Football never stops in the modern era.

No sooner is the Scottish Cup won but the transfer rumours start in earnest, followed by the European Championships and then it’s pre-season. We’ve not even had a summer holiday!

However, back to being 24/7 digital natives. Rumours are rife that Celtic are pursuing the option to purchase Paulo Bernardo from Benfica permanently. He is not in their first-team plans. That’s a red flag right there. If he is not considered good enough for Benfica’s first-team squad then why would Celtic be interested? The answer to that could be very nuanced depending on many factors including the existing personnel at Benfica and the areas they feel they need to strengthen.

For Celtic, a known quantity can be seen as less risky. The “try before you buy” model has been successful – Odsonne Edouard, Jota and Cameron Carter-Vickers are the standouts. We discussed this possibility on the latest Huddle Breakdown. Long story short, and spoiler alert, it was a “no” from me. However, a hot podcast “take” and a more studied data-driven opinion are two slightly different things. And the boy from Almada deserves a more considered perspective.

I am not going to compare Bernardo to Matt O’Riley as that will be fruitless. The Dane is a clear standout whichever way you form your opinions and if / when he leaves for bigger things, the question of “how to replace” can be addressed. For now, Celtic need a central midfield “eight”, a box-to-box all-rounder to compliment O’Riley and do some of Callum McGregor’s running for him.

I’m going to use 2022-23 Aaron Mooy and Reo Hatate as comparators. Mooy was a highly experienced but veteran midfielder who added guile, tempo and prime decision-making to Postecoglou's treble-winning side. Given last season’s injury woes, that same season saw the best of Hatate in a Celtic shirt. So, I believe those are fair benchmarks for Bernardo to be compared against as opposed to peak O’Riley.

As usual, I will combine Statsbomb data with my own. Statsbomb data covers SPFL matches only whilst mine covers ALL competitions. We are looking for all-round midfielders here so we will consider defending, ball progression, creation and goal threat.

Defending

Firstly, here’s the Mooy comparison.

And Hatate (2022-23).

A strong start for the Portuguese. He came to prominence in the Champions League campaign as a strong defensive midfielder. This data does not capture those matches but the trends are clear.  Bernardo is in the top 20 per cent in the league for the volume of pressures and counterpressures, and for subsequent regains he is in the top 10 per cent. Both Mooy and Hatate are near the bottom in the league for pressures, improving to around 50 per cent for counterpressing.

Bernardo completes around three times the aggressive actions of the other pair. Brendan Rodgers likes his teams to have the ball, and that means when without it, to press aggressively to win it back. Bernardo is very strong in this regard. If we consider the impact of all defensive actions...

Mooy and Bernardo trend similarly with the Portuguese youngster just slightly more successful in winning possession back. This aspect of Hatate’s play is arguably his weakest. A strong start, then, and last season's Champions League campaign showcased the mobility and athleticism Celtic needs in their midfield from Bernard.

Ball progression

Here are the relevant buildup profiles for Bernardo versus his two peers.

With Mooy, you could be guaranteed a high volume of passes synonymous with the control he brought to a game, allied to a range of passes – a high deep progression rate, and a high passing on-ball value (OBV) of 0.19 indicative of the quality of passing relative to improving scoring probability.

Given that Mooy’s game - especially at 32 years old - was built on decision-making and not athleticism, it is slightly surprising his ball-carrying rate is also higher than the more athletic Bernardo’s. Again, this is indicative of the canny Australian choosing his moments to drive forward wisely. An interesting vignette is that Bernardo found himself being pressured when passing more often than Mooy (again, decision-making) but was slightly more successful in his passing when pressured (more robust physical frame).

Versus Hatate, it is a very similar picture. The Japanese player’s ball-carrying is a strength and he profiles by these attributes almost identically to the Australian. The main difference is in the overall pass quality as measured by OBV with Hatate on 0.06. Bernardo’s is -0.09. I suspect this reflects his tendency to play the ball backwards and square rather than forwards as we see here.

