Ninety-five minutes. That is the average session length for Baldur's Gate 3, measured across hundreds of thousands of cloud gaming sessions. Fire up Rocket League instead and the number collapses to a few minutes. Same player, same couch, same free evening. The gap has nothing to do with your willpower and everything to do with which icon you clicked.
We tend to think in accumulated hours. "I played too much this week." But the hour is a strangely blind unit. It says nothing about how your evening was cut up — and that cutting is exactly what determines when you can reasonably stop. The length of a match, the spacing between save points, the duration of a run: someone designed all of it, deliberately, before you sat down.
Session length is a design decision
That is the central finding of a study presented at CHI 2026, the leading academic conference on human-computer interaction. Thomas Byers, Martin Gibbs and Bjørn Nansen, from the University of Melbourne, conducted twenty semi-structured interviews with game professionals working across AAA, indie, mobile and live-service studios. Their paper earned a Best Paper Honourable Mention.
What those developers describe is not accidental. Player time is raw material, shaped with dedicated tools and tracked by dedicated metrics. The authors name the tension at the heart of the craft: on one side, user-centered temporal design, which structures time around player autonomy and lived experience; on the other, data-centered temporal design, which structures time around metrics such as session length and retention.
Put plainly: "session length" is not only something observed after the fact. In some studios, it is a target. Daily quests, seasons and battle passes are the instruments aimed at that target. None of them forces your hand — but together they set the shape of your evening long before you sit down.
Genre decides more than motivation does
The numbers show how wide the gap runs. A network traffic analysis by Yifan Wang, Minzhao Lyu and Vijay Sivaraman, drawing on hundreds of thousands of cloud gaming sessions collected over three months at an internet provider, compared thirteen popular titles. Baldur's Gate 3 topped the list at 95 minutes per average session, with Cyberpunk 2077 close behind at 82 minutes. At the other end, Rocket League and CS:GO posted the shortest durations in the sample.
The logic is obvious once you see it. A role-playing game advances through dialogue, cutscenes and zones to cross. Its natural unit is long, and interrupting it costs something. A competitive game built on short matches hands control back to you every few minutes. Between those poles sit run-based games — roguelites — offering an intermediate unit: a complete arc, won or lost, that closes cleanly.
This minimum unit of play is the real variable. The question is not "how long do I feel like playing," it is "how long is the smallest slice I can finish without losing anything." If that slice runs ninety minutes and you have forty minutes before bed, you do not have the choice you think you have. You have a choice between not starting and running over.
What if hours never measured what we assumed?
Here comes the uncomfortable part. Counting your gaming hours looks like the responsible approach by definition. Recent research suggests it is mostly the easiest approach to implement — which is not the same thing.
In 2022, Matti Vuorre, Niklas Johannes, Kristoffer Magnusson and Andrew K. Przybylski published a landmark study in Royal Society Open Science. Rather than asking players to estimate their own playtime — a notoriously unreliable exercise — the researchers obtained objective behavioral data for nearly 39,000 players, supplied directly by seven publishers, across six weeks. Their conclusion: there is little to no evidence of a causal connection between time spent playing and well-being.
Three years later, an expanded version of the same team pushed further. Nick Ballou and his coauthors analyzed data from 703 casual adult Nintendo Switch players, covering 150 games and more than 140,000 hours of play, cross-referenced with self-reported well-being. The title says the essential part out loud: perceived value of games, not hours played, predicts mental well-being. How a person judges the place gaming holds in their life matters more than the quantity they consume.
This does not mean time never counts. It means the raw total, stripped of context, is a poor indicator. A two-hour evening that ends when you decided it would and a two-hour evening that overruns because the next save point was still far away are not the same experience — yet they produce an identical line in your counter.
Genre also shifts the risk profile
A set of preregistered studies published in 2023 in Frontiers in Psychiatry by Strojny and colleagues points the same way on more clinical ground. The team asked what predicts gaming disorder risk beyond time spent. Genre preferences surfaced: role-playing games (β = 0.15 in the first study, 0.19 in the second), MMORPGs (β = 0.14) and shooters (β = 0.16) emerged as predictors.
Honesty requires quoting the limit the authors state themselves: the final model explains only 11% of the variance in the first study and 8% in the second. That is modest. Genre is not destiny, and nobody should reshelve their library by presumed danger level. But the order of magnitude supports a simple observation: if you want to understand your relationship with games, genre carries more information than a total number of hours, which explains almost nothing.
Reading your library differently
There is a practical consequence here, and it is a liberating one. If the length of your evenings is largely a property of the games you launch, then changing it does not require extra willpower. It requires looking in the right place.
Three questions cover most of it.
