The title race may not be decided by flair but by defensive margins: goalkeepers, set-pieces, pressing and defensive structure. Analyst Alan Morrison digs into what the numbers say about Celtic, Hearts and Rangers ahead of the final run-in.
Previously, I focused on how normal variance in football is having a significant effect on the league title race this season.
To recap: normal variance includes things managers cannot really control, such as your finishing and goalkeeper play, and the opposition’s finishing and goalkeeper play. Then add in the impact of own goals, penalties and red cards.
In Scotland, with Celtic performances more reflective of economic realities, such variance has had a relatively minor effect on the title in recent years due to Celtic’s dominance. It has significantly affected the relegation spots; for example, last season St Johnstone were severely hampered by negative variance, with Dundee in particular benefiting.
But this season, with leadership issues at Celtic, and a very shaky start from the new owners at Rangers, a Tony Bloom boosted Heart of Midlothian are genuine challengers. They have excelled at recruitment and on the pitch, and are also benefiting hugely from positive variance. As are Rangers to a lesser degree. Whilst Celtic’s poor outcomes are partly self-inflicted, failure to recruit adequate forward players and persisting with an ageing goalkeeper.
To build on that, we return to another hoary old football maxim. It is credited to Alex Ferguson that he once opined, “Attack wins you games, defence wins you titles.”
In the modern era, that sentiment may be challengeable, but let’s test the defensive mettle of the SPFL hopefuls.
I am going to focus on Celtic, Hearts and Rangers. Whilst Motherwell will have a huge impact on where the title ends up, being 13 points behind Hearts at this stage is likely too big a gap for them to be considered for the crown.
Goalkeeping
Whilst both Celtic and Rangers have persisted with ageing keepers, Hearts have had some rotation, but overall, it seems Alexander Schwolow is their first pick.
This shows the percentage of saves above or below expected each has saved, and the average goals per game against avoided thanks to keeper saves. In simple terms, what value is each keeper adding in the most fundamental task: saving the ball?
In this core respect, Hearts appear to have a slight advantage with Schwolow. He has saved more shots than expected and is avoiding around 0.2 goals per game thanks to his saves.
Kasper Schmeichel is almost exactly on the average for SPFL keepers. This is the level Joe Hart got to in the season he retired. If you are average at the SPFL level, but historically you have been considered one of the foremost keepers in the world, it may be a time to consider how far you want to fall. Just a thought.
Jack Butland has been feast or famine this season. Often being the difference maker, especially in early European ties, in a positive sense. Also, he has had several poor actions resulting in important goals being lost, none more so than at Tynecastle.
The averages are no respecters of highs and lows, and so far, he is the one from the three challengers below average.
This view speaks to positioning and overall value added. Positioning error is the average distance in yards from the perfect position the Statsbomb model says you are when facing a shot.
The goalkeeping On Ball Value (OBV) is an aggregate of all keeper actions expressed as an expected goals value added or subtracted from the team based on those actions.
So, this is more about decision-making overall.
In this view, Butland scores the highest with a positive OBV (rare in the SPFL) and the best positioning when facing shots in the league.
Schwolow’s positioning is not far behind, but his value is overall xG neutral when considering all his on-ball actions.
Schmeichel is, again, almost exactly on the league average for both. In such a tight title race, Celtic may reflect that they need more of a difference maker in the net.
Opposition Threat
Think of this as avoiding danger.
This view maps the average xG each team is generating in their shots against you, with the number of entries into your box.
So far, it is Rangers who are best at restricting the quality of the chances each team is generating and managing to keep the opposition out of their penalty box to the greatest extent.
This has been Danny Rohl’s major focus area on taking over – becoming defensively more solid, and he has been successful on a relative basis.
Hearts are not far behind. Derek McInnes avoids danger by getting the ball as far away from their own box as quickly as possible with direct passes from the back.
