The most impactful midfielders in the Scottish Premiership this season
written by: @TheGersReport
We have data!!!!!!!!!!!!!!!!!
As you likely know, Modern Fitba has formed a partnership with Ortec Sports to have access to the data they collect for Scottish Premiership matches. This means that we can apply our models & statistical categories to provide you, the lovely readers, with in-depth analysis that is often unique to any other outlet in Scotland.
So, let’s just get to some numbers. Thankfully, with the data that we now have…I can dust off some stats that I have put in the forefront of my work in the past.
In this post, I’m going to share some Shot Creation Passing Stats for the season & re-introduce Relative Shot Creation Passing Stats to see which midfielders are having the biggest impacts on their teams.
We’ll begin with the Key Passes per 90 leaders from the Scottish Premiership this season (note: Key Passes are the passes that set-up a shot):
Notice than more than half of the players on this list play for either Rangers or Celtic. This is part of the reason that I like to use Relative Shot Creation Passing stats when assessing a player’s impact. This is one way to add context to the numbers.
Relative Shot Creation Passing stats compares a player’s per 90 passing stats to the rest of the midfielders/wingers on the team. For example, Tony Andreu has actually had a bigger impact on the creation of shots for Hamilton than Ryan Christie has had for Celtic.
How is that possible when you don’t even see Andreu on the list above, while Christie averages the fourth most Key Passes per 90 in the league?
Tony Andreu, on loan from Coventry City, average 1.63 Key Passes per 90 which is nearly a full Key Pass per match less than Christie.
The catch is that it doesn’t seem fair to simply rank players based on their raw data when a team like Celtic has taken 278 MORE shots this season than Hamilton.
When Andreu joined Hamilton this January, he walked into a lineup that truly struggled to generate shots & he has gone on to average 1.63 Key Passes per 90 minutes, which represents 1.09 more Key Passes per 90 then the average output from the rest of the midfielders on team.
Compare that rate to Ryan Christie, who averages 0.81 more Key Passes per 90 than his Celtic peers (which is also very good).
Relative Key Pass rates, like the ones above, do not measure whether Tony Andreu is better then Ryan Christie (he’s not), however it highlights the fact that his passing has had a bigger impact for his club then his Celtic counterpart.
Below is a list of midfielders & wingers who have the best Relative Key Passes per 90 rates in the league this season (with a minimum of 600 minutes played).
- David Turnbull, Motherwell +1.53
- Olly Lee, Hearts +1.27
- Chris Burke, Kilmarnock +1.22
- Scott Wright, Dundee +1.17
- Niall McGinn, Aberdeen +1.09
- Tony Andreu, Hamilton +1.09
- Stevie Mallan, Hibs +1.05
- Kyle Magennis, St Mirren +0.87
- Cammy Smith, St Mirren +0.81
- Ryan Christie, Celtic +0.81
- Daryl Horgan, Hibs +0.79
- Daniel Candeias, Rangers +0.63
- David Wotherspoon, St Johnstone +0.60
- Steve Lawson, Livingston +0.58
- Matthew Kennedy, St Johnstone +0.55
- Glenn Middleton, Rangers +0.41
- Steven Lawless, Livingston +0.41
- Ryan Kent, Rangers +0.39
- Scott Arfield, Rangers +0.31
- Gary Mackey-Steven, Aberdeen 0.29
- Connor McLennan, Aberdeen+0.28
I wrote this about David Turnbull back in July:
I can’t overstate how big of an impact David Turnbull has had & while his goals get a lot of the headlines…it’s his playmaking influence that really seems like an indicator that this is a budding superstar. No other midfielder in the league averages as many Key Passes per 90 then Turnbull…on a team that struggles to generate shots (Motherwell rank 8th in the league for total shots).
And…he’s only 19.
I also talked about Scott Wright in that same blog post that featured Turnbull. He was a player whose play was being criticized by his own manager for not making enough an impact with the opportunities he had been given. The irony was that the numbers highlighted the kind of impact that the ‘eye test’ doesn’t always see.
Wright only played a little over 400 minutes this season for Aberdeen before being shipped out on loan to Dundee. Last season he averaged 1.95 Key Passes per 90, this season…it’s 2.10. Not only is he among the very best when it comes to his Relative Key Pass rate…his actual per 90 rate is better than both James Forrest & Daniel Candeias. It’s time to embrace the fact that Scott Wright can be a game changer at this level & is just now entering the prime years of his career.
