The lessons of Edmond Tapsoba

Timing. The secret of comedy, staying onside, and a vital component in transfer market success.

Anyone can spot a good player, playing well. If you can’t do it by eye then data can pick it up for you very simply.

What they can’t accurately measure is the potential of that player. Are they playing at the best possible level they can already? Would they look as good (or even better) in a different system or in a different league?

Getting that right can be worth millions of pounds.

A good example is Edmond Tapsoba.

Our system flagged him in Portugal’s second division playing for his club’s reserve side last season.

Now obviously Vitoria will have known they had a very good player on their books and would not have sold him in the summer for a low fee.

However I very much doubt they would have expected to sell him for £18m+bonuses just 16 league appearances later.

We would look at data like this and see a player who is very comfortable progressing the ball both through passing and running, but also a highly accurate volume passer. His penalty goalscoring record shows a player who has self-belief.

This transition from the second to the first team was seamless, and possibly a good example of the merits of B teams.

So given the data was identical in the second division to the first should a team have moved in for the player last summer?

Most top 5 league teams would not even consider moving for a player at that stage. They simply would not be confident spending money on an unproven player.

But what is a proven player?

Are those 16 league games (plus some European games) enough for a player to be “proven”? It is a fairly small sample size in a league where team quality is variable.

Are any players currently playing in non-top 5 European leagues proven?

Is there a way we can measure which leagues are producing players who, when they join a Premier League club, go on to regularly play minutes for them?

One of the things we’ve been working on is a model to assess success in transfers. Our hypothesis is that players who are playing well remain in the team. Of course there are other reasons, such as injury, why a player may not feature but in general a manager will select the best players available to him.

We, therefore, have created a model whereby we look at the league players have come from and the number of minutes they average per season after signing for a club.

Looking at signings made since 2015, domestic signings from within the Premier League (164 signings,1024 mins average played) and in the Championship (150 signings, 805 mins) dominate in terms of volume.

Excluding them we see:

The colour of the dot indicates the average fee paid for the transfers, green indicating a lower average fee, and red higher.

The top right quadrant contains purchases from the top divisions of France, Spain, Germany, and Italy. This is where the majority of Premier league teams buy their foreign players from.

It is probably fair to say that the more expensive the signing the more chances a manager will give them a chance to prove themselves. However, these players are the ones playing the most minutes, the proven quality.

An interesting discovery is that players signed from Portugal have been very successful in the Premier League with the average signing averaging more minutes per season than players purchased from Ligue 1 or Serie A despite a lower average transfer fee.

Breaking this down further we can look into Liga Nos (Portugal 1) and see the success rate of signings into that division.

What we discover is that transitioning from Portugal 2 to Portugal 1 is successful with 169 transactions averaging 900 minutes of game time. Pretty much identical to transfers within Portugal 1. This excludes moves from B teams to A teams.

When we include B teams we find that a significant proportion of the current Portugal national team have played in the second division, with an average of 45 games at that level.

Given that Portugal 2 to Portugal 1 is a successful route, and Portugal 1 to the Premier League is a successful route, does it follow that Portugal 2 to the Premier League would be successful?

It would be a very brave manager who put a player from Portugal 2 straight into a Premier League line up.

There is no point in signing Tapsoba from Portugal 2 if you don’t play him. Sure you could sign him, put him in the U23 team, give him the odd run out in cup games but ultimately the pace of his development will likely slow.

So is it just better to pay the premium?

Let’s make the assumption that you could have got Tapsoba for £6m pre-debut in the first team. Vitoria are not a tiny club, they are well run, and competitive within their league. They knew the potential of the player.

Are you better to gamble on that potential at £6m, or wait 6 months and pay £18m when you are confident you have a guaranteed starter?

Certainty is worth a premium once you reach an elite level.

But what about teams trying to break into the elite?

Once a player is a certainty it is almost certainly too late for you to buy them.

This works at all levels. That player with great data in League One is no longer an option once they’ve backed up that performance in the Championship.

A good way of thinking about building a football team is trying to think of it in a similar way to a company building up its product line. The classic example of this is the Boston Consulting Group’s Growth Share Matrix.

The basic idea is that you want a mix of proven products in proven markets and also want to take the risks of launching new products into growing markets.

For a football team you could relabel the X-axis as “Evidence of ability” and the Y-Axis as “Potential to improve”.

Your “Stars” are the high potential players who have shown evidence of their ability.

The “Cash Cows” are proven players who have done a good job at the level you are playing at.

“Dogs” are players who aren’t good enough and have low potential to get better. These are the players you want out.

What we are interested in are the “Question Marks”, these are unproven players who have been assessed as having high potential.

But how do you turn a Question Mark into a Star? We see there being two solutions available.

The first is to make sure the people who assess potential are the best you can afford at your level. In our modern recruitment department article we looked at how the key people are the decision-makers. They know the system the club want to play, they know the options in place, they have a real stake in the club’s future.

If you have a “trusted eye” who really understands what you need then they need to be making the final decisions. But they can be massively helped in that by companies like ours who can provide data, first filters, video scouting, and player shortlisting. The more time they can dedicate to the final decision the better, we can help with getting them that time.

The second is to join a network of clubs with a similar development ethos, a similar playing style, and a similar coaching style. You can take the “Question Mark” player at an earlier stage and get them game time at the right level for their current ability.

That way a player can be system, style, and environment proven before taking the final step of moving to a higher level.

Club networks are a core focus of ours. We see huge mutual benefit in clubs working together to maximise the use of the squads. These can and should be made up of clubs at different levels to get the most benefit from them and to build in the player pathways required.

This doesn’t guarantee you’ll get the next Tapsoba, but taking players early, and critically not then blocking their development, is a hugely important part of improving the efficiency of your recruitment.