Rangers need some new faces to start scoring goals

Michael O'Halloran, courtesy of SNS
Michael O’Halloran, courtesy of SNS

written by – Rangers Report

Back in November, while Kenny Miller was mired in a steep goalless drought, the advanced statistics were pointing to an inevitable return of the veteran’s scoring touch.  While looking at his Expected Goals numbers, I wrote – “Even though the goals have dried up of late for the veteran…if Miller keeps getting chances, where he’s getting them – the goals should follow suit.”

Four months later, Miller has scored nine non-penalty goals in 1,174 league minutes.  That’s only three less then team leader Martyn Waghorn – who has scored 12 non-penalty goals in the 2,070 minutes before he went down injured.

As a team, Rangers average 2.68 goals per match, which is nearly a full goal better then the next best team, Hibs (1.71).  Rangers have been thriving off a varied attack that has seen five other players chip in four or more league goals this season.  Jason Holt (8 goals), James Tavernier (7 goals), Lee Wallace (5 goals), Andy Halliday (4 goals), & Barrie McKay (4 goals) have provided a layer of secondary scoring that has been invaluable.  However,  early signs indicated that certain players, especially Holt & Tavernier, would unlikely sustain their high rates of scoring.

Jason Holt was riding a hot hand for much of the season with a shooting percentage that was never going to be sustainable.  His last league goal came on December 28, 2015 against Hibs.  At that point his shooting percentage was 0.615 – meaning 62% of his shots on target beat the keeper for a goal.  Since then his success rate has been regressing towards normality.  In the new year, Holt has seen his shooting percentage dip down to 0.444 – which is still pretty high.

James Tavernier got off to a scintillating & sensational start to the season but given his role & where he shoots from – it wasn’t going to last.  The right back has scored one goal since early November, after scoring six goals in the first 13 matches of the season.  Of his 64 shots this season, 72% of his shots have come from outside the penalty area.  Some of those are free kicks & some are hopeful shots at glory.  Shots from outside the box keep a defence honest & forces them to close down the player on the ball but very rarely to they actually beat the keeper.  Tavernier scored some breathtaking goals at the outset of the season but did anyone really believe that he could keep that form up?  Even the world’s best players struggle to score consistently from long range.

So where does that leave Rangers?  Rangers have hit a bit of a goal drought of late, having only scored five goals in their past six league matches.  Martyn Waghorn is out for the long term & now the statistics are suggesting that Kenny Miller’s form may actually take a dip.

Let’s layout the data for you.  First, a look at Rangers leading goal producers this season.  Rather then simply list a table of goals scored, let’s see which players average the most goals per 90 minutes.  Penalty goals are not factored in.

Non-penalty Goals per 90 minutes

  • Billy King has only played 115 league minutes for Rangers so far – so there is no sense in looking at his rate (yet).  The same with Michael O’Halloran, who has played 206 minutes.  It should be noted that, O’Halloran averaged 0.11 goals per 90 this season with St. Johnstone – which was a significant decline from his 0.32 rate last season.
  • Kenny Miller’s efficiency & Martyn Waghorn’s sheer volume of shots has carried Rangers offence this season.  Waghorn’s shot totals are gone & later we’ll look at why Miller’s success rate may not last. So who will step up to make up for the goals lost through injury & potential regression?
  • King’s career goals per 90 minutes is 0.31 & 57% of his career goals came while playing in the Championship with Hearts.  Look for management to give him an extended look to see if he can replicate that scoring rate with Rangers.
courtesy of stats.com

Expected Goals (xG) can really ground statistical analysis to begin assessing which players/teams are over-performing or underperforming.  When Kenny Miller was struggling to get goals, his xG remained among the best on the team.  He was getting quality scoring chances, the goals just weren’t there…yet.

Expected Goals simply place a probability value of a goal being scored based on where a shot is taken & whether is was kicked or headed.  With more data available, Expected Goals models get more &  more complex. However, given the limited information available from Opta’s coverage of the Scottish Championship – my model is somewhat limited.

The red shaded area in the above visual is where most goals are scored & the closer to the goal the higher the rate of success.

When you take a player’s Expected Goals & compare it to his actual goals – you can begin making some assumptions about his future form.  If a player has scored more goals then expected, he may be due for the goal success rate to take a dip.  If a player is scoring well below his xG, some natural questions come to the surface.  Is this a player who is in a bit of a rut but history suggests he will likely bounce back (see: Kenny Miller) or are there flaws in that player’s ability to finish off his chances?

