Loanee Paulo Bernardo passed the magic (statistically significant – there is no magic in data analysis) 900-minute mark in Celtic's home victory over Ross County the past weekend. This is for all competitions.

Rather than do an isolated piece, it is worth comparing him to Reo Hatate as it looks like that will be the choice manager Brendan Rodgers faces on the Japanese player's return from the Asia Cup. Callum McGregor is the captain and heartbeat, whilst Matt O’Riley has been Celtic’s most impressive performer this season. That leaves just one midfield berth.

We need to keep in mind that 33 per cent of the Portuguese players’ minutes have been in the Champions League whereas for Hatate in 2022-23, his time on the field at that level accounts for only 14 per cent of his minutes.

Statsbomb radar

We’ll start with the StatsBomb radar comparison, which is a nice visual. Remember this is SPFL Premiership matches only.

The first impression is that we are looking at two very different players. The new man scores heavily on all the defensive metrics. Most striking is the 37.53 pressures compared to 17.91 from Hatate. This is even more pronounced if we consider Champions League matches. In 2022-23 Hatate managed 10.65 pressure per group stage match compared to 36.56 from Bernardo. At that level, the Portuguese is in the 99th percentile across all teams in the competition – an elite ball presser at the European level.

READ MORE: Why is Celtic's Kyogo Furuhashi struggling this season?

Bernardo’s volume of tackles and interceptions is also significantly greater, and he is much less likely to be tackled or dribbled past. Contrary to that, Hatate has significantly better deep progressions and overall passing ‘on ball value’. Indeed, Bernardo is one of the worst in the whole league for passing according to Statsbomb’s ‘on-ball value’ metrics. He is no better in the dribbling and carrying the ball domain, but their respective assist rates are not too dissimilar.

Let’s dig deeper into those skill areas.


The pressure statistics are particularly discriminatory as regards their specific skill sets. Bernardo averages 12.82 defensive actions per 90 minutes whereas for Hatate that is 9.4. You simply get more defensive solidity with the Portuguese.

Here are the defensive actions success rate and possessions won from defensive action metrics:

Again, in terms of the success in defensive actions and winning back possession, the Portuguese is a clear favourite. Winning 57 per cent of his duels, there are no forwards or midfielders with a better success rate this season. Hatate won only 41 per cent of his aerial and ground duels. They are a little closer in terms of ball recoveries with Hatate retrieving 9.12 balls per 90 minutes and Bernardo 10.1.

Ball Progression

For ball progression, we’ll look at pack passing rate (the number of opponents taken out of the game) – forward passes that take out at least one opponent – and ball carries – running with the ball either past an opponent or at least ten yards forwards inside the opponent's half.

Again, there could not be a starker contrast. Hatate takes over 22 opponents out of the game each match, yet Bernardo is around five. Hatate has nearly six times the volume of ball carries per game. It has perplexed me that Bernardo is so unproductive in this area given how attack-minded his basic data looked especially for the Portuguese international age grade sides.

Bear in mind that Hatate is not particularly prolific in this area compared to Aaron Mooy and Matt O’Riley, based on last year's data. We have seen glimpses from Bernardo – his chipped-through ball for Greg Taylor’s goal at St Mirren was an excellent piece of vision and execution. We just need to see much more of it.

Deep Creativity

For this category, I am comparing passes into the danger zone per 90 minutes (passes that are complete into the centre of the opponent's box) and secondary assisting pass xA (expected assist).

Again, Hatate is easily the more productive in setting up establishing passes and getting the ball into dangerous central areas. To be fair, Hatate has often been deployed deeper in the team than Bernardo, who is often the one pushed up whilst O’Riley supports McGregor more directly.

Chance Creation

We’ll consider both chance volume (chances created) and chance quality (chance xA).

Whilst there is quite a distance in the volume of chances created with the Japanese player almost doubling Bernardo’s efforts, the overall quality is more similar. Indeed, the average xA of the chances Bernardo creates is 0.15 whilst Hatate’s average is 0.14.

Where Hatate has been missed is that he was able to connect successfully with main striker Kyogo Furuhashi. Last season he provided five direct assists and 11 chances overall. This season Bernardo has only managed to provide one chance for the talismanic forward. This failure to get Kyogo into the game has been one of the key Celtic weaknesses this season.

Goal Threat

Considering again volume (total shots) and quality (total xG) we get:

Bernardo has the edge both in the volume and quality of shots attempted. His average xG per shot is lower though at 0.12 compared to 0.13 last season from Hatate. Bernardo has 3.65 touches in the opposition box compared to 3.25 from the Japanese.

In terms of overall attacking threat, their actual scoring contributions are identical at 0.56 goals and assists per 90 minutes. However, Hatate has the higher expected scoring contribution (0.5) compared to Bernardo’s 0.41.


I like to end on a negative!

Wastefulness looks at the volume of possessions that each player gives the ball up in the final third without a positive outcome for the team. Also, I map this with pack turnovers – the volume of occasions the ball is lost putting your teammates wrong side of the ball relative to your own goal.

This is Hatate’s Achilles heel. He tends to give the ball away in dangerous areas and often. Bernardo is less so, but not so infrequently that it offsets his rather sparse overall packing scores.

Net all packing activities (passing, receiving, dribbling, recovering, and minus turnovers) Hatate averages an overall packing score of 97.32 versus 53.51 for Bernardo. Hatate is better at passing the ball forwards and receiving the ball in space but tends to give it away in costly areas. Bernardo is safer on the ball, but we have yet to see the expansive and aggressive forward passing.


On the face of it, we are considering who partners McGregor and O’Riley in central midfield. However, what we have is two very different players. Again, Bernardo’s data is far more influenced by having to compete at the Champions League level than Hatate’s. Given 85 per cent time against domestic Scottish opposition, the Portuguese’s numbers may look different, especially the creative ones.

READ MORE: Why Celtic's 'scattergun' strategy COULD be their undoing

If the need is for expansive forward passing, preferably from deeper areas, but with the downside of defensive fragility and the risk of high-cost turnovers, then Hatate is your man. If you want pressing intensity, defensive solidity, and goal threat but with minimal progressive and creative passing, then Bernardo is your man.

They have very different skill sets and arguably playing them on the same team would be complimentary with Hatate deeper and Bernardo more box-to-box - which would be another dilemma!