The one 'huge but subtle' Celtic improvement behind Greg Taylor's rise to prominence - Alan Morrison

The sentiment is of a player whose limitations, both physical and technical, were apparent to all as a young recruit from Kilmarnock but who is perceived to be improving and thriving under Ange Postecoglou.

That qualitative assessment is just begging to be tested using the data.

Defending

My default assessment method for defending capabilities plots a lot of data aggregated into two key metrics.

Both are a mouthful, for which I apologise. So let me try and put it plainly:

Defensive Action Success Rate (%) – this looks at all defensive actions (e.g. tackles, aerial duels, interceptions, fouls) and indicates what percentage were successful

Possession Won from Defensive Action (%) – this tells us of all defensive actions, what percentage resulted in Celtic having possession. For example, you might ‘win’ an aerial duel, but the ball goes to the opposition.

It is a very reliable method for assessing centre-backs. A warning on full-backs: at Celtic, they tend to get forward a lot and so are prone to losing tackles in the opposition half, which dampens the numbers. These metrics do not differentiate where on the field the event happens.

However, since we are only comparing the Scotland international with himself, this should not be an issue.

Admittedly the scales make this more dramatic, but, in essence, the 24-year-old is no more defensively effective in his actions than in previous seasons. 

In fact, last season, he was winning the ball back with around 62 per cent of his defensive actions compared to 57 per cent this season. 

What it does say is that Taylor scores reasonably well by these metrics as a full-back and his defending is not an area most fans would be overly critical of him on.

A noticeable trend would be to lose possession in your own defensive third increasingly. All defenders do this more under Postecoglou due to the style asked, but the former Killie man was coughing it up 1.04 times per 90 minutes last season, and this has reduced to 0.84 this term.

This may be an indicator that the support sees he is playing the way the manager wishes and adapting effectively. 

Ball progression

This will plot overall passing accuracy – the Scotland international is a relatively high-possession player – as well as pack pass score (the score accrued from forward passes that take opponents out the game).

The 24-year-old's pack passing score has gone down slightly this season compared to last. Both campaigns saw big increases on previous seasons, mainly due to the way the manager wants the team to play by being more aggressive in their forward passing.

One must remember the standard of opposition is a lot higher this season in a relatively small sample – this holds true for all these comparisons.

Where Taylor has improved is in simple pass completion or accuracy. He is now completing 86 per cent of his passes from open play.

This may not sound exciting, but that equates to over 20 passes more completed this season. This was a recurring bugbear about the player when he joined from Kilmarnock. The better the quality of opposition, the more hurried and inaccurate his passing would become. 

The left-backs composure in possession, under pressure in his own defensive third, has been a huge area of improvement.

Watch him against Shakhtar Donetsk, for instance, where he often played out a high press effectively. When he first joined, it would have been a hurried hoof up the park.

His pass completion average in the Champions League has been 86 per cent - the same as in domestic football. 

There were doubts that Taylor could scale his technical play to this level, but this shows that in the most high-volume activity (passing), he can perform as effectively at the highest level as he does domestically. That is undoubtedly a huge improvement.

Creativity

The job of a Celtic full-back means getting up the park and delivering service into the box.

To measure this, we will plot volume (number of chances created) and quality (expected assists) per 90 minutes.

We can see that in the two seasons under Postecoglou, his creativity has actually decreased.

Reasons for that may be the changing role of the full-back in this system. Inverting often means playing the pass from a more central position to advanced midfielders as opposed to bombing on outside the winger and getting crosses into the box.

The Australian likes his more creative talents in midfield to form the wide overloads rather than it always being the full-backs.

Taylor has two goals and two assists this season, so the perceived improvement is not being driven by outcome bias in this regard.

StatsBomb

StatsBomb ranks players across the SPFL – something my own data cannot do.

Here are the radars for Taylor over the last two seasons and the percentile ranking indicated per attribute:

As you can see, there are some tiny improvements this season, but this chart demonstrably is showing data for the same player. 

It is perhaps surprising that the number of pressures has decreased. A potential explanation for this is that, with Champions League games in midweek, Celtic have been slightly more conservative in SPFL matches.

That is, however, probably another piece for another day.

Back to the Scotland international. StatsBomb’s bespoke amalgamated on-ball value (OBV) metrics do not show any improvement from him either.

Indeed, they claim a significant reduction in both his pass OBV and defensive action OBV in SPFL games, although without knowing their basis of calculation, it is difficult to comment exactly why that is.

SofaScore, meanwhile, provides a single-game ranking for each player. In the two prior seasons, Taylor’s match average was a 6.97 out of 10.

This season he is averaging 7.08 and is ranked 20th overall in the league, and is the eighth-highest-ranked Hoops player.

Summary

So, back to the hypothesis: has Taylor improved as sentiment would indicate?

Not according to StatsBomb and only slightly according to SofaScore.

In this analysis, there is usually some glaring areas where the player in question has improved. 

The 24-year-old, however, is a far more subtle tale.

The major area of improvement from him has been his pass completion rates. While this may seem mundane (and is not accompanied by an uptick in the sexier chance creation data), it addresses a key area of weakness from him since he joined.

Taylor was not someone considered to be particularly secure in possession, and this worsened the higher quality of opposition the Parkhead side faced. 

What we have seen from him, especially in European football, is increased confidence and ability to keep possession under duress and not give the ball away cheaply. 

In addition, what fans see is a player continually looking to learn and improve. He also is one of the fittest players at the club, and supporters appreciate his non-stop style of play.

While we cannot say he has massively improved over the range of performance, he has got better in one key and fundamental area, and it is perhaps that which is underpinning the continuing positive sentiment towards him.