Deila vs Rodgers, by numbers
Overall Records
The Deila era yielded 2 (expected?) league titles, 1 League Cup and 1 post New Year European run in the Europa League. Those, like me, who have followed Celtic for 40+ years do not take European football after the Bells lightly (1 post New Year appearance in my 1st 27 years of support).
Competition 2014/15 2015/16 Champions League Final Qualifying Final Qualifying Europa League Last 32 Group Stage League 1st 1st Scottish Cup Semi Final Semi Final League Cup Winners Semi Final
Whether you feel the team stagnated or went backwards, in most sports you are dying if you are not moving forward. The overall results suggest slight decline (% Success is calculated by taking every game and awarding 3 points for a win and 1 for a draw and assessing the % of “Points” won over all fixtures):
2014/15 P W D L F A Pts % Success League 38 29 5 4 84 17 92 81 CL 6 2 1 3 7 8 39 EL 8 2 3 3 13 15 38 SC 5 3 1 1 13 4 67 LC 4 4 0 0 13 0 100 TOTAL 61 40 10 11 130 44 71% 2015/16 P W D L F A Pts % Success League 38 26 8 4 93 31 86 75 CL 6 4 1 1 10 5 72 EL 6 0 3 3 8 12 17 SC 4 3 1 0 10 2 83 LC 3 2 0 1 5 4 67 TOTAL 57 35 13 9 126 54 69%
We lost less games in Deila season 2 and scored more league goals but struggled in the European competitions and did not win a domestic cup. The “Denayer and Van Dijk” effect - the loss of a stellar centre back pairing in season 2 - can be seen with 31 league goals conceded to 17 and 54-44 conceded overall (in 4 less games).
The Rodgers era has started in spectacular fashion by any measure domestically. 97% of “points” have been won domestically with 2.71 goals per game compared to 2.4 in Deila’s 2nd season. In Europe our record and performances (more later) are comparable, but against better opposition (!small sample alert!).
2016/17 P W D L F A Pts % Success League 20 19 1 0 54 14 58 97 CL 12 3 4 5 16 23 36 LC 4 4 0 0 11 0 100 TOTAL 36 26 5 5 81 37 77%
But in pure results terms Rodgers is 8% better than last season with a League Cup trophy, a likely league title (expected?) but no European football post New Year. This can be tempered by getting, arguably, the hardest possible CL draw and the monetary and experiential riches that come with such a campaign.
The more things change…..
Firstly, and counter to the confirmation bias risk inherent in this article, here are some surprising performance statistics that show “little change” between the Deila and Rodgers teams so far. It is of course tempting to start an exercise like this thinking “Rodgers team is far better” and find the stats to back that up as the evidence of one’s own eyes in terms of exciting, attacking and winning football, is compelling. Breaking down the performances and looking at the data can, however, reveal a more nuanced picture.
All values are either average numbers of events per 90 minute game or a %.
KPI Deila 2014/15 Deila 2015/16 Rodgers 2016/17 Shots On Target 6.8 7.5 6.9 Shots Off Target 11.3 11.9 12.4 Total Shots 18.1 19.4 19.3 Shot Accuracy (% on target) 38% 39% 36% On Target Conversion (% on target = goals) 31% 30% 33% All Shot Conversion (% shots = goals) 12% 11% 12% Chances Created 6 5 5 Chances Creation Failed 8.4 5.6 7.5 Dribbles (or Take ons) Complete 5.9 9 8.8 Dribbles Incomplete 6.9 7.9 9.2 Fouls Conceded 12.3 12 11.4 Clearances 47 27 30 Interceptions 37 33 33 Tackles Won 46 45 38 Tackles Lost 39 39 43 Defensive Saves 3 3.1 2.1 Defensive Errors 4.4 4.3 3.4 Defensive Critical Errors 0.4 0.5 0.3
Under Rodgers, Celtic are taking no more shots, are no more accurate nor successful with them, creating no more chances nor attempting to, and defensively performing about the same number of actions but losing more tackles. Numbers for both domestic and Euro campaigns are similar to the Deila era, albeit against more difficult Euro opponents.
Blimin’ stats , eh? Same players though, so what did you expect???
Ok so what is different? Because as a Real Football Man would know, the only stat that counts is the one in the top right corner of the screen and, indubitably, those have been better.
The Rodgers (r)evolution(?)……..
