A Game Theoretic Study on Hearthstone Match Formats

Match Solver

Experiment with the HearthNash Match Solver yourself! You may find it useful for optimal strategies in your own Hearthstone matches, or to explore how different metas and formats interact.

Research Paper

Our research on Hearthstone match formats led to a variety of interesting observations. Read all about the methods, results, and discussion in our paper - Generation and Analysis of Optimal Strategies for Hearthstone Match Formats.


You will find a condensed presentation of this research on Hearthstone match formats in our HeartNash poster.


Hearthstone is a 1-vs-1 digital card game that is played competitively in a variety of formats. In a tournament, each player preconstructs several decks and then chooses at each point in the match what deck they will play, according to the restrictions of a chosen match format. Such tournaments measure players' skill and are exciting for viewers, but can take place in a variety of match formats which fans claim drastically affect the competitiveness and viewer engagement. Given that different deck archetypes have certain expected winrates against other archetypes, an optimal strategy exists for each player that determines with what probability they should pick each action available to them. We developed software that can solve for optimal strategies under fixed conditions and apply this process to winrate sets to generate metrics that can be compared across formats. We then used a large dataset collected from live Hearthstone players to generate metrics that reflect the competitive deck ecosystem, and used these results to draw qualitative conclusions about various match formats.

Source Code

GitHub Repo Mirror