The Impact of Game Live Streaming on In-Game Purchases of Chinese Young Game Users

This study examines how viewing game live streams relates to in-game spending among young Chinese gamers, situating the question within a rapidly expanding market. China had 654 million smartphone game users and 266 million live-stream viewers in 2022, and its smartphone game market grew 17.51% to roughly $320.65 billion in 2023, with MOBA, shooting, and role-playing games together accounting for nearly half of top-100 revenue. These genres therefore form the focus of the investigation, which is set against a global gaming market projected at $183.9 billion, led by the Asia-Pacific region at 46% of revenue.

The empirical core rests on a web survey of 856 valid respondents aged 20 to 40 in China. Weekday live-stream viewing time showed moderate positive correlations with both spending frequency (r = 0.362) and spending amount (r = 0.407), with comparable but somewhat weaker associations for playtime. In logistic regression, weekday viewing time emerged as the strongest predictor of spending likelihood (β = 0.704, p < .001), followed by age (β = 0.362, p < .001), while gender had no significant effect. Taken together, these results indicate that live-stream viewing is more closely tied to spending than direct gameplay is.

A subsequent analysis combining logistic regression with a structural equation model reinforces this conclusion. Viewing time exerted the largest positive effect on both spending frequency (0.759, p < 0.001) and spending amount (156.528, p < 0.001), and play time also contributed positively. However, the model's RMSEA of 0.213 exceeds conventional acceptability thresholds, signaling poor overall fit and the need for model refinement, so the structural conclusions should be treated with caution.

The authors interpret live streaming as an important driver of in-game purchasing, while acknowledging significant limitations. The findings are correlational rather than causal, the sample is self-selected, drawn from online respondents, and restricted to Chinese youth, and the data rely on self-reported behavior. These constraints mean the results are best viewed as evidence of a strong association that warrants further investigation rather than as definitive proof of causation.