…… dealt with in due course! The final piece analysing Celtic’s 2-0 win at Ibrox considers the attacking performances. An overall tactical and team performance assessment can be found here-> Derby Delight for Dominant Celtic An assessment of defensive performances can be found here-> Boyata and Simunovic Reunited: Derby Defensive Analysis
Green Bhoy Packers
The terms “packing” and “impect” are probably jargony and annoying but if you can read this-> Concerning “Packing” and return (I’ll wait). I hope you agree it provides a language and framework to assess passes (and dribbles) that are beyond the mundane – that is, passes that bypass or take out at least one opponent. “Packing” opponents eliminates them from the defence and therefore, in theory, should make it easier to score. Some players are more creative than others and we can now measure that. Playing high risk passes may result in poor pass completion stats, but if you have created scoring opportunities, the reward is surely worth the risk. I measure the number of passes and dribbles that take out at least one opponent (Packs), the total number of players bypassed (Pack Rate) and finally the Impect – assigning a score to the pass based on whether you took out defenders, midfielders or forwards, and how many. At Ibrox, the Pack Stats:

Armstrong completed 23 “packs”, accounting for 19% of all his passes, and includes 2 successful dribbles. Not one for the aimless square ball our Stu. This resulted in 40 opponents being bypassed and an Impect score of 83, the highest on the day for Celtic. In addition, Armstrong’s passes created 4 chances (4 shots resulted directly from his passing) – again the most in the team. As well as his assist, a whopping 38% of Roberts’ passes “packed” opponents. He created a further 2 chances. “Packing” rewards the movement, space creation and control of the pass receiver. In this regard, Griffiths’s game is improving hugely this season. He led the team with 7 passes received that took out opponents, for an Impect score of 67. Three of those passes were from Lustig, who is the master of the long ball forward. Lusting was 3rd behind Armstrong and Griffiths in overall Impect, on 70, providing 12 Pack Passes – the most in the team. But back to Armstrong. He was ranked 2nd in Pack Passing, 1st in Pack Dribbling and 3rd in Pack Pass receiving. Back to his best, he was the game Pack Man through an all-round display of midfield play.
Top CATs
I introduced a new aggregate metric called CAT Score, and it is described here-> Top CATs. In summary it aggregates a number of key attacking events to assess overall threat to the opposition goal.

No surprise that the one out-and-out striker carried the greatest threat, but bear in mind he was only on the pitch 80 minutes. Over 75% of all on field events are passes. Yet with Griffiths, he completed more attacking threat actions (12) than he completed open play passes (8). He led the way with 3 shots on target, and 3 shots inside the box (unusual for Griffiths) from 5 shots in total.

Griffiths considering material for his new curtains
Griffiths also had the most possessions in the box (6). The second goal, hit right footed on the run from Roberts perfectly weighted pass, killed the game as a contest. Detour Starts------------> xPts As an aside, the previous statement leads to the thought “what is the worth of a goal”? I try to address that with a concept called Expected Points (xPts). xPts considers only three factors – whether you are the home or away side; the time of the goal; the game score when the goal is scored. It does not, therefore, consider the relative strength of the opposition. This would require in-game match odds as calculated by the bookies, which is market sensitive information and not available. So, long story short, I calculate xPts for every goal which gives a value for the number of points the goal gives you on the journey to 3 points. That is, the model assumes the winner of any game gets 3 points (theoretical for cup games) and assigns a “Point” value for each goal given the factors I have defined. Based on those factors, Rogic’s goal was worth 0.8 points, and Griffith’s goal worth 0.7 points. Another way to think about this is that each team starts with 1 point as the game state is 0-0. Rogic’s goal took Celtic to a probability of winning 1.8 points, which is crudely a 60% chance of winning. Griffith’s goal took that to 2.5 points – an 83% chance of winning. It would have been higher if Celtic had a) been the home side and b) the goal had been later in the match. Over the season this will also allow me to asses who scores the most important goals. Scoring the 4th goal in the 90th minute of a 4-0 win effectively adds 0 to the xPts, for example. But scoring the only goal of the game in the 85th minute will generate a high xPts. Detour Ends-----------> Back to CATs. Sinclair made some eye-catching dribble attempts, but only had one shot, off target, and created 1 chance. He possessed in the box 5 times, the second highest. Roberts is less eye-catching, but created 3 chances including 1 assist. His one shot was on target and he had 4 possessions in the box. Forrest’s on fire. His 21 minutes generated 1 shot on target and 2 attempts, 4 box entries and 1 key pass. Another take away is that threat from Celtic runs throughout the team. Both Lustig and Tierney contribute to attacking output in different ways.
Scoring Contribution
I consider both Actual and Expected Scoring Contribution. Scoring Contribution is very simple equalling Goals plus Assists. It is not very insightful and you can get that from the newspapers – Rogic and Griffiths scored 1 goal each and Roberts got 1 assist. More interesting is xSC or Expected Scoring Contribution. This aggregates Expected Goals (xG) and Expected Assists (xA) to provide an overall view on what would normally have been expected from the chances created and efforts on target.

Griffiths leads the way and his 2 chances created and 3 shots on target had a cumulative xSC of 0.491. Next came Armstrong with 4 chances created and 1 shot on target, and xSC of 0.250. Nine Celts either had a shot on target or created a chance.
Bhoy of the Match
Note that Brown had the 8th highest Impect score – his passing tended to be short and safe. He also contributed 0 xSC. I mention this due to the ongoing online storm (well, a bit of debate on a couple of forums) regarding who the Bhoy of the Match was. Also, please read the defensive analysis article referenced at the top as this accounts for the bulk of his performance given his role. There were many good performances including an influential one by Brown. Simunovic was the outstanding defender; Lustig not far behind and was creative especially with his long passing; Sinclair and Roberts sparkled; and Griffiths done what Griffiths gone done and do. What a great problem to have? Having watched it back and analysed the performance data, it is Armstrong for me as Bhoy of the Match. (*ducks*)