Whilst Mooy (24.14) and Hatate (22.67) take out more than 20 opponents per game with their forward passing, Bernardo manages a mere eight. This single data point has been my main concern that Bernardo will not be the player Celtic needs. This is a exceptionally low packing figure for an ostensibly attacking midfielder in a dominant side like Celtic. This one feature would be a showstopper. Can Rodgers coax him into being more aggressive in his forward passing?

Getting the ball into dangerous areas was also a Mooy strength although his numbers are boosted by set piece deliveries. Hatate provides slightly more passes into the danger zone (central to the goal inside the box) and generates a higher expected assist rate for secondary assisting passes (the pass before the one that creates the chance). Mooy excelled in both fields.

If we consider chance creation again Mooy generated a higher volume of chances and a higher average expected assist value per 90 minutes (0.4). Bernardo’s and Hatate’s numbers are similar but perhaps tellingly whilst the Japan international generates more chances, Bernardo’s are of slightly higher quality. Mooy has 10 years of experience on Bernardo, so we also need to consider that factor.

Goal threat

Finally, let’s consider the goal threat comparison. Mooy took penalties, so we will focus on non-penalty xG to ensure a fair comparison.

These views are more nuanced than what we have seen above. Bernardo can generate more shots than Mooy, and a much higher overall xG although the Australian’s xG per shot average is better. Surprisingly Bernardo provides the higher volume of through balls despite Mooy being the more creative in all other respects. This is probably reflected in the shot OBV being -0.14 for Bernardo and -0.06 for Mooy. Celtic players have relatively many shots when considering the league, and I suspect this “dampens” the averages.

Again, Hatate and Mooy profile similarly and therefore Bernardo’s strengths are similar relative to Hatate’s. he generates more shots and a higher overall xG but also a significantly higher average xG per shot.

Across all competitions, Bernardo generates slightly more shots (2.29) and has the highest overall xG (0.28) per 90 minutes. He has scored 13 goals across Portugal’s national under 15 to under 21 sides and this is an aspect of his game we can perhaps expect to see develop.

One final view is the final third wastefulness. That is, the propensity to give the ball up in the opponent's defensive third without a chance, shot, corner or possession being maintained. The other attribute is packing turnovers. That is the impact of being turned over in terms of how many of your own team do you take out of the game.

By both measures, Bernardo is “safer” in the final third with fewer unproductive possession losses and by far the lower pack turnover score. This is a particular weakness of Hatate’s and Mooy was only slightly less wasteful.

Summary

These are my favourite articles as I managed to convince myself of a different position than I held when I started it!

I am still concerned that a fee reported in the region of £6 million is too much for a player who - whilst valuable in the squad over 50 matches - only completed the equivalent of 16 matches. Essentially, a squad player. Both Jota and Carter-Vickers were nailed on starters for similar money.

In terms of performance, Bernardo is a significant upgrade on Hatate defensively and likely a much safer option in Europe. He also provides more goal threat with underlying data suggesting a higher quality of shot taken.  My main concerns are midfield ball progression and specifically a chronically low pass packing rate. He simply must contribute more to getting the ball through opposition lines if Celtic are not to get bogged down domestically against low blocks.

Hatate is erratic but there are the “magic moments”. Mooy was calm, probing and consistent. Mooy was a highly experienced international player. Bernardo has the equivalent of 88 matches of career minutes across Benfica’s A and B teams, Celtic and a loan to Pacos Ferriera. What is his ceiling?

I suspect that Rodgers’s obsession with ball retention and pressing acumen are behind the continuing pursuit of the Portuguese. He also sees an uptick in midfield athleticism with him in the side, which will be vital in Europe. There is sufficient goal threat to be acceptable. Can he develop the ball progression and carrying skills needed to complete the package?

If Celtic can agree a significant reduction in the previously agreed option to buy, the deal starts to make a lot more sense.