In his excellent book, Inverting the Pyramid, Jonathan Wilson explains the evolution of football tactics. Today, there is very little new from an innovation perspective at the macro level (i.e. new formations or approaches), rather coaches try to find a) weaknesses in the opposition to exploit or b) devise ways to ensure your own better players are on the ball in sufficiently dangerous areas. In many ways, whatever the nominal formation, the game, tactically, is a veritable Rock-Paper-Scissors as coaches try to identify patterns in the opposition to exploit and adjust accordingly. Unless of course you have the superior players in which case you let your opponent do their worst. This was the Deila philosophy and predictability proved beatable, on occasions.
Derek McInnes cannot be accused of under thinking to overcome the challenge of matching up to Celtic. In the 5 matches this season, a variety of formations and approaches has been tried with little obvious success.

27/08/16 – Celtic 4, Aberdeen 1

McInnes opted for an ultra-defensive 5-4-1, sitting his team very deep which makes sense to counter the pace of the Celtic attack – that day featuring Sinclair, Forrest and Griffiths.
In some regards this was successful in that Celtic were restricted in chance creation, but broke through with a long range strike from Griffiths and well-crafted Forrest finish. More encouragement for Aberdeen was that they had 1 chance, from a Celtic error, that Rooney took. A Celtic team visibly tired 4 days on from the emotionally and physically draining experience in Hapoel Be’er Sheva, were happy to maintain possession.
The Dons pressed higher in the 2nd half, and the high press stopped De Vries playing out from the back as 3 players covered the first receivers. There were glimpses of trouble for Celtic when the Dons played diagonal passes onto the two Celtic full backs – that day Gamboa and Tierney. This has been a weak spot for Celtic over the season and if Lustig is not fit for Celtic, would remain a concern. Ultimately, Celtic were happy to contain. Aberdeen continued not to be able to create and a patient home side were rewarded with 2 late goals. A clinical Celtic scored 4 goals from 5 shots on target – Aberdeen scored from their only shot on target.
29/10/16 – Aberdeen 0, Celtic 1

Despite the close score line, it hides the fact Aberdeen failed to register a single shot on target. A late free kick was diverted by Sviatchenko towards goal and Gordon made an athletic save. Aberdeen had many dangerous situations usually involving crosses into the box, which the Celtic defence dealt with comfortably. At the other end, Lewis kept Celtic to 1, as Aberdeen failed to cope with Dembele’s hold up play, Rogic’s ability to find space, and Sinclair’s pace.
McInnes tried something different again, dropping Rooney and Stockley, and going with 3 quick mobile forwards – Burns, McGinn and Hayes. They pressed very high and were clearly ordered to “get in the faces” of the Celtic players, conceding 12 fouls in the opening 20 minutes (no bookings, surprisingly). Where this failed was in controlling the game to any degree. Completing 89 passes by half time, the Dons were starved of the ball as Brown, Armstrong, Rogic, Forrest and Sinclair out manoeuvred the under manned Dons midfield.
Once again, the Dons changed approach for the 2nd half with Rooney and Maddison coming on as orthodox 9 and 10s after 53m. Celtic however dominated the opening of the 2nd half and had 7 chances up to the 65m to kill the tie. Celtic, between the Glasgow Derby and Borussia Moenchengladbach Champions League tie, and after a trip to Dingwall, perhaps tired. Rodgers, sensing momentum shift, replaced Rogic with Bitton after 70m.
Stockley was brought on very late on 83m and the Dons reverted to a very direct long ball game. Much pinging about in the box ensued without clear good quality chances being created for the Dons. Celtic had played throughout with conviction, game-faces squarely on point.
27/11/16 - League Cup Final (Hampden Park) – Aberdeen 0, Celtic 3

Played on a slippy, wet surface, and 4 days after another humbling defeat to Barcelona, Celtic nevertheless delivered Rodgers his first Celtic trophy.
On this occasion, Aberdeen matched up with Celtic’s 4-2-3-1. The emphasis early was on keeping defensive shape, and sitting very deep. Consequently, after 8 mins, Celtic had racked up 100 passes to 2 from the Dons. Maddison provided the creativity at 10 and was involved in all Aberdeen’s early promise.
As we will see regarding chances created and shots, Aberdeen need to be ruthless to stand a chance in the final. On this occasion, the discrepancy in pass decision making was highlighted as, in short order, Maddison chose the wrong option in not playing in Rooney, and then Rogic was played in by Simunovic, and displayed quality control and finishing to put Celtic up.
With Forrest and Roberts wandering to cause trouble, the Dons lacked defensive discipline, McLean failing to track Forrest who ran half the pitch to finish for 2-0 before half time.
It was very comfortable for Celtic, but in contrast to Rodgers ability to detect momentum shifts and make changes, McInnes waited until 0-3 before shifting approach to 4-4-2 and again going more direct. Celtic allowed only 1 shot on target in the 2nd half.
01/02/17 – Celtic 1, Aberdeen 0

