More Games, More Gains? What LoL Data Says About Practice

When you're playing, you try to be as good as you can be. However, depending on your rank and server, your practice may look different. The study we'll discuss today looks at the differences (if any exist) in practice behavior across different server regions and ranks. Here's what they found.

With the increase in gaming and esports, the professionalization of competitive gaming has caught up to the hype. You don't just see it on your screen when watching a tournament, but also in the number of training facilities run by esports orgs, the amount of support staff for players—such as coaches, analysts, and psychologists.

"Consequently, innovation in performance optimization, talent identification, and expertise development is now at the forefront of professional organizations looking to succeed at domestic and international levels. [1]"

(⬆️ That, btw., is also the area my research revolves around.) However, in order to get into a professional team in the first place, your average gamer Adam (apparently that's one of the most common names for characters in video games—so, you know, close enough) is typically left to practicing by himself in an unstructured way. How Adam typically practices is unknown and the focus of today's study. How does Adam's practice behavior contribute to his expertise compared to players outside a professional or even semi-professional tier?

In order to examine what's going on, the researchers pulled data from League of Legends (Riot Games). They analyzed the last 100 ranked games for 913 players from the EU, NA, KR, and CN servers in the ranks of Emerald and above. Here's one remark I'll make, I'd argue that the word "practice" may not fully be appropriate, especially among lower-tier players. I could ̶e̶a̶s̶i̶l̶y̶ be Emerald and just playing for fun as—I furthermore argue—as I'd argue most Adams, even in Masters, do. They play for fun. Anyway...

💎 Are you practicing like a Challenger — or just playing like Emerald Adam?

"Challenger players (highest tier) across all servers (except Europe West) tended to have the greatest number of ranked solo/duo queue matches per day, whereas Emerald I players (lowest tier) had the least (excluding Japan and Vietnam)." [1]

That shouldn't really surprise anyone. Challenger players are typically professionals that see playing as practice and have, because it is their job, more time to do so. On the other hand, Adam—who is 35, just came back from his double shift, and has a family—has way less time and ambition to "practice". On average, Challenger players played 4.7 ranked games, whereas Emerald Adam did only 3.4. Korea, btw., had the highest average number of matches per day.

When looking at the game length, Emerald Adam and his team's matches lasted 27.5 min. on average, whereas higher-tier players were quicker (25.4 mins. at the Challenger tier). Another aspect where Emerald Adam and Challenger players differ is the amount of time played—Adam played about 1.4h, whereas pros put in 1.98h of practice.

"The average of most days without a match increased as tier decreased (Challenger = 4.0 ± 2.0, Grandmaster = 4.4 ± 2.1, Master = 5.2 ± 2.0, Emerald I = 5.3 ± 1.9)." [1]

This finding would support my assumption of Emerald Adam having less time to play. It also shows in the amount of matches played within one week. Challenger players had, on average, 53.7 (and up to 69) matches on their clock, whereas Adam "only" played 25.4 (up to 34). The figure below visualizes some of the findings across tiers and servers.

When analyzing the "practice" behavior more closely, multiple clusters emerged.

• Cluster 1: large blocks of play within a short period and few days without play (pros and Adam?)
• Cluster 2: short blocks of practice and more days without playing (Adam?)
• Cluster 3: "had occasional very large one and three-day blocks, but the player may go many days without practice." [1] (definitely Adam)

Thanks for reading, and I hope you all have a great week. Cheers,

Christian 🙂

References

[1] Bennett et al., 2025