Celtic are only around the league average as far as limiting the opposition's opportunities is concerned. In previous seasons, this is a view by which Celtic would be an island at the virtuous top right of the chart. Injury to the two most effective defenders, plus an ageing defensive screen in Callum McGregor, all contribute to a decline in overall defensive team performances.
Goal Threat
We get to xG eventually! This plots both xG conceded in open play and the increasingly important xG conceded from set pieces.
Despite media focus and hype, Celtic are the third best side in the league in suppressing chances per game from set plays. Motherwell and Hearts eclipse them. Motherwell are a superbly coached team, aren’t they?
Despite the focus on overall defensive shape, Rangers are weakest in chance creation by the opposition, with below-average set-piece xG conceded, but the second-lowest xG conceded from open play.
Open play xG is more prevalent, and so Celtic, despite letting the opposition have the better field position, are the most effective at suppressing overall opposition chance quality.
Countering the Counter
With all teams (even Dundee against Celtic in the Scottish Cup) committing to a pressing game, the ability to defend chances created from counterattacks and high pressures has become a more important aspect of the game.
This is where what is called the “rest defence” is important. When you have the ball, how well set up are you to avoid danger when you lose the ball?
The top three are the most effective sides at suppressing both counterattack and high pressing shots. This is important as although low in volume, such shots tend to come with a high xG value as they are usually against a defence low in numbers and relatively disorganised.
Celtic are the most effective at avoiding shots from opposition high pressing, and we’ll see why that is below. Teams that avoid counterattack shots tend to be those that sit deeper, as Hibernian do, as opposed to having a high defensive line, like Celtic.
Both Hearts and Rangers are above average at avoiding high press shots but are average or below in supressing counterattacking shots. Both teams are more direct than Celtic and more wayward passes from the back can lead to such situations.
Squeeze It!
Considering more general defensive structure, which side is most effective at limiting the opposition in terms of field position and space and time to pass?
Celtic maintain by far the highest defensive line in the league. Whilst this serves to squeeze play into the opposition half and limits the space the opponents must play out, it can increase risk to counterattacks, although, as we see above, Celtic do smother teams quite effectively.
The passes/defensive action is the number of completed passes the opposition manages to make per your defensive actions, such as a tackle or an interception.
The lower the number, the more you are restricting high-volume passing moves. It is a proxy for pressing intensity and effectiveness. But also, most sides in the SPFL tend to play more long passes than Celtic and focus less on build-up through high-volume passing.
Nevertheless, Celtic are effective at limiting the opposition build up compared to the rest of the league.
Pressing
Finally, which teams are most effective at pressing disruption by regaining the ball?
The volume of regains from pressing and counter pressing is not possession adjusted and are averages per 90 minutes. This will serve to dilute the values for high possession sides like Celtic. However, this season, Celtic’s possession rate are more variable, especially under Martin O’Neill.
A feature of the season is the aggressive pressing game instilled at Hearts. Given their likely trajectory under the guidance of Tony Bloom, I suspect this is the last season for a while where they will not have European football.
Meaning, this season, they have largely been playing one game a week. Hence, they have been able to implement a hyper aggressive pressing and counter pressing game model, with some success as measured by regain rates.
Rangers are bang average by league standards. A reflection again on Rohl’s priority over shape and coverage versus being too front footed on pressing.
Celtic’s relatively low regain rates highlight their higher possession values, but also the strains of a game very three or four days, and a squad carrying many injuries.
Can this be maintained when stress increases against the better teams especially after the split?
Not all analysis leads to nice, neat conclusions with a bow on top!
Indeed, rigorous analysis rarely should.
What we see is a complex picture of defensive strengths and weaknesses across the three title challengers.
There are no clear favourites when considering defensive solidity.
I would classify both McInnes and Rohl as defensive-oriented managers by nature, whilst O’Neill is not lacking in pragmatism either.
The danger for all sides is the title slips away amongst a slew or draws – a fate most clearly facing Rangers I would argue given their creative paucity and focus on risk avoidance.
The winning team will need a strong defence, for sure, and a reliable goalkeeper. But who will dare to win?
Share