Below you’ll find the data for Rangers players this season.  It shows the difference between non-penalty goals & Expected Goals.

Non-Penalty Goals - Expected Goals

  • First off, the three greatest outliers in the positive direction are three we have mentioned throughout this post – Holt, Miller, & Tavernier.  This trio have been scoring more goals this season then the Expected Goals suggest they should.  For some context, only six players have a higher differential this season in the Championship.  Can they sustain this success for the next couple of months?  Of course they can…but Holt & Tavernier have already been slumping – is Miller next?
  • The three outliers on the other end of the spectrum are a bit of an enigma.  Waghorn’s recent struggles prior to his injury brought his actual production below his expected production.  He probably would have turned that around if it weren’t for his injury.  The fact that he has scored 12 non-penalty goals & his Expected Goals is a league high 13.23 suggests he would have been just fine.  Gedion Zelalem has zero goals from his ten shots this season.  The stats say he is due for a goal, but the ‘eyeball test’ says he may go the rest of the season without scoring.  Zelalem has three scoring chances this season (shots from that red area) but so far has lacked that instinct to fire the ball past the keeper.  Harry Forrester has only played 161 minutes for Rangers but already has 14 shots, which averages out to 5.6 shots per 90 minutes.  But only three of those shots have been scoring chances as the majority of his shots come from the edges in the box or from outside the penalty area.  Forrester averages 0.22 goals per 90 minutes in his career & 0.28 in the two seasons in which he played 2000+ minutes.  He has a history of scoring when given extended playing time – so the goals may be coming – but don’t expect them to come in bunches.
  • What about Billy King & Michael O’Halloran?  They have combined for more then 2,200 minutes this season in the Premiership.  Thankfully, Matt Rhein has been doing analytics work covering Scotland’s top league & has been running a similar site to ours over at The Backpass Rule.  He graciously shared his data on Rangers newest players.  This season King scored two goals for Hearts & had an xG of 1.16, while O’Halloran also had two goals with an xG total of 2.69.  So, both players were producing roughly what their projections suggested they would.
Mark Warburton & Billy King, courtesy of SNS

Is there hope here?  Can Kenny Miller maintain his current pace?  Can Jason Holt start scoring again?  Are one of the January signings the answer?  How about Nicky Clark?

Our last presentation of stats, looks at which players are finishing at a high/low rate & who is getting the best shots on goal.  The graph below looks at shooting percentage (goals /shots on target) & Expected Goals per shot.

Shooting Percentage v xG per shot

  •  The trend line running through the heart of the graph can act as a gravitational pull.   Given a player’s xG per shot you can move up or down to his shooting percentage.  The outliers are the players outside each of the curved lines.
  • For example, Holt was once an outlier on this graph & has been steadily moving back to the norm.
  • Kenny Miller has the highest Expected Goals per shot largely because 74% of his shots have been scoring chances (in that aforementioned red box).   His shooting percentage is high (.529) because he has been a very good finisher when he gets those high quality shots.  His shooting percentage from that High Danger area is 0.538, for a comparison – Waghorn’s rate was 0.346.  But can a 36-year old Miller be called upon to play 90 minutes every match, or has his success this season come due to Mark Warburton’s management of Miller’s minutes played?
  • Nicky Clark seems to be in line to see an uptake in scoring, but the jury is still out on whether he is up to the task.  His shooting percentage is among the lowest of the players who have scored this season despite generating  respectable chances.  Are there more goals there beyond his winner against Kilmarnock?
  • Again, we have to look at King & O’Halloran’s statistics in the Premiership to get a sense of what they may offer.  O’Halloran average 0.16 xG per shot for St. Johnstone & his shooting percentage was 0.286.  If you place that on the above graph – that glimmer in your eyes is hope.  If he can get a similar quality of shots going forward, you’d expect his shooting percentage to increase – which means more goals.  King’s xG per shot was 0.10 with Hearts & his shooting percentage was 0.333, which would fall right on the trend line.

The bottom line going forward this season is that Michael O’Halloran may become the player relied upon the most to carry the weight of goal scoring for Rangers.

You can follow Rangers Report on Twitter @TheGersReport

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s