Consider the following, again based on 90 min averages or %:
KPI Deila 2014/15 Deila 2015/16 Rodgers 2016/17 Passes Complete 478 464 514 Passes Incomplete 95 88 79 Total Passes Attempted 573 552 593 Possession 61% 60% 62% Goals For Per Games 2.13 2.21 2.25 Goals Against Per Game 0.72 0.95 1.03 Possession Events Won 565 552 599 Possession Events Lost 202 195 187 PEI % 74% 74% 76% Win or Draw from being Behind 3 5 5 Lose / Draw from being Ahead 9 11 3
The most startling difference in team performance is the passing data. 50 more successful passes per game compared to Deila season 2 – an increase of 11%. Also less passes given away. The Deila numbers show a decrease in both passing completed and incomplete as the play became slower in season 2 with combinations of Brown, Bitton and Johansen playing passes from deep central midfield, often in circulation with central defenders. This is an aspect we will explore in more detail in another article – the impact of having only 1 deep lying midfielder under Rodgers compared to a rigid 4-2-3-1 of Deila. Celtic move the ball quicker, and it appears, with more purpose and to a better end result. Also, the Rodgers team look to get the ball out wide to full backs with the attacking wide players inverting inwards to link with the on coming central midfielders, of which there are now 2 attacking instead of 1 under Deila. But what do you think has caused the increase in pass completion?
A word on PEI – this stands for “Possession Effectiveness Index” and is a measure of the proportion of all events where you maintain possession. Higher is better generally (at an individual level it will be higher for defenders than attackers) and a 2% increase to the Rodgers era is a positive increase. This increase will mainly be due to more passes completed as passes account for over 50% of all on field actions, and as we have seen tackles won has actually gone down (more possession?).
We are both scoring and conceding more goals but the 0-7 hammering from Barcelona is a real outlier in a small sample and will therefore distort the averages.
The other most notable positive difference as far as Rodgers is concerned is that we have won or drawn 5 games from losing positions, in line with Deila’s team (but in 60% of the season), but only lost or drawn 3 from winning positions so far compared to 9 and 11 draws/loses from winning positions in the Deila seasons. The team appear to be more resilient.
Additionally 25% of all Celtic goals under Rodger are coming from the 80th minute onwards compared to 12% in Deila season 2. Are we also a fitter team?
The opposition scored 17% of their goals in the last 10 minutes in Deila season 2 and it is 5% under Rodgers.
Finally, in terms of opposition performance, the opposition stats for passing and possession highlight the different challenges of playing Scottish domestic football and Champions League. And Champions League opposition performance dwarfs that achieved by Europa League opponents:
Domestic Opposition Performance
KPI – Domestic Oppo Deila 2014/15 Deila 2015/16 Rodgers 2016/17 Passes Complete 245 248 197 Pass Incomplete 101 102 91 Pass Complete % 71% 71% 68% Possession % 37% 38% 31% Shots On 3 3 3 Shots Off 6 5 5 Shot Accuracy 31% 37% 37% On Target Conversion % 17% 27% 21% All Shot Conversion % 5% 10% 8% Chances Created 2 2 1 Chances Missed 4 4 4
European Opposition Performance
KPI – Domestic Oppo Deila 2014/15 Deila 2015/16 Rodgers 2016/17 Passes Complete 344 345 413 Pass Incomplete 102 104 96 Pass Complete % 77% 77% 81% Possession % 47% 47% 50% Shots On 7 4 5 Shots Off 7 7 7 Shot Accuracy 49% 34% 42% On Target Conversion % 25% 40% 38% All Shot Conversion % 12% 13% 16% Chances Created 4 NA 4 Chances Missed 10 NA 6
Domestically however it is noticeable the Rodgers team is denying the opposition as much ball albeit the opposition number of shots, chances and conversion rates are consistent across all 3 seasons.
(Note the “NA” are due to changes in how I have captured the data for the last 3 seasons).
Summary
On many key measures, especially around attack effectiveness, there is little discernible difference between performance of the Deila and Rodgers teams. Results do not lie and passing data shows a significant increase in passing attempts and accuracy suggesting quicker passing and movement with the corollary that the opposition are having significantly less of the ball to hurt us. Celtic under Rodgers are more resilient and are scoring more and conceding less late goals.
There is a growing body of analytical and statistical based work on Scottish football and I highly recommend some excellent resources complimentary to this data.
I am happy to get feedback and thoughts from supporters in the comments section.
We’ll do this again at the end of the season. Even if Celtic get to the Scottish Cup Final, we are 61% through Rodgers season 1.