Selection Dilemas
I hope you all had a Merry Christmas and will have a safe and healthy New Year. The final article of the year considers the selection possibilities for the final match before the Winter Break. With an 8-point lead, Celtic know a win could dee that into a double-digit lead if other results are favourable.
With injuries to key creative talents Rogic and Roberts, and with defensive players Gamboa, Bitton, and Sviatchenko seemingly out of favour, there are some near certainties within the line out. I posit that the following will play:
Gordon
Lustig
Tierney
Boyata
Brown
Forrest
Sinclair
That leaves 4 starting places up for grabs:
1 centre back from Ajer or Simunovic
2 centre midfielders from Armstrong, Ntcham and McGregor
1 forward from Dembele, Edouard and Griffiths.
Centre Back
Simunovic had a nightmare game away at Hearts and has lost his place to Ajer for the last three matches, although he was reported to have had a knock for the Partick Thistle game. This has required Boyata to move to right centre back with Ajer slotting in at left centre back. Collectively, the defence have done will with highlights including:
Three clean sheet.
Only 3 shots on target have been allowed in those three matches.
The opposition have averaged 0.057 xG per match compared to the season average of 0.544.
No Big Chances have been given up.
Only 2 crosses have been successful into the box (from open play - excludes free kicks and corners).
An average of 9 Celtic defenders have been bypassed – the season average is 12.
The opposition have average 8.3 possessions in the box compared to the season average of 10.8.
Ajer has been particularly impressive. Against Aberdeen he won 21 challenges – the most by any player in any game this season. He has won 36 challenges in his last 180 minutes. And he has yet to be attributed a defensive error this season. Furthermore, he adds attacking threat with 2 efforts off target and averaging 2 successful dribbles per match, each taking out 3.44 players on average.
Defensive Effectiveness
Comparing the key defensive metrics:
All these figures are averages per 90m. Also, if you are unsure of the terms and their meaning, the Glossary is Santa’s bestest statto Elf!
Simunovic has played 1622 minutes (equivalent to 18.02 games, representing a 49% utilisation rate).
Ajer has played 910 minutes (equivalent to 10.11 games, representing a 27% utilisation rate).
Yes, it is a small sample, but it is perhaps surprising that Ajer is leading the way on all the defensive metrics. Simunovic hasn’t missed blocking any crosses, has been dribbled past less, and effects more clearances. But overall, Ajer has a whopping 10% higher DASR (Defensive Action Success Rate). The DASR takes a basket of defensive stats to come up with a composite view of defensive effectiveness. For context, here are the DASR scores for the Celtic central defenders:
Simunovic had led the way all season but has dropped off recently. Sviatchenko has had little game time. It will be interesting to see whether Ajer can maintain such a high score, and indeed whether he performs in the Europa League ties.
Possession Effectiveness
In terms of possession effectiveness, the tale of the tape is:
Ajer is more influential in that he has greater possession (by 15 events per 90m) and completes 13 more passes in open play, more accurately. He also wins more possessions. His overall PEI of 94% may be difficult to maintain over a bigger sample, and add in exposure to European football.
Verdict:
Ajer is clearly in great form. The Aberdeen game was a great audition, showing he can cope with the best of the rest in Scotland. Albeit the Dons played into his large hands (or head) by attempting a long ball game.
On the other hand, Simunovic represents more experience generally and in big games.
There are always two questions in these situations. 1. “What do I think Rodgers will do?” and 2. “What would I do?”.
I think the answer will be the same: three clean sheets, safe against Aberdeen, don’t change a winning back line. Ajer will play.
Central Midfield
If McGregor plays he will be the closest to a “10” Celtic have but this will not stop him coming deep for the ball. Ntcham and Armstrong are more similar in that they will try and break lines from deeper using a range of passing, or in Armstrong’s case, strong running. McGregor is a seeker of space in which to receive on the half turn and link play. So, there is an element of tactical decision making required: a more solid midfield 3 with Ntcham and Armstrong or a 2 with McGregor pushed slightly further on.
Defensive Effectiveness
The central midfield must support Brown and the back four as well as support the attack, so a wealth of data to sift through.
It is remarkable how similar Armstrong’s and Ntcham’s defensive data is. Only thing I’d point out is Armstrong’s tendency to be easy to run past with the ball whereas Ntcham positions his body more intelligently. Otherwise Celtic will get the pretty much the same defensive output from either.
McGregor is predictably the lest effective defensively. He plays further forward and therefore is the subject of most challenges rather than initiating them. Also, he is the slighter man.
Possession Effectiveness
Again, Ntcham and Armstrong are remarkably similar. Ntcham is a little more accurate and will be less likely to lose possession.
