clock menu more-arrow no yes mobile

Filed under:

Length Matters

Why no one knows how long video games are — and why maybe we should

Ringer GIF

A line of text at the top of this page told you that this story would take 10 minutes to read. That’s only an estimate, based on average reading rates; you might go slightly slower or faster, and along the way, you might pause to pretend to be working. Even if you don’t reach the end in exactly 10 minutes, though, the guess is still useful in setting expectations. That’s long enough to learn something, but not so long that you won’t have time to finish (if you feel like finishing) before you head home to your hot date/Sears catalog.

In almost every medium (not to mention Medium), this is how entertainment operates: The consumer opts into an experience after the creator gives guidance on how long the content will last. Movies, videos, TV shows, plays, albums, songs, board games, and even amusement-park rides offer either approximate or precise runtimes. Books provide page counts; e-book readers, like this website, include estimated times to completion. Most sporting events include clocks. The benefits of being upfront are twofold: First, consumers can make more informed decisions about whether the payoff will be worth their investment, and second, they can pursue that payoff with a feel for how much of their lives they’ll need to set aside.

Video games don’t work that way. They’re exceptions to the status quo, stubborn black boxes that leave more to the imagination than most media. You won’t find an estimated playtime on a given game’s packaging or its listing in an online store, even though many games last longer than entire TV seasons, with enormous (and, at the extremes, increasing) variability from one title to the next. That doesn’t keep people from guesstimating, though: “In addition to the forever game that has the capacity to be a lifelong hobby (Overwatch), and the 50-hour game that is the standard, canonical video game consumer product (The Witcher 3), we now have the 5-hour game (Inside), the 3-hour game (The Beginner’s Guide), the 1-hour game (Frog Fractions), the 15-minute game (Dys4ia), the 1-minute game (Passage), [and] the 30-seconds-per-day game (VESPER.5),” says Frank Lantz, a game designer who serves as the director of the NYU Game Center.

According to the Entertainment Software Association, the video game industry’s trade association, the average age of gamers has hovered between 30 and 37 since 2004. Although many younger gamers are glad to stretch their budgets and fill their free time with open-ended experiences, adulthood brings competing commitments. Couple those commitments with the ever-widening range between the longest and shortest games and the leisure-time-management problems posed by “peak TV,” and there’s an increasing allure to content that discloses its length. In a saturated media market, some consumers put a premium on forecasting their fun, scheduling their dopamine delivery in predictable parcels. Maybe it’s time for games to get with the program.

Many games go part of the way by including in-menu progress percentages, but those reports are usually pegged to the rare achievement of crossing off every activity a game has to offer, rather than the more common goal of completing the core story and watching the credits scroll. In the absence of any standardized source for playtimes, gamers have historically relied on anecdotal reports by reviewers, comments by developers, or baselines established by previous games in the same genre. For experienced players, those methods work well enough that most gamers get to the start screen somewhat prepared. As Lantz says, “I don’t think anyone bought The Beginner’s Guide expecting a Witcher 3–style experience.”

Nonetheless, it’s clear that the community craves more detailed data; more than one crowdsourced website has sprung up to satisfy the curiosity that game publishers aren’t quenching. The most popular example is How Long to Beat, the brainchild of Minnesota IT technician Randy Gleason. Gleason started HLTB in 2011, when his game backlog grew long enough that the lack of length information frustrated him. “I’d be going through a game and be like, ‘God, how long is this going to take me?’” Gleason says. “A lot of times it was just looking at my Steam game list and [thinking], I’ve got like three or four hours to burn here — what can I get through? What can I knock off my list?”

When Googling failed to supply enough answers, Gleason started a thread on the popular gaming forum NeoGAF, which led to a shared document that quickly became unwieldy enough for him to start a site instead. An audience was waiting: Gleason says the site welcomed more than 665,000 unique visitors last month, and its database, which extends to the early Atari era, houses 350,000-plus playtimes submitted from more than 100,000 registered accounts. To cope with the problem of differing play styles, Gleason allows users to classify their submissions in one of three ways: “Main Story,” (for those who do only the minimum required to complete the narrative), “Main + Extra,” (for those who prefer a more leisurely playthrough, including some exploration and side quests) and “Completionist” (for those who really wring the game dry). As a rule of thumb, he treats 15 submissions as the threshold at which averages become reliable, although they can take more time to stabilize, since the fastest players are also the first to submit.

HLTB data provided by Gleason allows us to break down the average “time to beat” by genre, after excluding multiplayer-only, MMO, and other “endless” games that technically can’t be completed. Among genres with more than 20 titles in the HLTB database, role-playing games lead with a 23-hour “Main Story” average.

How Long to Beat
How Long to Beat

We can also draw on HLTB’s data to display the average game length by year back to 1980, revealing a clear trend over time. The graph below presents the annual average in unweighted form (which counts each game in Gleason’s records the same), and also in weighted form, which gives greater influence to games with more playtimes submitted to the site (a proxy for popularity).

How Long to Beat
How Long to Beat

Games got bigger (and more motley in length) throughout the 1980s and 1990s, as chip makers buffed computers and consoles and storage space increased. Around the turn of the century, that mission creep plateaued. In recent years, game length seems to have shrunk slightly, although the decline appears more pronounced for the “all platforms” group than the “consoles only” crowd, maybe because the flourishing mobile market has dragged down the average. Regardless of platform preference, gamers who gravitate toward rapid playthroughs have more options available than they did a decade ago.

