Packing Them In

How effective is a pass? How does the data help show who the most creative players are? Which players combine most effectively? Using the concept of Packing can help answer those questions.

My analysis is limited to players who have played more than 7 matches worth of minutes. It was a natural cut off based on squad appearance data. So no Ralston, Miller, Sviatchenko, Benyu, Musonda, Johnston.

Please read Concerning Packing to understand the concepts being discussed here. Essentially, the most effective passes are often forward passes and we can measure those in terms of how many opponents those passes “take out”. By take out, this means that before the pass the opponent was between the ball and their goal when Celtic have possession. And after the pass they are in front of the ball. That is, they have been removed from defending by the pass. The more players you take out the better, and taking out defenders is more likely to lead to chances than taking out midfielders, so we score the passes accordingly.

The score of how many opponents you have taken out is called the Impect (ugh, I know, blame the Germans who invented it).

But at the most simple level the more opponents that can be bypassed the less opponents are defending the goal therefore the higher chance of creating something.

Total Pass Impect – Top 7

These are the players with the highest overall Impect scores – i.e. their passes have taken out the most opponents during the season.

The reason for a Top 7 is these are the players who have accumulated over 1000 Impect points this season. Below Armstrong it is a 357 drop to Simunovic. Brown led with 400 forward passes that took out opponents. Totals reflect partly the number of minutes on the pitch, so we need to assess the per 90 averages for a more accurate picture.

Pass Impect per 90m – Top 8

The eight players averaging over 30 Pass Impect per 90m are:

These are the players who break the lines from deeper positions and find the creative talent in the final third.

Before the season started and before I started recording this I would have predicted Armstrong as being the Pass Pack King (or Pack Man). However, injury and loss of form blighted his season, and Ntcham emerged. Whilst I believe they can play together, they have many similar skill sets. Being able to break midfield lines with passes foremost amongst them.

Nearly 10 passes a match from Ntcham take out at least one opponents and the average number of packing passes per 90m soon falls away to 5 or less illustrating how rare such passes are given Celtic play over 550 per match.

Bitton splits Ntcham and Armstrong and this shows his value to the team playing deep. With more time and space playing in defence, Bitton’s rather languid style flourishes in possession. Especially in domestic games where Celtic dominate the ball, having such a deep lying play maker adds to Celtic’s ability to break teams down. As I have shown elsewhere, in Europe, his defensive error rate makes this too much of a risk.

I hope Lustig critics can start to see how his long passes especially are a real benefit to the team. The diagonal pass in behind defences is a useful alternative to the “horseshoe of despair” – the endless recycling across the back and down the sides. Lustig has a better range of passing that I think many give him credit for.

Boyata is also surprisingly high. I have made the comment before that he is the worst passer at the club and this is proven by his incredibly high passing stats – he completes 95% of all passes the joint highest in the squad. That obviously sounds counter intuitive. What I mean by that is he keeps it very simple and reduces risk in his passing. He does this because he is not a good passer. However, a distinct improvement this season based on having the 5^th best Pass Impect. Passing through opponents is more risky than passing in front of them based on the potential for interception. Boyata does not seek to find attackers (we’ll look at team passing connections later) but tries to advance the ball to the likes of Brown, Ntcham, and Armstrong.

Interesting to note Kouassi has very similar Impect rates to Brown.

Finally, Tierney’s numbers are impressive because much of his passing is in the final third. Again, I will cover this in more detail later.

But overall, Ntcham is the Pack Man!

Pack Passing % – Top 10

The following players all by pass opponents with 10% or more of their passes. Not all players get the ball as much as central defenders and central midfielders. This metric normalises differences in possession.

Again, we see the value of De Vries as a ball playing goalkeeper. Despite not being as agile or commanding as Gordon, 17% of all his passes take out opponents. 31% of his passes are incomplete. This is the second “worst” percentage in the squad. The corollary is he is attempting relatively high risk passes to reach as far forward as possible.

After that once again Ntcham and Armstrong are locked together. Both bypass opponents with 15% of their passes. They are the players most likely to get Celtic forward from midfield.

Roberts and Rogic both score 13% underlining their importance when available, which sadly has not been enough.

Lustig may lack pace but he is effective getting the team up the park, which he does with 11% of his passes.

Finally Brown deserves mention as he has completed 12% of ALL passes Celtic played last seasons. And 10% of them took out at least one opponent.

Defenders Packed

Passing around pressing strikers and attacking midfielders is relatively easy. What we want to see is opposition defenders taken out the game. Take out defenders and you are really disrupting the opponent’s defensive organisation. These are the players that are doing it where it matters.

Surprisingly it is Tierney that has packed the most defenders – 117, 19 more than the next highest Ntcham. But it is Rogic who bypasses the most defenders on average and is the only player to pack more than 3 defenders per 90m. Rogic plays as the most advanced midfielder but is operating under pressure with little time and space.

The subtle differences between Armstrong and Ntcham are highlighted by Armstrong being the second highest packer of defenders. Armstrong operates in a more forward position than Ntcham and although he packed 74 defenders, 24 less that Ntcham, he achieve more per 90m than the Frenchman.

