Why Celtic should be open to a Giorgos Giakoumakis transfer - James Dailey

In Giorgos Giakoumakis’s relatively short tenure at Celtic, he has scored 18 goals across all competitions in just 1,650 minutes of action.

The Greek striker is due to turn 28 next month and, with a huge Champions League and domestic campaign dead ahead, surely the idea of selling him this month is crazy? Welcome to crazy town!

One of the concepts in financial market analytics is assessing the performance of securities relative to the overall market versus the idiosyncratic properties.

To put it more simply: how much of the return is due to a ‘rising tide’ lifting all boats versus the specifics of the boat itself? A related idea in real estate is that it can be better to buy the worst house in a good neighbourhood than the best house in a bad neighbourhood.

Relative to Celtic and player performance data, the thesis on this occasion is that the ‘market’ is the interplay between ‘Ange-ball’ and the significant financial disparity between Celtic and 10 of their other 11 domestic league opponents.

These generate a sort of ‘rising tide’ which offers a platform for attacking players to post eye-watering statistics.

For example, here is a radar comparing some widely-used traditional attacking metrics for Giakoumakis and fellow striker Kyogo Furuhashi for all league minutes to date playing for Ange Postecoglou:

We can see from this comparison how similar the various metrics have been, including the number of shots, touches in the opposition box, expected goals, and also xG per shot.

Indeed, even in his very limited minutes last season, Albian Ajeti’s output was similar to what Giakoumakis has offered (with the vital exception of goals scored):

Daizen Maeda – and even Liel Abada in his single game up front – have posted very high output while playing the striker position in Postecoglou’s system domestically.

If the system is providing a huge platform for these players, then how does one attempt to assess the ‘value’ a player might be offering?

First, we can review each player’s performance history prior to their arrival at Celtic, as well as looking at their performance output versus higher-level opposition.

Giakoumakis’s historical data via Wyscout data includes 12,251 minutes across all competitions, in which he scored 70 goals on 75.18 xG. He has finished very well during his tenure at Celtic so far but how much of that is idiosyncratic and how much is systemic?

Here are shot maps for him and Kyogo:

We can see from the maps how close and central Giakoumakis’s shots have been, with Kyogo offering similarly with the addition of more varied locations. Both have enjoyed quality finishing rates since arriving, with Giakoumakis’s post-shot xG of 11.14 and Kyogo’s 14.83.

That introduces another variable into the performance measurement equation – strikers are not the sole factor in goal scoring.

The relative quality of play from opposition keepers is also vitally important. To date, Kyogo has faced better shot-stopping compared to Giakoumakis. 

Next, let us compare the two using on-ball value (OBV):

Here is where we begin to see some signs of idiosyncratic value for Kyogo, as his OBV has been fairly pervasive across various aspects of play, whereas Giakoumakis has been almost all finishing. 

If we take a closer look at Giakoumakis’s track record so far at Celtic, his goal total has been extremely impressive and it is certainly not unusual for Celtic strikers to pad their stats versus lower-quality competition. His tally includes 13 goals against Alloa Athletic, Raith Rovers, Dundee, St Johnstone, Ross County and Motherwell combined, with three versus Dundee United and two against Hearts. 

In his 571 total combined minutes – admittedly not a large sample – in European competition at Celtic and AEK Athens, as well as in derbies versus Rangers last season, he has scored zero goals. Expanding this record include elite league games in Greece and the Netherlands versus PAOK, Olympiakos, AZ Alkmaar, Feyenoord, PSV and Ajax, the total comes to four goals scored in 1,518 minutes or 0.24 goals per 90 minutes.

In contrast, Kyogo has played 1,345 minutes combined between last season for Celtic in Europe and in the Asian Champions League, averaging 0.60 goals per 90 minutes. Add in the six goals scored in the 1,088 minutes he played in the J1 League versus Kawasaki Frontale and Postecoglou’s former club Yokohama F Marinos and the total becomes 0.55 goals per 90 minutes over the 2,433-minute sample.

Fans like Giakoumakis and enjoy rooting for him. However, the weight of the evidence is significantly tilted towards his performance value generally being compiled versus weaker opposition.

Like a good portfolio manager, a good player trading model requires the ability to know when to ‘hit the bid’ – i.e. sell.

If the price is comparable to the going rate in England, then it should not be difficult to identify a comparably motivated and skilled player as a replacement to produce versus domestic opponents.

Better yet, Celtic could use the proceeds to sign another Kyogo and in the process improve both their depth and ability to compete in Europe.