, aerial crosses are usually delivered with power — you can see how this all contributes, right?
It gets worse, though. Unfortunately, just over a quarter of Brighton’s big chances were aerial crosses, so this alone does not answer the question as to why a painfully low 55% were on target. Now for a couple of contextual factors which complicate the problem for Potter more since they aren’t Brighton controllables — the effect of opposition on pressure and shot clarity.
Pressure makes Diamonds, but it might make Brighton miss chances
A 2018 StatsBomb article quantified the effect of opposition pressure on shot outcomes. Simplified, they found shots under pressure were converted at half the rate (7%) that shots under no pressure were converted at (14%). Contextually, pressure direction is important. Conversion rates dropped from 12% to 5% when the pressure was in-front of the opponent, with pressure from either side of the shooter dropping conversion from 10% to 6.5%.
Marrying this concept into Brighton’s big chances, then. The ‘pressure’ on the shooting player has purely been administered from a visual perspective, so this may not be as synergic with StatsBomb’s pressure model as it could be, but as a concept pressure is clearly identifiable through the eyes, so we can definitely draw conclusions.
20 (Twenty) of the big chances Brighton had were under some form of opposition pressure; 7 times pressure was from either side, 4 times from behind and on a big 9 occasions, from in-front. Of course, every single footballer and every single football team has to shoot under pressure at times, but it is interesting to see that Brighton are not frequently creating high xG chances in spaces which show an absence of defenders. As a point of reference, pre-lockdown, Brighton had conceded 38 1v1 chances.
Pressure impacts headed shots more significantly, too. Headers were found to be under pressure 10% more than foot shots. 5 of Brighton’s 6 headed shots were in the presence of pressure, with that percentagised at 83%, comparatively beyond the 75% of foot shots which were under pressure. Pressure puts players off, players can see, hear and even feel players in their peripheries, which may force them to rush their shot, or even just provide an additional distraction to prevent their full focus being transmitted to the shot.
The Accomplices — Deep Blocks and Unclear Shots
StatsPerform (Opta) found that shot clarity was really important, a concept which acknowledges the role of players ‘in the way’ of the shooter. High clarity shots are 1v1s, where there are no opponents between the shooter and goalkeeper. Moderate clarity sees up to one defender in the way, with low clarity shots any instances where two or more defenders are positioned between the shooter and the opposition ‘keeper.
To deviate from this calendar year slightly and time travel to the 2017–18 season, where Brighton used a deep-block under Chris Hughton to cement a 40-point total in their debut Premier League season. This deep defending block prioritised putting players between the opponent in-possession and the goal, with the intention of preventing high clarity shots.
As the graphic above shows, Brighton conceded more low clarity shows and less high clarity shots than the league average, even with a relatively low rate of pressure. Why then, is shot clarity important?
The answer: clarity is of similar significance to pressure. Players with a clear sight of goal convert a third of chances when not under pressure, outscoring the xG relatively considerably by 0.05, which shows just how desirable such shooting scenarios are. Simultaneously, though, the conversion rate of high clarity shots plummets under pressure, down to less than 20%.
Big chances are naturally more likely to be of higher clarity, as to get into zones close to goal you firstly have to bypass opponents — the people who can affect clarity. As football is evolving further, more and more teams — Brighton included — operate with higher defensive lines in order to offside trap opponents.
Over a third of Brighton’s shots saw opponents in the way — moderate and low clarity shots. Second balls from set pieces are primary conspirators, as naturally teams drop back more defenders when faced with opposition dead balls, in order to match the increased number of attackers they are faced with. Neal Maupay’s two big chances in injury time at Bournemouth proved this; no less than 9 Bournemouth players (Ramsdale included) were inside the 6-yard box. Unsurprisingly, both his shots were blocked. Although on paper — given the fact he is between the posts and around 6 yards from goal — the chance is a high quality one, the low clarity means he would require an incredibly precise finish to avoid hitting any Bournemouth players.