Hi, Glenn Middleton! I see you…not sure your manager still does….but you keep doing you!
Before we move on…let’s have some fun (or dread…depending on your loyalties) & look at which midfielders are having the least impact on their team when it comes to Key Passes. Below are the players with the worst Relative Key Pass rates in the league:
- Scott Brown, Celtic -1.14
- Marvin Bartley, Hibs -1.10
- Ross McCrorie, Rangers -1.07
- Eros Grezda, Rangers -0.86
- Alex Gorrin, Motherwell -0.82
- Danny Swanson, St Johnstone -0.80
- Allan Campbell, Motherwell -0.79
For a few of these players like Brown, Bartley, McCrorie, & Campbell it’s really no surprise to see them here given that they do more work earlier in the build-up play. Meanwhile, it’s safe to say that Grezda & Swanson (whom the numbers really liked last season) have been pretty big disappointments for their fan bases. However, in the case of Eros Grezda…stay tuned for the next stat because…well - you’ll see.
When looking to find impactful playmakers, you also need to look beyond Key Passes. Often, the pass that truly changes the build-up is the Secondary Shot Assist (the pass before the Key Pass). Sometimes this is the pass that gets the ball out of a tight situation & suddenly puts the defence on their heels, or other times it’s the visionary pass that gets defenders chasing the ball which ends up creating openings for teammates to finish off the play.
The opening goal in the example below is one I’ve used often to highlight what a Secondary Shot Assist can do:
The nice thing about keeping track of shot assists like this is the fact Candeias still gets credit for creating the play…regardless of whether Morelos scored or not. Last season, Josh Windass was tenth in the league in Secondary Shot Assists but not a single one led to an actual goal. When only honing in on assists, you truly are limiting the scope of player analysis.
Anywhoo….here’s your Secondary Shot Assists leaders this season & then we’ll look at the Relative leaders.
Of the 30 players listed above, 60% play for either Rangers or Celtic. Again, the volume of shots they create - especially from sustained build-up - is at a whole different level….& highlights the appeal of the Relative stats listed below….which propels one midfielder who isn’t even on the list above all the way up to #6 when comparing rates to his midfield peers on his team.
- Olivier Ntcham, Celtic +0.90
- Kyle Magennis, St Mirren +0.83
- Eros Grezda, Rangers +0.75
- Callum McGregor, Celtic +0.64
- Gael Bigirimana, Motherwell +0.64
- Keaghan Jacobs, Livingston +0.61
- Martin Woods, Dundee +0.53
- Vykintas Slivka, Hibs +0.52
- Ryan Kent, Rangers +0.46
- Olly Lee, Hearts +0.45
- Niall McGinn, Aberdeen +0.44
- Daniel Candeias, Rangers +0.40
- Stevie Mallan, Hibs +0.40
- Sean Clare, Hearts +0.38
- David Wotherspoon, St Johnstone +0.37
- Daryl Horgan, Hibs +0.31
- Mayindou Madianga, Dundee +0.31
- Gary Mackay-Steven, Aberdeen +0.30
- Chris Burke, Kilmarnock +0.26
- Cameron MacPherson, St Mirren +0.23
- Jack Hastie, Motherwell +0.23
Keaghan Jacobs has been one of the most impactful midfielders for his team in the league. When you look at the very basic stats of goals & assists - he doesn’t stick out at all (one goal, two assists)…but then when you look at a number like Secondary Shot Assists in relation to the rest of Livingston’s midfield - that’s where you see the impact. He averages 1.10 Secondary Shot Assist per 90, which is 0.61 better than the average rate of the rest of Livi’s midfield. This is an indicator that he has the kind of vision to change the pace of build-up play in the final third which is often leading to shots for a Livi team that struggles to generate much in attack.
Twenty-year old Kyle Magennis has been sidelined with an injury since the turn of the new year, which is a shame given the overlooked impact his was having on a poor St. Mirren side.
Gael Bigirimana was a player who showed well in in the Relative Shot Creation Passing stats last season as well & his increase in minutes hasn’t lessened his impact (he was also has a Relative Key Passes rate of +0.87).
