We all know the importance of software updates and patches for our phones and computers. In the world of esports, patching takes on a whole new level of significance, especially in MOBAs. These patches are often introduced to bring about changes and shake things up in the game. They can eliminate building types, introduce new heroes, or completely transform existing heroes.
However, it’s not just major changes that occur. Subtle tweaks like cooldown reductions or adjustments to item prices can also impact gameplay. The true effects of these patches are not always immediately apparent and are often observed through how professional teams adapt to the new game rules.
In the dynamic landscape of esports, patches create a challenge for creating betting odds. Side markets for MOBAs typically focus on objectives like killing certain monsters, whose properties and priorities frequently change with patches. Failing to update models accordingly leaves room for clever punters to take advantage.
Unfortunately, adjusting models to reflect patch changes takes time and requires new data. Expert analysis and human traders can fill the gap temporarily, but they are not infallible. Gathering sufficient data to ensure model accuracy for side markets can take weeks, during which time these markets either cannot be offered or carry higher risks.
Imagine if there was a software that could predict how patches would affect the meta game. This would be a game changer for the industry. One possible solution is using self-play, where a computer steers teams in games like League of Legends or DotA2 until they develop a new meta. By comparing shifts in strategies and collecting statistics, we could create new prediction models much faster than waiting for relevant matches from top teams.
While this may sound futuristic, it is not beyond reach with the capabilities of modern AI. OpenAI has already demonstrated success in teaching machines to play Dota2 at a professional level. With the right team and access to high-quality data like Bayes Esports provides, exploring patches with AI assessment could become a reality in the near future.
Dr Darina Goldin, the director of data science at Bayes Esports, is passionate about the esports industry. She has developed predictive models for popular games like Counter Strike, Dota2, and League of Legends. In her free time, she enjoys training in Brazillian Jiu Jitsu.