McGregor is the most accurate passer at 91% and also loses possession less than the other two.
Creative Effectiveness
Getting to the meaty bit now!
Armstrong and Ntcham continue to be incredibly close on many key metrics. But Armstrong is creating more chances per 90m than both team mates. He is also the most effective dribbler, completing nearly three times the take-ons per 90m that Ntcham and McGregor. And his xA assist per 90m is the highest yet slightly below his actual assist rate of 0.29 per 90m. Slight concern is that Ntcham’s xA is half his actual, suggesting he is overperforming and may regress to the xA average.
Armstrong and Ntcham lead the squad in Pass Impect indicating you need at least one of them to break the opposition defensive lines.
McGregor’s stats hint at “competent at all but Master at none”. Sidebar: Rogic’s numbers are better than all three in many cases.
Scoring Effectiveness
Goals are shared around at Celtic and therefore the central midfielders are expected to contribute.
Armstrong had such a stand out season with 17 goals last, it was always going to be a stretch to match that given the number of outside the box shots that went in – the odds are not in his favour. And it is proving so this season. He is taking more shots from outside the box but is easily the most accurate of the three, but has the lowest goals per 90m.
Caution with Ntcham again as his actual goal tally far exceeds his xG. Many of his goals have benefitted from deflections. He has the highest ratio of shots from outside the box of the three and easily the highest shots off target average. His luck won’t hold for ever!
McGregor has a healthy 0.3 goals per 90m with the highest xG. He also has the highest Pack Receive rate, indicating his ability to find space and receive and control passes that take out opponents. I like McGregor’s ratio in favour of shots in the box too, over 50% being taken in more favourable positions.
Overall Attacking Effectiveness
Armstrong is the packing king when considering passing, receiving, and dribbling. Only Rogic has a higher overall Impect score.
McGregor is averaging 0.554 Goals and Assists per 90m, which will be a career high for him. Note that both he and Ntcham are outperforming their Expected Scoring Contributions. Whereas Armstrong is closest to his expected and has the highest expected yet the lowest actual – has he been unlucky this season?
Despite playing deeper, Armstrong’s overall CAT score is only slightly below McGregor’s. McGregor is managing scoring contribution every 114 possessions.
Ruefully I must note Rogic beats most of these numbers.
Verdict
CBN loves Armstrong and whilst never a sacred cow, would play him when fit. I would also play Ntcham for a more resilient midfield to win the initial battle. Playing both Ntcham and Armstrong provides maximum deep passing capability and defensive solidity without greatly reducing overall attacking threat. McGregor is excellent in broken field with space to play and therefore ideal as a substitute.
Striker
All three strikers have been rotated of late. Whilst it would be unusual to put the young Edouard under Derby pressure, he is worth considering as Rodger’s now speaks of having three choices.
He has now played the equivalent of 6.68 matches with 601 minutes. Dembele has played 1071 minutes (11.9 matches) and Griffiths 1611 (17.9). None have appeared more than Griffith’s 48% utilisation.
Creative Effectiveness
Griffiths takes many free kicks and corners and therefore has the highest actual and expected assist rates. He also has the actual assist rates closest to his expected. Dembele and Edouard are over performing theirs by some distance.
Consequently, Griffiths leads the way on overall Chance Creation. The young French strikers both complete more dribbles than the Scotsman and complete more passes that take out opponents. For Griffiths, a pass is a waste of a shot!
Scoring Effectiveness
The young Edouard, on less than 7 games evidence, is just ahead of Dembele on goals per 90m. Small sample and all that. But impressively he has the highest xG as well, and the best shots on target ratio.
Dembele has the most Big Chances, and the most shots on target per 90m. He has over 3 shots in the box per 90m. Griffiths is now on the positive side of that ratio as well. The French pair also average over 1 possession in the box per 90m more than Griffiths.
Griffiths hold up play has improved this season and is evidenced by the fact his Receive Pack data matches the France U21 star’s. This is one area Edouard can clearly improve.
Overall Attacking Threat Effectiveness
Whilst Edouard leads the squad with over 1 Scoring Contribution per 90m, this is unlikely to be maintained over a season including European games. Dembele is just edging Griffiths although he is outperforming his expected Scoring Contribution to a greater degree.
Dembele is the Top CAT, with an average 1 more per 90m than Griffiths. He also edges the Scot in terms of minutes per Scoring Contribution with 1 per 97m.
Verdict
Edouard has posted impressive numbers on limited starts and will have a squad role to play over the season.
For this match, I’d go with Dembele. Griffiths has improved, yes, and Dembele appears not to have hit the form heights of last season. Yet he is still posting impressive performance numbers and can usually be relied upon in the big games.