It’s interesting to see players like Magennis, Ryan Kent, Olly Lee, Niall McGinn, Daniel Candeias, David Wotherspoon, Gary Mackay-Steven & Chris Burke feature on both Relative Shot Creation Pass leaderboards here. They are clearly keys to their team’s play in the final third, which is especially impressive for the central midfielders (Magennis, Lee & Wotherspoon) to be dictating the attack while so much of shot creation originates from wing play in the Premiership.
I know…it feels like I’m burying the lead here, right?
Eros Grezda has been seen as a pretty high priced bust so for & I was pretty shocked to see how high he ranked in this stat.
He’s actually doing something…positive…in attack…when he’s on the ball? Mind blown?!?
That’s the fun of analytics…the numbers have shown us something. The next step (quite possibly for a future post) is to review those plays in which Grezda’s getting those Secondary Shot Assists to highlight what he’s doing right in those moments.
Who has the worst Relative Secondary Shot Assist rates? Well, let’s take a gander:
- Ross McCrorie, Rangers -0.89
- Marvin Bartley, Hibs -0.76
- Scott Sinclair, Celtic -0.67
- Mark Milligan, Hibs -0.59
- Lassana Coulibaly, Rangers -0.50
- Lewis Ferguson, Aberdeen -0.47
- Peter Haring, Hearts -0.46
- Lewis Spence, Dundee -0.46
- Matthew Willock, St Mirren -0.45
- Tom Taiwo, Hamilton -0.44
- Danny Swanson, St Johnstone -0.44
- Scott Wright, Dundee -0.39
- Ricki Lamie, Livingston -0.37
- Gboly Aryibi, Motherwell -0.33
Let’s not jump to many conclusions (yet) with this set of data. However, a couple of things stick out:
Ross McCrorie? Wow…but let’s hold that thought until we look at the next list of stats
I think Danny Swanson has hit the nasty side of the aging curve. He had a very good season last season but given the role he’s expected to play - it’s not good to see him at the bottom of both the Relative Key Pass list & then again on this list.
It’s been a while since I’ve had the chance to share data on the Establishing Pass, which is what I termed the third shot assist. It’s probably not as important as the other two types of Shot Creation Passes we’ve discussed…but I do think the outliers in this category deserve some attention. Here’s what I wrote back in November 2017:
Below are the leaders for Establishing Passes per 90 this season & then I’ll share the Relative leaders.
Again, this stat is dominated by the Old Firm clubs. Of the 34 players on this list, 59% of them play for Rangers or Celtic (that doesn’t include Glen Kamara given he played most of season with Dundee). Again, that’s why we will look for outliers in the Relative Establishing Passes category to see who really is impactfully driving play for their teams (rather than just piling up the numbers because they play for a team that generates tons of shots).
- Callum McGregor, Celtic +0.85
- Lassana Coulibaly, Rangers +0.66
- Glen Kamara, Dundee/Rangers +0.60
- Ryan Christie, Celtic +0.58
- Marvin Bartley, Hibs +0.57
- Alan Power, Kilmarnock +0.56
- Olivier Ntcham, Celtic +0.46
- Olly Lee, Hearts +0.45
- Drey Wright, St Johnstone +0.44
- Graeme Shinnie, Aberdeen +0.42
- Stevie Mallan, Hibs +0.39
- Ryan Jack, Rangers +0.37
- Andy Rose, Motherwell +0.36
- Connor McLennan, Aberdeen +0.28
- Stephen McGinn, St Mirren +0.28
- Alex Gorrin, Motherwell +0.27
- Scott Pittman, Livingston +0.25
- Keaghan Jacobs, Livingston +0.19
- Steven Lawless, Livingston +0.17
- Aaron McGowan, Hamilton +0.16
- Peter Haring, Hearts +0.15
- Gary Dicker, Kilmarnock +0.14
- Gael Bigirimana, Motherwell +0.12
It’s safe to say that Celtic’s attack goes through Callum McGregor. He has shown up on nearly every (good) list presented here, which is evidence he has taken charge of replacing the impact Stuart Armstrong had last season. In 2018-19, the only Relative Shot Creation Passing Stat in which McGregor was an outlier was in Secondary Shot Assists. For context, last season his Relative Establishing Pass rate was -0.03. This season he’s the pace setter & really is justified as a potential Player of the Year winner.
