We need to talk about Celtic midfielder Reo Hatate

Was this the Ryan Christie replacement fans craved? Rested against Alloa Athletic, he exploded into Scottish football consciousness by battering the ball past Craig Gordon as Celtic gained an important victory at Tynecastle.

A week later, a global television audience saw him record two Glasgow Derby debut goals as Ange Postecoglou’s Celtic sent shudders of doubt through the current Premiership champions. Two outside-the-box goals swung momentum in the title race; here was a player not only capable of midfield multi-tasking but also of influencing big games.

Since then, though, no goals have followed and only two assists – the second of which dubiously attributed on Sunday when his shot off Allan McGregor led to Tom Rogic’s equaliser.

More important has been the impression that he looks a little fatigued and even bewildered by the rough and tumble of Scottish football.

Now that Hatate has accumulated 10 matches worth of data, it's time to dig into his performance numbers in comparison to the other number 8s. 

We have included Christie’s data from the 2019-20 season (his best) as the benchmark for the type of midfielder that would thrive in Postecoglou’s system. Callum McGregor is included as he is such a talented all-rounder and often plays 8 as well.

Defensive

This plots defensive action success rate with possessions won from defensive actions percentage.

No surprise that McGregor rates highest by defensive metrics, as this is more the job he is asked to do. Also, no surprise that Rogic and David Turnbull are respectively less effective. Both are often found more advanced in '10' roles. However, this does illustrate their limitations for the all-action style Postecoglou’s approach demands.

Of the more box-to-box 8s, Hatate lags behind in defensive action effectiveness with both Matt O’Riley (did we tell you how good he is?) and peak Christie.

Deep progression

Let’s now look at progressing the ball from deep.

We measure this by looking at 'pack' passes – passes that are forward and take opponents out of the game – and ball carries (progressive runs + pack dribbles).

Hatate has the fewest ball carries of the group and slightly more pack passes than Rogic and Christie. In terms of total volume of productivity by these measures, he would rank bottom.

Chance creativity

Chance creativity is illustrated by mapping expected assists (xA) – the quality of chances created – with chances created, the volume of chances created.

O’Riley, Turnbull and Christie have similar volume and quality of chance creation. Rogic creates fewer chances but they tend to be good ones.

McGregor’s numbers are depressed due to his deeper role. Hatate’s creative numbers trend him more as a defensive midfielder than attacking.

Goal threat

This is measured with xG, of course, as well as the volume of shots taken.

Hatate is in the position of taking quite of few shots but they are generally of low quality. This is the problem with scoring a few goals outside the box. Remember Stuart Armstrong? Seven goals outside the box one season and one the next. xG overperformance breeds overconfidence.

Hatate again trends closer to the defensive-minded McGregor than to the more threatening trio of Rogic, Turnbull and Christie. O’Riley’s one slight disappointment is his relative lack of goal threat so far.

READ MORE: Why Maeda and Giakoumakis personify Ange Postecoglou's team-first ethic

Attacking threat

Finally, we aggregate attacking threat in two ways. Expected scoring contribution is xA + xG. CAT Score is Celtic attacking threat score and is an aggregated count of goal-threatening actions.

The other main reason for including McGregor in this analysis is to benchmark against what a defensive midfielder or '6' looks like versus the '8' role.

We see again here that Hatate’s overall attacking threat data has him aligned with McGregor more than the other 8s in the sample.

O’Riley has the highest volume of attacking threat actions per 90 minutes while Christie's xSC in 2019-20 was an exceptional 0.87 (i.e. he was expected to get 0.87 goals/assists per 90 minutes). That's terrific for a central midfielder.

SPFL benchmark and O’Riley comparison

Here is StatsBomb’s comparison of Hatate and O’Riley, who both joined at the same time and fulfil the same role:

O’Riley’s performances have been excellent and consistent. Hatate really only scores highly on expected assists by dint of playing for Celtic. We’ve seen when benchmarked against his team-mates, his numbers are not so impressive. What you will also notice is that O’Riley is 21 and Hatate 24.

That was a lot of sidebars but, hopefully, it some insight. Let’s go back to Hatate.

Conclusions

To summarise all the data above, we have a player whose overall performances trend more like a defensive midfielder than the all-action template the 8s play in Postecoglou’s Celtic. Only, his defensive data is not as strong as McGregor’s.

All that may not sound encouraging. The eye test says there is a player here with all the raw materials. Quick over the first few yards and so capable of getting out of traffic; a wide range of passing, much with creative intent; an ability to score important goals in important games; intelligence shown by the movement that led to the cross for Liel Abada’s goal in the February derby.

What we also need to consider is the human aspect to all of this. A brickbat thrown (usually unfairly) at data analysts sometimes is that there are real people behind the spreadsheets.

In Japan, professional footballers often come through the university system. Education is highly valued in Japan while this would be a very unusual route in Britain. Remember, Frank Lampard was considered some kind of boffin in football circles just because he managed a few A-Levels (ex-footballer Daddy was well off and he went to an independent school, lest we forget).

But not so in Japan. This means, of course, that players don’t typically play for professional clubs until they are 20 or 21. Hatate joined Kawasaki Fontale in 2018 while in college, but didn’t play for them continuously until 2020 when he was 23 years old.

Prior to joining Celtic, Hatate had played the equivalent of 56 matches of professional football. Furthermore, 21 of his appearances had been as a central midfielder.

This is not a seasoned box-to-box midfielder but a relatively inexperienced young man still learning his trade. He is also adapting to a totally new culture, weather, professional environment, employer, language, food and so on. Not all players adapt at the same rate to all these new variables.

Christie was also included in this analysis for an ulterior motive. Christie made the Inverness Caley Thistle first team at 18 years of age. On joining Celtic at 20, he had only 50 matches worth of professional football under his belt, six fewer than Hatate.

He spent most of his first three seasons at Celtic being loaned back to Caley Thistle and then Aberdeen. By the time he became a regular at Celtic in the 2018-19 season, he had another 50 games worth of minutes under his belt and was 23 years old.

The conclusion? That we need to be patient with Hatate and let him grow into his Celtic career. I suspect next season the midfield will have other more athletic competition and there will be less pressure on Hatate week on week.

He has the tools for the job and can eventually flourish with care and nurturing, just as Christie did.