McInnes chose Paper to cover the Celtic Rock as his team went man for man all over the park in a high full court press, falling back to a deep block when Celtic had controlled possession. This approach requires incredible stamina and concentration and it is to Aberdeen’s credit they successfully shackled Celtic.
By half time in a tactically fascinating match, both sides had registered 3 shots off target and 0 on. Celtic had been restricted to 263 passes with Forrest starved of possession – completing 4 and losing 3. The Dons targeted Brown and Bitton who were involved in many midfield challenges, and whose passing was lower than average. Whilst Aberdeen had, in previous games struggled with the mobility of Armstrong, Bitton and Brown together, as evidenced many times under Deila, is a defensively sound but relatively immobile midfield, lacking creativity.
Characteristically, Rodgers did not hesitate to change things, adapting to a 3-4-3 at half time. Celtic also adopted their own higher press, sure that the Dons did not have the stamina to maintain their first half approach. This resulted in greater Celtic ball control, but neither side could fashion chances. Celtic took theirs, Boyata losing Taylor to meet Sinclair’s free kick. In a tight match, it was the Dons who made the defensive error and Celtic who scored from their only shot on target.
This game remains the closest template the Dons may have for final success. Albeit they were successful in nullifying Celtic, they did not create. Celtic lacked a recognised striker as both Griffiths and Dembele were out, and the aforementioned Bitton played instead of Armstrong. For the Dons, this was their first X1.
12/05/17 – Aberdeen 1, Celtic 3

Potentially the most worrying game of the series for Aberdeen. On a poor quality pitch in difficult weather, Celtic raced into a 3-0 lead with speed and clinical finishing. Dons mistakes were punished. Aberdeen were desperate to lay some psychological markers in advance of the Cup Final.
From 0-3 down they made the game a war of attrition, pressing and harrying Celtic, and bombarding the box with crosses and long, direct passes. Simunovic and Boyata coped fairly well.
The home side also ran out of steam – they have not finished any game well against Celtic apart from the first home game of the season. Celtic could bring Bitton on to add control, and Forrest to add energy and counter attacking threat, whilst the Dons blooded a 16-year-old. An utterly deflating massacre was avoided, but Celtic restricted Aberdeen to 0 shots on target in the 2nd half.
Celtic v Aberdeen 2016/17 By Numbers
All the numbers below are averages across the 5 matches Celtic have played Aberdeen this season.

Based on matches this season (not always a great predictor!), Celtic will have at least 2/3rds of the possession and twice as many shots. The real problem Aberdeen have is that Celtic are averaging 2 goals per game more than the Dons. Being 2 goals better in each and every game is a large gap to close.

Aberdeen have, on occasions, been able to break up Celtics passing flow. 471 passes are nearly 100 passes below Celtic’s average (563) in league games.
210 passes for the Dons is 20 passes more than the average SPFL opponent for Celtic. As we have seen from the match summaries above, the Dons have so far failed to reconcile both sides of the equation – can they stifle Celtic AND have enough possession to create chances?

Mirroring Celtic’s season, they have outperformed their xG. Scoring 12 goals against the Dons compared to a cumulative xG of 8.4. One would expect this to regress to the mean, however it is perhaps reflecting Celtics goal threat. With 6 players in double figures, and Rogic, Armstrong, McGregor, Roberts and Sinclair all capable of scoring from low probability distances and angles, Celtic have superior fire power to overwhelm the statistical model. Aberdeen, perhaps, do not.
Celtic have been simply more clinical – converting 17% of all shots compared to Aberdeen’s 6%. Despite having the excellent Lewis in goal, Celtic have scored with an incredible 48% of on target shots – they average 37% across the season.
Celtic created more chances and complete more dribbles although Aberdeen do a good job of restricting Celtic dribble success.

There is hope for the Dons in the fact they have pressured Celtic into 40 clearances per match, compared to their season average of 28. High clearance rates usually indicate a side under pressure. Celtic have been forced to make key defensive saves – 2.8 compares to an average against SPFL opponents of 1.3. And 3 defensive errors per 90m against Aberdeen compares to 2 against the rest of the league.
But they must take any chances resulting from forcing errors….
Summary
Celtic stand on the brink of unprecedented history. A domestic treble achieved without a single defeat. Even the Lisbon Lions never achieved that. So far, Celtic have not shown any signs of buckling under pressure. The torture of Be’er Sheva was negotiated eventually; 2 Cup Semi-finals and 1 Cup Final have been navigated with barely a drama to a collective score of 6-0; the league was secured at the first time of asking by 5-0 at the most intimidating arena in the league.
Rodgers has proved flexible and has the better hand. The Dons boss must be creative as arguably he is competing against not only a superior 11 starting players, but a superior match day 16. Perhaps Lewis and Hayes would make the Celtic bench?
McInnes has tried all manner of tactics and approaches against Celtic. He has successfully defended deep and successfully pressed high to restrict Celtic control for periods, but lost anyway. He has tried a quick, mobile front line and a direct approach to a large striker. Neither approach has resulted in sustained pressure nor chance creation.
Ultimately Celtic have the better players, and a manager who is more reactive to in game momentum changes. But in a 1 off game, and especially if Aberdeen score first to have something to believe in and fight for, there is always hope. If Lustig is not fit for Celtic, Aberdeen go in with potentially 6 tall players (Taylor, Reynolds, Considine, O’Conner, McLean, Rooney) against Boyata and Simunovic. Who knows what sort of game Mr. Madden will have?
In this week of nostalgia and celebration, of being shown what is possible with unity of purpose, these Celtic players must realise they have momentum, and that this is their time.