Scouting and Recruitment

Finding Similar Players

By StatsBomb | July 3, 2024 | 5 min read
Scouting and Recruitment

Finding Similar Players

At StatsBomb, we are committed to providing our customers with the tools they need to succeed.

Professional teams at all levels worldwide use our StatsBomb IQ analytics platform and market-leading event data to identify players, analyse them in-depth, and curate their shortlists and squad planning ahead of the upcoming transfer window.

Among the many recruitment features currently in IQ, one that can act as a starting point for any scouting process is our Similar Player Search tool.

The concept is simple: a player is selected as the subject of the search, and the algorithm then produces a list of players with similar statistical profiles, ranked on a scale of 0-100, with 100 being an exact match. The subject could be a well-known player, a potential transfer target or even one of a team’s own players for whom they require a backup or future replacement.

Various filters can then be applied to more closely tailor the results to the requirements of the end user. These include:

  • Gender
  • Position
  • Minutes Played
  • Age
  • Competition: open selection or by region or tier grouping
  • Season

It is also possible to adjust the importance assigned to each metric. If ball retention is the number one attribute you desire in your new central midfielder, or you want a full back that is excellent at defending in 1v1 duels, you can adjust the algorithm to account for this.

The results can act as an ideal jumping-off point for a more involved scouting process, providing a list of names to investigate further. StatsBomb Data covers over 120 competitions and 75,000 players worldwide, so there is plenty of opportunity to come across some lesser-known names. That is particularly valuable for teams operating on a tighter budget.

It is important to note that the simple fact of approximating the statistical output of another player doesn’t necessarily suggest similarities in ability, potential, or physique. Think of the subject of the search as an avatar for a particular set of statistical outputs.

Let’s look at a few examples.

Andy Robertson

Though still a high-level performer when he’s on the pitch, Liverpool will likely soon start to succession plan for life after Andy Robertson, if they haven’t already. This could still be two or three seasons away, but with a couple of injury layoffs in recent seasons and a 30th birthday celebrated in March, Old Father Time may just be starting to catch up with the wing back.

Were Liverpool to be in the market for competition and a potential long-term successor this summer, who might they look at?

We know they typically shop in the Top 7 European leagues, so we’ll include those plus the Brazilian Serie A, Belgian Pro League, and English Championship, and we’ll also filter to players under 25 on 2023/24 and 2024 season data.

The top result will be familiar to some already at Liverpool.

Teams can go beyond our default positional radar templates and get even more specific. IQ allows users to create their own radar templates, and analysts can harness the StatsBomb IQ API to select the metrics to go on the radar based on those their specific player stands out in.

Toni Kroos

Let’s use Toni Kroos as an example, with his playing career ending and Real Madrid hypothetically in the market for a replacement if they hadn’t already spent the last two years hoovering up some of the elite midfield potential in Europe.

As we all know, Kroos is an unequivocally elite passer and ball progressor, and also incredibly adept at retaining the ball under pressure in his team’s half as they attempt to build up, with defensive output less of a requirement for his role.

The StatsBomb IQ API confirms this. As well as the usual pass progression and retention metrics, Kroos also stands out for:

  • Line-breaking passes to >5m of Space
  • Pressured Pass Completion %
  • Ratio of Ball Receipts in >5m of Space in Attacking Half

…compared to central midfielders in the Big 5 over the last six seasons.

The resulting radar looks like this:

We laughed too.

But can we find a player with similar output?

Again, we’ll look at the Top 7 leagues, plus Brazilian Serie A and Argentinian Liga Profesional, and under-26 players from the last two seasons.

Of the top five players in the resulting list, #1 and #3 are both Enzo Fernández. It’s easy to understand this one from a profile point of view, but it’s hard to see a transfer happening any time soon given the transfer fee and contract Chelsea invested in him in January 2023.

Plus, our expected pass model, which evaluates how often a player completes their passes against the expected completion rate based on the condition of the pass (e.g. length, height, pressured or unpressured), isn’t quite as hot on Enzo (~80th percentile) as it is on Kroos (99th percentile). That’s not to say Fernández couldn’t develop in the future though, still only being 23.

The most interesting might be player #2, given his potential to also be an heir at national team level.

Aleksandar Pavlović.

Pavlović only turned 20 in May and would’ve been in Germany’s squad for Euro 2024 were it not for illness, having made his senior international debut in June. All the statistical indicators show that he could become a player of similar style and substance to Kroos – and the passing model loves him already, rating him as a 96th percentile passer.

We’re talking about a player who’s made just 19 Bundesliga appearances, but the potential is clear. A positive 2024/25 season for Bayern Munich, and who knows if Real might decide to add another midfield gem to their collection.

Vivianne Miedema

We can also apply this analysis to women’s football, where, as of 2023/24 and the 2024 summer leagues, we cover 13 domestic leagues. Teams competing in those competitions can access StatsBomb IQ for free as part of our commitment to elevating women’s football.

Vivianne Miedema leaves Arsenal this summer after seven highly successful seasons. As such, there are some big shoes to be filled at the Emirates for a forward to play off Alessia Russo and Stina Blackstenius.

Taking the same approach we took with Kroos, the StatsBomb IQ API tells us that, as well as the usual goalscoring metrics, Miedema stands out for her ability to add value through dribbling and carrying, play passes into the box, create plenty of chances and throughballs, and rarely conceding possession.

We use Miedema’s last season with significant minutes (2021/22) for the template and look at forwards in all our women’s competitions in their most recent seasons and… look who comes in with a 93% similarity score and the most similar forward:

It’s only new Arsenal signing Mariona Caldentey(!) Some would call it serendipity, others would call it an excellent piece of player profiling by the recruitment team.

The Similar Players tool is just one of many recruitment solutions in the StatsBomb IQ platform, providing a simple way for analysts and scouts to identify players that may interest their team.

Do you work for a team and want to become more data-driven in your scouting and recruitment? Contact us to schedule a demo of our platform.

By StatsBomb | July 3, 2024