The ability of Roberts has been missed greatly over the season. Despite his sparse appearances he makes it to 3^rd in the list per 90m.

Ahead of attacking talents such as Sinclair, Forrest, McGregor and Hayes, Tierney is the 4^th highest defender eliminator. Bypassing the most defenders overall (117), this is remarkable for the nominal left back. He is the last player on the table to achieve an average over 2 defenders per 90m.

We’d expect the defenders and goalkeepers to be near the bottom as well as the forwards. After all they are the furthest forward and are more likely receivers of forward passes than generators of them. The disappointment is perhaps Sinclair who achieved less defenders bypassed per 90m than Kouassi. He is probably more a player to get on the end of through balls, to be fair.

Finding Strikers

Ideally the team wants to get the ball to the strikers as they provide the highest goal threat. Which players are most adept at finding the striking talent? There are four players that stand out in their ability to get the ball to the central strikers. The values shown are total Impect scores when passing to strikers.

This shows that Tierney has taken out more opponents to connect with the strikers than any other player. Surprisingly Lustig is second, his trademark diagonal helping set up goals in two consecutive Scottish Cup Semi Finals.

If we look at the % of each players Pass Impect that is accounted for by passes to strikers there are a few surprises. The following players have at least 1/3 or more of their Pass Impect accounted for by passes to strikers.

Two goalkeepers are in the top 3 and Gordon is at 22%. This is probably a combination of small samples for Bain and De Vries, the fact ‘keepers make relatively few passes and finally that they are often forced to go long and will often aim for the most advanced target i.e. the striker.

Another surprise is Kouassi whose successful forward passes predominantly find strikers. Again his is a small sample but he found passes to Edouard, Dembele or Griffiths far more than any other players indicative of a direct but accurate passing style. It’s certainly an interesting outlier.

Also fascinating is Edouard. The fact 46% of his Pack Passes are to a striker hints he is effective as a second or deeper forward. He links effectively with a more advanced central striker. I thought his most effective performances were playing instead of Sinclair and with Dembele.

Tierney and McGregor again emphasise how effective they have been connecting the Celtic attacks.

Connections

Speaking of which, whom connects with whom most effectively to take Celtic forward? The diagram below takes the players in each position who have played the most and shows the number of successful Pack Passes between them. The thicker the line the higher the average Impect score between the two players.

There is a lot going on in this diagram and here are the highlights.

1. Tierney and Sinclair! Despite teams doubling up on them, man marking them, and Sinclair’s form dipping in the second half of the season, the pair connected for 568 Impect points. The highest between any pair by miles. 31% of Sinclair’s Pack Passes are to Tierney and 23% of Tierney’s to Sinclair.
2. McGregor – an absolute ball magnet and joins the team together like no other. The deeper players find McGregor evenly and regularly and he finds all the attacking players equally too. He is the perfect “linker”.
3. Celtic are a very left biased team. Just let your eyes compare the lines moving centre and right to left and vice versa. Predominantly line splitting passes go down the left.
4. However, Dembele is more often connected to by the right side of Forrest and Lustig. The left side sees Sinclair as the target.
5. Brown plays a lot of Pack Passes but they tend to be shorter and to the next line of attack.
6. Dembele’s most common target for forward passes is Sinclair.
7. Ntcham is the Pack Man but has a heavy bias to passing to the left side.
8. The most Impect-ful connections are Tierney to Sinclair scoring 341 and Nthcam to Tierney scoring 305.

From the players not shown on the diagram there are some other interesting observations to be drawn.

1. When Rogic plays he equally finds Sinclair (18%), Dembele (16%) and Forrest (18%) with his Pack Passes.
2. When the French trio play together they link up effectively. Ntcham found Edouard with 9% of his Pack Passes despite them rarely playing together relative to the other strikers. 27% of Edouard’s Pack Passes were to Dembele.
3. Armstrong and McGregor combine particularly effectively. Armstrong had McGregor as a recipient with 17% of his Pack Passes and McGregor reciprocated with 12% of his.
4. Griffiths has a radar finely tuned to Forrest’s runs. 29% of Griffith’s Pack Passes found the wide man.
5. Lustig loves Griffiths runs when he spins in behind defences and he connected for 13% of his Pack Passes. Lustig to Griffiths is the 9^th most prolific connection despite Griffiths rare appearances. AND despite Lustig being a defender and Griffiths a striker – i.e. with a large distance between them.
6. Emphasising the left bias, neither Ajer, Boyata nor Simunovic have Forrest as a top receiver. Ajer and Boyata connect with the central midfielders whilst Simunovic often connects with Sinclair.

Summary

Measuring forward passes using the Packing concept tells us who the most effective players are at unlocking opposition defensive organisation. Ntcham does this most effectively from a deeper midfield position. Rogic is the most adept at taking out defenders. He also manages to connect across the front line evenly. This is helpful as Celtic remain a team biased towards passing down the left-hand side channels.

Finally, McGregor’s fine season is seen through yet another lens as the fulcrum of the side in terms of being available to take passes and then connecting with other attackers across the team.