To shift the burden of playtime collection from a user-submitted model to a publisher-provided one would require a well-choreographed campaign. According to an ESA spokesperson, though, any initiative would have to be based on elective decisions by individual publishers, rather than a top-down, industrywide mandate. And as Gleason notes, times reported by publishers wouldn’t always be trustworthy; given the prevailing (if often erroneous) belief among gamers that bigger is better, sellers have some incentive to stretch the truth, as a few standouts did on the back of the box during earlier eras. “Every time you see a developer say, ‘Oh, our game is 20 hours long,’ I’ll play it, and it’ll be like 8 hours long for me,” Gleason says.

To keep developers honest, then, a watchdog would have to arise. The League for Gamers, an advocacy group formed in 2012 as an intended rival to the ESA, could be one candidate. The group’s president, former World of Warcraft developer and Red 5 Studios CEO Mark Kern, says that stumping for length disclosures isn’t on LFG’s radar right now, although he concedes that “it’s an interesting problem that could use more thought.” Another option is the Entertainment Software Rating Board, the self-regulatory organization established by the ESA in 1994. The ESRB already assesses games’ content, handing out ratings that mirror the MPAA’s PG, PG-13, and R. In theory, the board could add playtimes to its purview. According to an ESRB spokesperson, though, that’s not under consideration: “The length of a game can vary depending on the decisions the player makes and has never been considered as a factor in the rating process.”

The ESRB’s stance makes some sense; pinning down playtimes is complicated. “I think it is simple to say that games ‘should’ do this, but harder to say how that number might be calculated for each game,” says Tracy Fullerton, the director of USC Games. Unlike most movies and music, video games are interactive, and even games with scripted campaigns are subject to the player’s approach. “Part of what it boils down to is delivering on expectations,” says Josh Scherr, a writer for Uncharted and The Last of Us developer Naughty Dog. “Is this game really 20 hours long? Does that include cutscenes? Is that with or without all the optional quests in your open-world game? Is that 20 hours for an average player, or someone who’s good at games? And so on and so on and so on.”

The picture is even more muddied for games that are highly replayable, with different outcomes or unlocks at the end of each playthrough. Through prerelease playtesting, though, developers can account for at least some of the stylistic variation from player to player. “By the time any major title is released, we know from thorough testing how long the average playthrough takes, how long average leveling up takes, and how long it takes different play styles to play through the game,” says game writer Anna Megill. “We know the times for players who skip cinematics vs. players who read every line of conversation, explorers vs. achievers, and so forth. We even have internal speedrun records. A studio could easily estimate an average gameplay time.”

However, Megill believes that just because a studio could estimate average playtime doesn’t mean it should. “There’s danger in viewing games as a time investment rather than an experience,” she says. “Journey took me four hours to play through the first time, but it changed the way I see games forever. How do you quantify that?”

There’s nothing inherently better or more satisfying about longer games, but the perception that they’re superior — based, in part, on better bang for the buck — persists. “There might be something particularly ‘gamic,’ if that’s a word, about the idea of needing to enumerate game length explicitly,” says Ian Bogost, author, game designer, and professor of interactive computing at Georgia Tech. “I’m pretty sure book readers and filmgoers don’t principally construe their entertainment experiences as a page-per-dollar or frame-per-dollar affair. But it’s undeniable that gamers do.”

If we filter the HLTB dataset for games with a user-submitted rating above 85 percent, we find some support for that statement. With the sample restricted, the year-to-year fluctuations are more noticeable, but on the whole, higher-rated games tend to be longer than the overall average.

How Long to Beat
How Long to Beat

It’s possible that length and quality are weakly correlated, but the belief that size matters might simply stem from the fact that games are expensive. The pain of paying $60 for a full-length release is amplified when an early ending forces the player to fork over another $60 sooner than expected, knowing that a longer game would have postponed the next trip to the store. Even the best downloadable indie games, which compress a fully realized experience into a four-hour package, can leave gamers conditioned by bigger-budget releases feeling unfulfilled.

The problem is that posting playtimes reinforces the concept that an identifiable ending is something to strive for, an outdated idea in the age of No Man’s Sky and everlasting e-sports. “Once playtime is foregrounded, it assumes the linear conclusion of narrative media,” Bogost says. “And many games resist that.”

Moreover, a publisher’s signpost could steer some players the wrong way. “As a developer I want players to play the game at whatever speed they want, and not tell them what an ‘appropriate’ amount of time is,” says Richard Rouse III, creative director at Paranoid Productions. “And if a player finishes it faster than a posted time, they’ll complain the box ‘lied to them.’” The latter concern could lead to an unintended consequence: If all games advertised their durations, the pressure to appeal to gamers who divide dollars by hours before buying could cause developers to pad their playtimes at the expense of pacing.

Unless you’re planning to preorder or purchase a game on day one, relying on How Long to Beat; similar, less-trafficked sites such as GameLengths; or an automated stat service such as Steam Spy makes gaming’s cost-benefit calculus a little less of an outlier. And a multitiered system like HLTB’s might also work on the back of a box, with exemptions for games that don’t fit the finite, single-player profile. For TV viewers, a 24-episode season and an eight-episode miniseries might be equally satisfying, but for life-planning purposes, it’s still nice to know which one you’re watching before you binge. Similarly, there’s value for gamers in knowing whether a certain title can be beaten in a single sitting or over a lone, lost weekend, or whether weeks or months might melt away while a player squeezes in sessions around real life.

Still, the developers opposed to published playtimes have a point: There’s something to be said for the thrill of discovery, without a timer ticking away in one’s head. Games empower the player in a way that linear media don’t, and a lack of guidance needn’t always be a bug. Maybe we’d be better off if we reframed games not as obstacles to be beaten, but as experiences to be savored. However long they last.