- What is this game's minimum unit? A match, a run, a chapter, the stretch between two save points. Time it once, honestly. That is the true cost of entry.
- Does this game hand control back to me? Some titles close cleanly at any moment; others punish interruption with lost progress. That difference is not a comfort detail — it decides whether you stop when you planned to.
- What does my library look like by genre rather than by total? The same thirty hours spread across twenty-minute sessions or across four full evenings tell completely different stories.
That third question is the one conventional counters answer worst. A Steam library displays cumulative totals, not session shapes — we covered elsewhere how to read your Steam library honestly and what those numbers actually contain. Cross-referencing hours against the genre of titles played, instead of simply adding them up, often changes the diagnosis entirely: it is not "I play too much," it is "I play games whose unit does not fit inside a weeknight."
The same reframing applies to a calm look at your Steam playtime set against the rest of your week — not to impose an arbitrary ceiling, but to see which titles cost a whole evening and which adapt to whatever time you have.
Choose the shape, not just the quantity
The practical conclusion is an inversion. Rather than deciding in advance how long you will play — a decision the game will renegotiate without asking — decide what format your evening can hold, then pick a game whose unit fits inside it.
Forty minutes before bed? A short-match game, or a roguelite whose run closes on its own. A whole evening ahead of you? That is the right moment for the role-playing game you have been postponing for three months, precisely because it demands that format. The same weekly total, distributed according to the real shape of each game, produces far less frustration.
People looking to reduce their video game time usually gain more from changing the structure of their sessions than from setting themselves a quota. A quota can be renegotiated. A twenty-minute run cannot — it simply ends.
The reasoning echoes what we see on other dimensions of play, where fine-grained measurement proves more revealing than raw volume. Our piece on reaction time and what esports reveals about the brain illustrates it well: there too, what matters is not the pile of accumulated hours, but what happens inside them.
FAQ
What is the average video game session length?
There is no useful cross-genre average, and that is precisely the point. Across the thirteen titles analyzed in the cloud gaming study cited below, the range runs from 95 minutes for Baldur's Gate 3 down to a few minutes for Rocket League or CS:GO. Averaging them means adding units that are not the same kind of thing.
Why are some games so hard to put down?
Usually because their minimum unit is long and interruption costs progress. Developer interviews conducted for the CHI 2026 paper show these stopping points are design decisions, arbitrated between player autonomy and engagement metrics. A game that closes cleanly at any moment is a game built to hand control back to you.
Is tracking my gaming hours actually worth it?
The raw total, taken alone, predicts well-being poorly. The Vuorre study, run on nearly 39,000 players using objective data supplied by publishers, found no clear causal link. Hours remain useful as a factual reference point, provided you read them with context: which games, which session shapes, and above all how you actually felt.
Are some genres riskier than others?
The preregistered studies by Strojny and colleagues identify role-playing games, MMORPGs and shooters as predictors of gaming disorder risk. But the models explain only 8% to 11% of the variance: that is a weak signal, not a verdict. No genre makes anyone ill, and playing an RPG is not a problem in itself.
How do I adjust my evenings without quitting games?
Start from the time you actually have, then pick a game whose minimum unit fits inside it. Time the match, the run or the gap between save points in your usual games just once: you will know which ones cost a whole evening. It is an adjustment of format, not a deprivation.
Ready to look at your evenings differently?
Watching the shape of your sessions rather than their total is often the moment you stop judging yourself and start understanding. Kantise helps you cross your playtime with the rest of your habits, with no rankings and no imposed targets.
Sources
- Byers, T., Gibbs, M., Nansen, B. (2026). "From Quarters Per Minute to Daily Quests and Seasons: Developer Perspectives on Temporal Design in Video Games." Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems. Read the study
- Wang, Y., Lyu, M., Sivaraman, V. (2025). "Games Are Not Equal: Classifying Cloud Gaming Contexts for Effective User Experience Measurement." arXiv. Read the study
- Vuorre, M., Johannes, N., Magnusson, K., Przybylski, A. K. (2022). "Time spent playing video games is unlikely to impact well-being." Royal Society Open Science, 9(7). Read the study
- Ballou, N., Vuorre, M., Hakman, T., Magnusson, K., Przybylski, A. K. (2025). "Perceived value of video games, but not hours played, predicts mental well-being in casual adult Nintendo players." Royal Society Open Science, 12(3). Read the study
- Strojny, P. et al. (2023). "It's not just about how long you play. Indirect gaming involvement and genre preferences in predicting gaming disorder risk." Frontiers in Psychiatry. Read the study
Kantise is an observation tool, not a medical device. The information in this article is provided for general purposes and does not replace professional medical advice. If your gaming habits worry you, speak to a professional.