Celtic scored 147 goals in winning a treble last season, in all competitions. 2.77 per game.
This season was always going to be some kind of transition from Ange Postecoglou to Brendan Rodgers and we do not have to rehash the recruitment story.
Those goals came from shots rated with an overall xG of 139. Some nice overperformance then but not ludicrously so.
This term Celtic are averaging 2.18 goals per game and with a maximum of 10 matches left in league and Scottish Cup, are on track to hit 108 goals.
Those are currently coming from xG of 102 – 2.23 xG per game. 87 goals so far from 102 xG is quite an “under” performance.
2022-23 Player
At the player level, here are all those who played at least 5 matches or more in terms of their individual goals to xG per 90 minutes differential:
Kyogo Furuhashi had an exceptional season and many of his goals were from inside the six-yard box and central and probably underestimated using the basic xG model I have.
Many of the “over” achievers were those that often came off the bench and gorged on tiring opposition – James Forrest, David Turnbull, Oh Hyeon-gyu.
Giorgos Giakoumakis and Matt O’Riley were notable for their “under” performance relative to their xG.
But the general point here is that half the squad over-performed their xG and half underperformed, roughly. Let’s call it a “normal” distribution.
2023-24 Player
This season we know as a team that some mean reversion hoodoo is going on, but how does that manifest at the player level?
Eeek.
Only three players are in the overperforming category, and one is now sitting on Cardiff City’s bench whilst the other two, who cares?
Top scorers Furuhashi (15) and O’Riley (11) are nearly at 0.1 goals per game under their xG.
Why does this happen?
You would have to go through each shot and categorise:
- Poor finishing
- Good goalkeeping/defending
- Bad luck
Likely for each player, it is a combination of all three.
Let’s consider some fundamentals:
Celtic are taking LESS headers as shots (always lower xG efforts) and the same proportion of shots inside and outside the box as last season.
Those would be two obvious areas to consider.
The one major difference is shots taken from Zone 1 – central to the goal within the six-yard box – very high xG territory. That has fallen from 10 percent of all shots to seven.
Celtic need to get back to the low, hard cross across the six-yard box and utilise the special movement skills of Kyogo to get on the end of them – simples!
As well as getting on the right side of the variability of football in terms of finishing skill, goalkeeper performance and sheer luck, naturally.
Celtic’s wingers aren’t getting to the byline very often this season, and that is restricting the opportunities for low, hard crosses, or cutbacks, into the six-yard box. I think the main factor in this has been Rodgers’ selection of inverting wingers, particularly playing the very right-footed Palma on the left wing.
Of course, Hatate not being available hasn’t helped; he’s very good at releasing Taylor or a winger into a position behind a defence. Rodgers might have better advised to play McGregor as the left-sided 8 to replace him instead of the more conservative Bernardo.
Alan sorry if I’m being stupid andissing it but what is our xG per shot?
Wasn’t there some talk with Sevco under Beale that whilst their xG looked ok it was a lot of small % chances and that this was skeing their numbers
xG per shot last season was 0.152
This season its 0.143
“This term Celtic are averaging 2.18… currently coming from xG of …2.23 xG per game…”
I’m struggling with the “Math” here, as we appear to be 0.05 goals down per game ie 1 per 20 games, 2 per 40 etc.
I’m not sure where the 15 goals down are coming from.
Clunky writing on my part
22/23 – 2.77 goals per game
23/24 – 2.18 goals per game
22/23 – 2.62 xG per game
23/24 – 2.54 xG per game
Ta, that makes more sense.
The last couple of games we’ve had more of the low cross through the 6 yard box than probably all season. Hopefully this isn’t yet another false dawn, but a sea change for the run in.
Time will tell.
I don’t think the change in approach has helped Kyogos stats one iota this Season.
He has for the most part been forced to drop deeper and deeper for the bulk of the Season as we have struggled to get the ball from defence to attack with anywhere near the intensity we seen under Ange.
Brendans focus on ball retention and possession has had a large hand in that as has the change in role for Callum McGregor within it.
Under Ange he was operating as the Celtic tempo setter as a deep lying playmaker but his role under Brendan has been more of that of a traditional No6.
Neither player has in my opinion benefited from these changes but none moreso than Kyogo.
I think we have seen a shift in tempo and approach in the last 2/3 games ,not quite Angeball but a lot closer to it than early to mid Season.
As a result we now see Kyogo in the box ,smiling and doing what he does best.
Putting chances away and more importantly getting on the end of them where he can hurt teams most.
Timing is everything in Football and I think the timing of this might just be the difference in deciding where the League Trophy ends up.
Agreed, although the counter argument is that had the timing been 6 months ago there would be little doubt where the league trophy was going to end up.
That’s down to the Managers decision to change the approach I’m afraid.
The change back to a more aggressive,front foot higher tempo Celtic is the reason for the upturn in fortunes for Kyogo.
The argument all Season I’ve been reading is Calmac and Kyogo haven’t been themselves?
Well that’s because the change in approach has asked them both to be something different.
McGregors form under Steve Clarke meanwhile seen no change or drop in performance.