Olly Lee is the only other player in the league to feature in each of the Relative Shot Creation Passing tables. Lee has been an ‘analytics darling’ from the very outset of the season:
Lassana Coulibaly?? He came out of the gates this season impressing on so many levels & then…not so much. The 22-year old has only played 94 minutes in 2019 as his loan deal looks to fizzle out in the closing weeks of the season. But there he is…displaying numbers that suggest he may have been a better anchor to Rangers midfield than Ryan Jack or Ryan McCrorie? I know that’s a stretch & I, like many Rangers supporters, became very tired of seeing Coulibaly in the line-up this past winter. However, like in the case of Eros Grezda, the numbers suggest there was an impact in Rangers attack that easily could have been overlooked.
What about McCrorie? He’s been one of the few untouchable players amongst Rangers support who has been immune from criticism. Given his role, I’d expect to see his Establishing Passing numbers to be grouped with the likes of Glen Kamara, Alan Power, Ryan Jack…shit, even Marvin Bartley, right?
Below, I regretfully present the players with the worst Relative Establishing Passing rates.
- Scott Sinclair, Celtic -0.80
- Glenn Middleton, Rangers -0.67
- Rory McKenzie, Kilmarnock -0.62
- Elton N'Gwatala, Dundee -0.51
- Darian Mackinnon, Hamilton -0.44
- Craig Sibbald, Livingston -0.40
- Eros Grezda, Rangers -0.38
- Adil Nabi, Dundee -0.37
- Gboly Ariyibi, Motherwell -0.35
- Martin Boyle, Hibs -0.32
- Ross McCrorie, Rangers -0.32
- Chris Cadden -0.31
Unfortunately, the numbers show that Ross McCrorie has had very little impact on the build-up play leading to shots. He ranked in the bottom of each of the Relative Shot Creation Passing stats listed here. He is known for his ability to read the play & to make quick, decisive decisions on the ball…as is Glen Kamara, who was one of the outliers (in a good way) when it comes to Establishing Passes (as he was last season when he had one of the best rates in the league).
Maybe it’s time to go back to developing McCrorie as a center-back, where he showed an awful lot of promise until the new manager came in proclaiming that playing center-back nearly “destroyed his career” despite the fact he had spent most of his youth career playing….center-back.
Ultimately, I’m wondering if McCrorie’s value would have a higher ceiling as a ball-playing center back rather then a holding midfielder? What if it was McCrorie who got Joe Worrall’s minutes this season alongside Connor Goldson? Could Rangers midfield survive with the likes of Ryan Jack, Lassana Coulibaly, & eventually, Glen Kamara playing in McCrorie’s role……or even Jordan Rossiter?
I guess this post ended on a tangent…but it’s basically a feeling I’ve had since the summer that it was strange to derail McCrorie’s development as a defender to turn him into a midfielder after years of playing mostly as a center back. I wasn’t expecting the numbers outlined in the post to be so…..negative when it comes to McCrorie & while it doesn’t change my opinion about him as a player…it does make me question some of the decisive decisions made by Rangers management this summer.
Stats are courtesy of data provided by Ortec Sports.
Stats don’t include this past weekend’s matches.
I probably shouldn’t have finished this blog post after watching Sunday’s Old Firm match. It put me in a mood & it devolved into a rant regarding Ross McCrorie’s deployment this season. But, honestly - I never really understood this Gerrard quote from the summer: “We had a player (McCrorie) playing at centre-back position that wasn't his real position, who was 19 years of age and thrown in at the deep end, who could have been destroyed in terms of his career.” McCrorie began playing center back in Rangers set-up as a 15-year old, six years before Steven Gerrard joined the club & made his (very impressive) debut...as a center-back in the “deep end” that is a match against Celtic. His play back in 2017, had him being touted as a ‘next great’ center back for both club & country.
Ok….let’s focus on some of the positives — these numbers really highlight the impact that young players like David Turnbull, Kyle Magennis, & Scott Wright have had on their clubs. It will be an interesting summer to see if/when clubs come inquiring about their services.
Yes, you did see Borna Barisic topping some lists here. I guess he plays for Croatia’s national team for a reason.
I hope that some these numbers begin conversations that go against the current assessment about players in the league. So often, group-think creates a narrative that might actually be overlooking the underlying reality that analytics can add to the discussion.
Yes, I know that all shots are not created equal…so be on the look out for similar Relative stats posts involving Expected Assists, xG Chain, & Scoring Chance Key Passes.
This was written under the influence of Spiritualized, Crazy Horse, the Mark Lanegan Band, & Karp.