Thanks to Moneyball – Michael Lewis’s bestselling, eye-opening book on the baseball team Oakland A’s use of analytics – sports analytics in popular perception is all about player evaluation and recruitment. Even though I am not particularly fond of that tag, I love writing transfer recommendations in my blog posts at thefutebolist and other sites. It is the most fun, easy and practical thing to do in soccer analytics.
While writing them, I follow a three-step plan. In this article, I explain my plan. Here’s the first step.
1. Underline a few rules.
Always make sure you only look for players who are:
- Under 25 years old: In the transfer market, it is crucial to stay on the right side of the aging curve. Buying players who are close to a spike in their performance is better than hiring players who are already in their prime or are just outside of their prime and will decline soon. Plus, younger players are usually less expensive.
- Not playing in the Premier League: The Premier League is the most popular and competitive league in the world. This also means that the clubs are rolling in money and don’t need to raise funds selling players. Also, more scouts from other (bigger) clubs will probably be scouting the same prospects you’ve found because of the popularity of the Premier League. All this drives up prices.
- Not from a big team: This is roughly similar to the earlier rule. Players on better sides will put up better stats. The other issue is that table-topping teams don’t need the money from the sale of good players. For example, it would have been easier to sign Ademola Lookman from Charlton Athletic than signing Jese Rodriguez from Real Madrid. This rule may not apply if the player is from a slightly smaller side and isn’t getting playing time or is out on loan, like Michy Batshuayi, Bertrand Traoré or Memphis Depay.
These are the standard rules that I follow for every article. Of course, other pre-scouting rules can be created according to a club’s situation.
2. Move on to statistics.
Of course, this is coming from me as a soccer stats enthusiast. I could give a one-hour talk on why stats are essential to recruitment in sports. But the main reasons are as follows:
- It’s way, way cheaper than live or video scouting.
- Instead of scouting several random players, you can just scout the players who show good results in your KPIs (key performance indicators) or scouting algorithm.
- Objective scouting reports using stats are not subject to cognitive biases or opinion.
There a couple of things you need to sort out while scouting with stats. First and foremost, you need to make sure the KPIs you’re using are predictive, repeatable and relevant to the position of the player. Second, the difference in style of leagues should be accounted for. For example, a defensive midfielder playing in a high-pressing German side may make more interceptions and tackles than a defensive midfielder in a reactive Italian unit. Biases also exist in data, and identifying and solving these biases are important.
3. Look at other factors.
In other words, apart from looking at the cost of the player, ask other questions which are important to the signing of the player, such as:
Is the player injury prone?
Check how many games and minutes the player has played over the past few seasons. Has he missed any games due to injury? If so, what injury did he suffer? Is that an injury that may come up again? These questions are important as some injury-prone players can bounce back and stay fit.
Will he fit into our tactical system?
This is not easy to do with stats alone, but look at the tactics and style of the player’s team. Was the player at home in that system? Then take a look at the player’s style: how much a striker dribbles and how many key passes/xA he’s putting up can tell us his style of play.
These are just two of the many post-scouting factors that must be taken into account while signing players, and the factors to be considered will differ from one club to another. By now, the list prepared using statistics will have been narrowed down.
In the end
These are the steps I follow whenever I write a blog post with a transfer recommendation. I will publish a couple of transfer dossiers soon that use the same method. These steps are not perfect, of course. Send in a comment below or hit me up on Twitter (@thefutebolist) if you find anything you don’t agree with.
The next step that I should follow is video scouting, but I am no qualitative analyst. But I’m learning. It’s also easy to present your claims with video clips to support your data (or the other way around). Another thing is that you can present any narrative with video.
— Ted Knutson (@mixedknuts) January 14, 2017