Esports offer a unique opportunity to conduct human performance studies, as they use modern hardware and software as an operation platform. Insights on gameplay and underlying processes may push the development of new and optimal practice methods. The aim of this study was to investigate performance indicators from in-game data to predict the outcome of the matches in StarCraft II: Legacy of The Void.

Data from 6509 games (game records) provided by 5 players at the level of Master or GrandMaster were used. The distribution of analyzed players concerning the preferred in-game race was as follows: “Protoss” (n = 3), “Zerg” (n = 1), “Terran” (n = 1). Each game record contained data for both the winner and the loser. In total, 3719 game records and 9 performance indicators were obtained after applying the inclusion criteria.

Logistic regression with 5-fold cross-validation was performed to predict the game outcome. The model was able to discriminate the game outcome (won, lost) with an out-of-sample accuracy of 0.728 ± 0.021. The performance indicators which showed the strongest effect in predicting the game outcome were “minerals lost army” [p-value< 0.001, std_odds_ratio: 0.069], “minerals killed army” [p-value< 0.001, std_odds_ratio: 6.446], “minerals used current army” [p-value< 0.001, std_odds_ratio: 4.081], and “minerals killed economy” [p-value< 0.001, std_odds_ratio: 2.896]. It seems evident that winner optimized interaction with an opponent by keeping his/her own army intact while inflicting damage to the opponent’s army or economy. In conclusion, the effective use of the army, based on optimizing the ratio between units lost and units killed, may be significant in predicting the game outcome.