As Glenn Murray will tell you, and his record will prove, shooting when between the posts is the most desirable for any forward — it opens up a whole host of finish options which become increasingly less available the further wide you move, due to narrow shot angles, the limited nature of the human foot to produce certain shot trajectories and the ability of Goalkeepers to cover parts of their goal.
55% of Brighton’s big chances this year have been from between the posts. Of the 16 that were located there, the breakdown is as follows:
· 7 on target — 2 goals
· 6 off target
· 3 blocked
Of the 13 located wide of the posts, the breakdown is as follows:
· 9 on target — 2 goals
· 4 off target
Again, Brighton’s weird ability to transgress many analytical norms is somewhat impressive here, though admittedly the sample size is not huge, so it is easier for potentially anomalous results to have a bigger impact on overall trends.
The Jury Verdict — Is this a Death Sentence?
No. Probabilities indicate that this poor finishing from big chances is occurring at an unsustainably low rate, so Brighton should start to regress — which would ironically see them progress in goal-scoring — towards the mean. Big chances are categorised by InfoGol at beyond 35% goalscoring probability (0.35xG), so Brighton’s calendar conversion rate of 12.5% should start to slide up. To further supplement that, Brighton may look to try and create on Maupay’s left foot more, even if he is right-dominant. The inclusion of Solly March in pre-season and against Chelsea at left wing-back could be a key supplement for this; he has shown consistently good creation rates across three seasons in the Premier League and created a big chance for Connolly against Chelsea.
Crosses from the left side are more suited to left-foot finishes as they fall in-line with the natural striking technique for a left-footer, on either the inside or the laces.
As strange as it might sound, Brighton may look to avoid creating big chances which are too close to opposition goalkeepers; of the 10 big chances they created in the 6-yard box, just 1 was scored (Maupay vs Sheffield United) and 50% have been saved. Particularly when shots are taken from outside the post, being too close to the goalkeeper can allow them to use spread or smother techniques to block the ball.
From a shot clarity perspective, Brighton may opt for a lower-possession approach, just like they frequently did post-lockdown.
Giving opponents more of the ball will naturally see them commit more players further forward, thus meaning if Brighton win the ball they will have less opposition players to bypass and will favourably be attacking against an unorganised defence. Naturally, the inverse of this is something Brighton struggled with last season. High possession approaches will normally be met by opponents defending with their players behind the ball, typically blocking off either the midfield third, or defending deeper in the defensive third.
Brighton faced this issue against Sheffield United at the Amex, back in December 2019. That game marginally misses the cut-off date for the 2020 Big Chance sample, though that would not matter anyway, as Brighton created a shamefully low 0.53xG. The single biggest chance for the Seagulls was a 0.1 chance for Adam Webster in the depths of injury time — they didn’t come even remotely close to creating a big chance. Sheffield United defended very deep, and very well that day. Over 40% of Brighton’s possessions began with 8 or more Bladesmen behind the ball (goalkeeper excluded), whom set up with a 5–3–2 out-of-possession, so consider that to be the defensive and midfield units.
The 69% possession Brighton had in that game was the most of the ball they managed in an Amex game last season, but Wilder’s men managed to create 3 big chances and register 1.45xG, deservedly winning by a solitary goal-to-nil.
In terms of big chance creation last season, Brighton were way more efficient as a low-possession than a high-possession side.
Importantly, this is not to say that the responsibility of every possession is to score a goal, it can be equally as effective as a defensive tactic. However, it is refreshing to see Brighton still capable of operating as a counter-attack side. After all, the best footballing sides are multifaceted, so to be able to hurt teams in numerous ways can only be a good thing. Albion’s ability to score low probability chances is also very beneficial; their last 4 Premier League goals have a combined xG of 0.22, with two scored from outside the box.
The bottom line is that, despite the feeling that it will never end, all the numbers suggest it will. Brighton have sufficient quality in creation departments to ensure the big chances keep flowing, plus Neal Maupay and Aaron Connolly both definitely know where the back of the net is.
Twitter: @AlbionAnalytics
