Nov 10, 2024 4 min read Gaming Science

Team History & Success: Skill isn't Everything!

Winning in DotA 2 might depend more on your team history than individual player skill.

Hello hello,

Another Sunday, another episode of Gaming Science (#75). Today, we'll discuss the influence prior shared success of team members has on performance. Furthermore, we will answer the question of whether individual skill is the most important factor in team performance. Enjoy. 🙂

💡 Highlights
• It is a commonly held belief that a team's success is linked to the presence of highly skilled players.
• Teams with fewer kills but a higher number of assists were more likely to win.
• "... teams with more successful previous co-play relations than their opponents were more likely to win [1]."
• In DotA 2, the expertise of the players is NOT the biggest factor in a team's success, raising the question: what is? 🤔
• The effect of prior shared success on a team's performance was modest in basketball, football, and DotA 2, but higher in baseball and cricket.

❓ What is Prior Shared Success?

“The idea of star players is a notion everywhere but nonsense in Germany.” - Hienric Spencer

This challenges the commonly held belief that a team's success is linked solely to the presence of highly skilled players. Looking at sports history, there are countless examples where teams with great players have failed to deliver. By now, the frequent reader of this newsletter know that a team's performance depends on many factors, such as individual talent, teamwork, and a support structure.

Another aspect that may play a role is prior collaboration (how often the players have played together and whether those games were successful—win/loss). These two elements of prior collaboration (referred to as prior shared success) could affect the future performance of a team.

💽 DotA 2 and Data

The authors collected data from 4,357 DotA 2 matches. To analyze the potential impact of prior shared success on future performance, the data were gathered over a two-week period.

🤯 Those are some Banger Results...

First, the study found that winning teams had a lower number of deaths on average. Similarly, teams with lower death rates and higher assist rates were more likely to win. These results shouldn't be a surprise to any gamer: fewer deaths mean fewer resources (gold) for the enemy team, and spreading resources (gold from kills) among the entire team (through assists) makes everyone stronger (because everyone gets to buy better items instead of just one player); you don't want to put all of your balls into one basket.

"... prior shared success had a significant positive impact on the outcomes of a match. That is, teams with more successful previous co-play relations than their opponents were more likely to win... [1]"

Second, the authors confirmed their what they had expected: players who played together and won were much more likely to win again. Interestingly, the effect size (the strength of the impact) was quite large (1.401), meaning prior shared success contributes "a lot" to a team's win percentage.

Third, the study also compared these results with other sports. What they found was that the impact of prior shared success on the odds of a team winning varies depending on the sport. In basketball, football, and DotA 2, the effect was modest, while it was much stronger in cricket and baseball.

"[Lastly,] in four out of five datasets (that is, with the exception of Dota2), talent plays the largest role in determining team success... [1]"

This is quite a banger result: in DotA 2, player expertise is NOT the biggest factor in the team's success, raising the question: what is? 🤔

"[To conclude,] we demonstrated how past successful interactions (prior shared success) significantly improved the odds of a team winning in basketball (NBA), football (EPL), baseball (MLB), cricket (IPL) and online games (Dota2) [1]."

🤔 What can We Learn from It?

I think there are two key takeaways for gamers who want to win more often. First, add players you've won with to your friends list and queue up with them again rather than relying on random queues. However, keep in mind that this will place you in a different (team) queue. You might be able to circumvent this by queuing up at the same time as someone on your friend list, increasing the chances of being put together in the same match.

Second, this mechanism may work against you. If you lost a game or two with someone on your friend list, he may be not willing to play with you again and instead role the dice again. What I mean by that is that queuing up instead of playing with someone you just lost games with may be perceived as a better chance of finding a better team/team mates. Don't be sad if people ghost you.

I hope you enjoyed this episode. Have a great week,

Christian 🙂