As we know Celtic have suffered injury upon injury to the most creative players. On the other hand, the talent pool has deepened in this area with the additions of a third striker (Edouard), a new deep lying creator (Ntcham), a new winger (Hayes) and new number 10 (Musonda). Has this help mitigate the injuries?
xSC
This article deals with the concept of Expected Scoring Contribution (xSC). Very simply, this is the sum of our good buddy Expected Goals (xG) and her partner in crime Expected Assists (xA). I’ve decided all football metrics are female and declare this a global standard.
Very simply, Expected Goals are an indication of the number of goals you would have been expected to score given average outcomes from that scenario. Similarly, Expected Assists are the Expected Goal value assigned to the passer that created the chance.
xSC is therefore an overall measure of expected productivity in terms of goals and assists given the chances created. It is expressed as a per 90-minute average to remove differences in time on pitch.
Finally, the analysis considers matches in the SPFL only. This is due to limitations on the data I have prior to this season, but also it standardises the quality of the opposition across the participants.
2016/17
Last season was an historic one for Celtic with an uneaten SPFL season. Whilst there were injuries and the usual vagaries of form, there was a consistency of selection. Celtic averaged 2.8 changes per game. All the most creative players in the squad managed between 14 and 30 matches worth of game time:
Only Griffiths and Rogic appeared in less that 50% of available SPFL game time. Sinclair put is a huge hunk of work, being on the pitch for over 80% of all SPFL matches.
In terms of xSC, the numbers were impressive from this cadre of creative talent.
The main attackers (Griffiths, Dembele and Sinclair) all managed virtually 1 goal or assist per 90 minutes. The supporting cast managed nearly 1 goal or assist every 2 matches with Rogic on 0.734 xSC90.
This was a focussed and highly functioning unit of players.
2017/18
This season the squad depth has been enhanced. All the creative talent was maintained, and new additions assimilated. But selection has been more variable with 3.3 changes per game. Injury has taken its toll and the utilisations reflect this.
Only three players scrape above 50% utilisation. The majority have appeared less that a third of the time. As we head into the last 8 matches of the season, 9 of the most creative players aren’t near completing 20 of those.
Consequently, it is no surprise to see the xSC’s suffering.
This season in the SPFL no player is breeching 1 goal or assist per 90m. Griffiths is showing again how much he is missed for sheer productivity. Of the new players, only Edouard is appearing near the top of the charts.
In terms of comparison to last season.
Only Forrest has shown marked improvement in expected productivity. His long run in the side yielding a 12% improvement over last season. Roberts gain is modest from a tiny number of minutes. Armstrong’s so-called demise is hugely exaggerated. What he is lacking in goals he is making up for in chance creation. If you follow me on Twitter you will know this!
Surprisingly, McGregor, lauded as having a great season, is producing 10% less expected output compared to last season.
Rogic and Dembele have suffered the largest falls in expected productivity of 20% and 22% respectively. Both are players that need game time to get up fitness. Neither hits the ground running on re-entry to the side.
Finally, the prolific Sinclair has seen a drop of by expected goals and assists but it is modest (5%). Don’t believe all you read in the papers!
All the players mentioned, bar Griffiths, appear ready for the run in of 8 SPFL and, hopefully, 2 Scottish Cup matches.
I hope this shows once again that lack of continuity of selection and injury to key creative talents has meant the heights of last season have not been reached. Next season will be interesting if everyone stays fit.
This article was written with the aid of StrataData, which is property of Stratagem Technologies. StrataData powers the StrataBet Sports Trading Platform, in addition to StrataBet Premium Recommendations.