Can an algorithm find you the perfect right back?

Imagine a league with 20 teams, each requires a right back and a pool of the best 20 right backs are gathered together, with the clubs, in a room.

Each club submits it’s list of the 20 players in ranked order from most to least preferred.

Each player submits their list of the clubs they are prepared to play for given the money on offer and the prestige of the club.

A matching process takes place.

Each player, in turn, is provisionally placed with a club.

If the club and player both ranked each other as first choice, for example. Trent Alexander Arnold chooses Liverpool, and Liverpool choose Trent Alexander Arnold, then the match is confirmed.

It is more complicated when people have different preferences, for example the 20th placed club may have 19 players they’d prefer ahead of the 20th rightback, and the 20th right back might prefer to play for 19 other clubs ahead of the 20th best club. But ultimately they are the best match.

In Economics this is called the Stable Marriage Problem and there is even an algorithm to show you how it works.

Basically everybody eventually ends up with the best possible match. The club won’t end up matched with a player who wants to move to another club, who also would prefer the player over the match they now have.

So, what has a Nobel prize winning matching algorithm about marriage got to do with football?

It shows us just how inefficient the transfer market is, and by how removing some of the barriers you can create better outcomes.

In our example earlier we effectively had a pool of free agents, a room full of the same number of clubs as players and the aim was that every club would be matched with a player.

Barriers:

(1) Contractual status of players – we don’t have perfect information about what the player is earning or the transfer fee the club would ask for the player

(2) The player doesn’t know which clubs want him – he may be at the absolute best club that would play him, but he might not.

(3) Club’s don’t know which players would play for them if asked. That brilliant young prospect in Barcelona reserves might be desperate to come and play in the Championship. You simply don’t know if they cannot express an opinion.

Problem:

So in the textbook example, you have complete openness of information. Everybody is there specifically to find a right-back (or get married). There are no middlemen, no demands for additional money at the last minute. It is a straightforward opportunity presented in a simple way. And this algorithm has proven extremely popular. It is used for allocating people onto their preferred course at their preferred institute the world over.

But in football we have middlemen and we have patchy, at best, information about the intentions of the parties. Does the player really want to move here or is it leverage for a new deal? Does the club really want this player or is it a smokescreen for a preferred option they are working on?

Solution:

Create a market within the market with better information

MRKT Insights are creating a network of like-minded clubs.

This network will allow clubs to optimize the use of their playing staff, creating development pathways.

Where one club has squad members who aren’t playing they will be offered the opportunity to move within the network. By allowing ranking and ordering of preferences the optimal solution can be found for the player and club. Information is clearly available, there are no middlemen or misunderstandings. Players and clubs can have a good long look at each other before committing to the deal.

The loaning/selling club are able to reduce their wage bill.

The borrowing/buying club get a player they want, and who wants to join them, who will improve their team.

The player gets to play for a club they want to play for and who wants them.

MRKT Insights LLP work with the network members to coordinate scouting and recruitment to allow the best possible players to be recruited, with the best possible development pathways, to have the best possible outcomes.

We currently have clients in three countries and are growing all the time. For more information contact@mrktinsights.com

Leave a comment

Your email address will not be published. Required fields are marked *