It turns out the more you play, the more you are trapped in a love-hate relationship with your favorite games—especially when multiplayer and paid titles are involved, according to a recent study.

On August 4, researchers Chetan Mudlapur and Om Prakash Singh from the Indian Institute of Information Technology Bhagalpur (IIIT Bhagalpur) published a study in the International Journal of Human–Computer Interaction (IJHCI), empirically analyzing what they call the 'Playtime-Sentiment Paradox'.

The team scoured 559,290 English reviews across 60 titles and 12 genres on Steam, employing the RoBERTa model for sentiment analysis. They took care to scrub the data, removing empty reviews, logs where total playtime was less than the time at the moment of writing, and extreme outliers exceeding 10,000 hours, which often suggest botting or idle farming.

The researchers segmented playtimes into five tiers: Early Quitters (0-2 hours), Trial Players (2-10 hours), Engaged (10-50 hours), Loyal (50-200 hours), and Veterans (200+ hours). Their findings debunk the simple notion that more playtime equals higher satisfaction.

On average, satisfaction peaks around 50.2 hours before beginning a slow, steady decline in an 'inverted-U' trajectory. Early quitters, unsurprisingly, provided more negative feedback, likely before forming a proper impression of the game. However, the 2-200 hour window saw negative sentiment drop to under 20%.

The 'Veteran' group, however, flipped the script. Negative reviews surged to 28.24% once playtime exceeded 200 hours. This 'Paradox Group'—those writing negative reviews after immense investment—accounted for 5.24% of the total dataset, or 29,308 entries. With an average of 976.2 hours and a median of 517.6 hours, these are clearly the players most deeply immersed in the games they claim to dislike.

War Thunder topped the list for these paradox-heavy titles with 24.66%, followed by Stellaris (20.87%), Marvel Rivals (19.88%), Dota 2 (18.56%), and Tekken 8 (18.11%)—mostly online, multiplayer-centric affairs. Conversely, titles like Sekiro, Black Myth: Wukong, Doom, and Code Vein saw zero paradox reviews, while single-player experiences like Subnautica, Golf With Your Friends, Hades, and Persona 5 Royal showed negligible figures. Interestingly, those who paid for their games were statistically more likely to leave these paradox reviews (5.40%) compared to free-to-play users (3.15%), a phenomenon the researchers attribute to sunk cost fallacy and cognitive dissonance.

Ultimately, this study suggests that developers and platforms need to look past basic metrics. Veterans write significantly longer reviews (averaging 330 characters) compared to the 257 characters of early quitters. Rather than dismissing this as mere 'churn', studios should recognize these as critical, albeit bitter, insights from their most dedicated player base.