Striking Consistency
The forward line has been disrupted by injury we know. How has this impacted the consistency of the strikers?
There are many measures for the forward players. The one linked most explicitly to productivity is Expected Scoring Contribution (xSC). Scoring Contribution is the sum of goals and assists and is a simple measure of productivity for attackers. “Expected” is because you don’t always get what you deserve (great saves, player misses from your excellent through ball, ball bobbles in 3 foot of grass). Using Expected Goals (xG) and Expected Assists (xA) tells us what you should have had, on average.
Players can go on streaks of scoring, and conversely droughts. Like all players, you want to see consistency of performance. Rather one goal every other game than five in one game then nothing for ten games. With the strikers hit by injury and/or loss of form, consistency of selection has been missing but what has been the impact on performance?
I’ll consider the main strikers – Dembele, Edouard, Griffiths and Sinclair.
The Glossary is the sniffer in the box, always alert to the half (expected) chance.
xSC90
xSC90 expresses Expected Scoring Contribution as a per 90-minute average. It therefore removes the vagaries of number of minutes played that reduces the usefulness of traditional measures such as “goals per game”. This season the aforementioned players’ xSC90s are:
Dembele and Griffiths are the primary central strikers whilst Edouard and Sinclair have tended to play either deeper or from wider. Their scores can be expected to be lower, therefore. Adding together the xG and xA and averaging over 90 minutes reveals that Dembele is the most productive in terms of the quality of the chances he creates and gets onto the end of.
But we are analysing consistency so let’s look at each player and their xSC across each game.
I have only included games where a player participated for at least 60 minutes.
Dembele
The orange line shows his average xSC across all minutes. It is very mountainous profile indicating a lack of consistency. More recent form has been good as we can see with the exception of the painful 0-0 with Dundee.
Edouard
As you can see the new kid has had relatively few (12) +60 minute appearances. But accepting the small sample, we can see a much shallower gradient of performance indicating more consistency than Dembele.
His only very unproductive game was the recent draw with Motherwell.
Griffiths
What stands out in Griffiths profile is a relatively good start to the season that has tailed off badly. No performances with a xSC over his 0.861 average since the St Johnstone game in August! There are notable troughs against Paris Saint-Germain and Bayern Munich, not unexpectedly.
One of the limitations of this analysis is that it does not show Griffiths’ effectiveness off the bench. He often creates and has good chances when coming on as a substitute.
Sinclair
Having played the most minutes Sinclair has the hardest job keeping up consistency. He has had three matches where he has not created nor got on the end of ANY chances (xSC = 0.000) - against Rosenborg, Bayern Munich and Aberdeen. He is also capable of very high contributions, but they tend to be against the weaker sides (Linfield, Hamilton Academical and Partick Thistle).
There is a nice consistent run of productivity before the New Year break but overall, it’s a very “Alps” like profile!
Striking Consistency
All that being said, we can use a simple r-squared analysis of their performance in each match where at least 60 minutes have been played to get a consistency measure.
Drum roll…..
For all the small sample size, across his twelve qualifying performance, Edouard displays the greatest consistency with a r-squared of 0.59 (where 1.00 would be perfectly linear consistency – i.e. the same xSC in every match).
Dembele had the highest overall xSC90 of the group and the second highest consistency indicator of 0.55. He is the No.1 Celtic striker.
Sinclair’s season has petered out after a strong start and not only is his xSC90 lagging the other strikers, his consistency indicator is below 0.5.
Griffiths, whilst often effective off the bench, and having the second highest xSC90, is by some way the least consistent performer.
This analysis does not say who is the “best” striker. It considers consistency as an attribute. I could equally apply the same principle to other data points such as shots on target or chances created. And there are limitations in excluding sub 60-minute performances, and due to the small sample sizes.
But it is an indicator that despite his age, young Edouard is capable of providing overall goal threat (creating chances and shots at goal) on a more consistent basis than his peers.