World Class Villains

One area I’ve been curious on for a while is what proportion of players actually complete the games they play. A lot of attention gets put on games about how their narrative sophistication is evolving and how players love playing for the story they uncover, but all of that is so much puffery if the majority of players don’t actually finish the games they start.

80% of people SHOULD have walked out on this film, and I should have been one of them.

This article prompted me to consider the point of completing games in more detail – it states that only between about 10% – 20% of players actually complete the games they start. If true, that’s appalling. I can’t imagine the movie industry considering itself successful if 80% of its viewing audience walked out before the credits rolled and it really scuttles the “players want story” if the strong majority don’t hang around to the end.

There’s also a side argument from players who feel “ripped off” if they buy a game at full price that takes them less than 8 hours to complete, so I’m interested to know if that is a meaningful argument – after all, shorter games should see higher completion rates.

So, how to find out for myself?

The Methodology

This is pretty simple – look to publicly available information to determine how what proportion of players have unlocked the “complete the main storyline / beat the main game” achievement for various titles. (Some games allow you to keep playing once the main storyline has been completed, or award achievements for completing the game multiple times / on different difficulty levels, but its the end of the story the first time that I’m interested in.) Thanks to Steam Achievements (given that Steam is believed to control up to 70% of the rapidly growing PC game downloads market, it should serve as a pretty reliable guide) and Raptr achievement tracking (a games tracking service covering PC and console titles with over 11m users) this kind of information is out there.

Some of these measures will be inexact, but they are the best free information sources that I'm aware of.

Both Raptr and Steam information will be included for comparative points of reference – they engage different (but possibly overlapping, particularly in the PC sphere) audiences. Steam doesn’t have as many game achievement lists as Raptr though, so some games will only have one completion rate mentioned. Also, I’m adding in figure sourced from BioWare about the completion rate for Mass Effect 2. You could question how representative these sites are, but I’m going to ignore that for now (because this intro is already long enough and I’m just interested, not writing for an academic journal).

In order to measure game length, I’m using the average Main Story (Required) Completion time available on HowLongToBeat.com as a source, rounding times to the nearest half hour for the reason of making it simpler for myself. These times may not be fully representative of an ‘average’ gamer’s behaviour (especially if only 10% – 20% of gamers actually finish their game) and sample sizes are small, but I’m including it as interesting information.

Regarding the games, I’m using popular, award-winning and / or mainstream titles such as those nominated by Gamespot as the best of the year (2009, 2010 and 2011) so that the impact of game quality on completion rates shouldn’t be a big factor. After all, these are the ‘best of’ games – if players aren’t completing them, they aren’t completing the rest either.

In order to find out the “game completion” achievement, I’m using sources like GameFAQs.com and Xbox360Achievements.org. This is actually the hard, time consuming part because actually a lot of devs get cute about naming the “You’ve Completed The Game!” achievement and I have to search for it.

Games I’ve Had To Exclude

The Dalish Elf story had the highest completion rate at 31%. Only 13% played through to get all possible endings.

I’ve had to exclude titles that don’t really have a “game end” achievement (e.g. The Sims 3, Forza Motorsport 3), titles that I can’t figure out which achievement to use (e.g. Dark Souls), games with multiple-yet-equally-valid ending achievements (e.g. Dragon Age: Origins, Bastion) and and any games that aren’t tracked on the sites I’m using or don’t have achievements (e.g. Super Mario Galaxy 2).

And Jamming Everything Into One Chart…

The following chart is very cluttered, but it provides an overview on 25 games – their title, the achievement (let me know if I’ve picked the wrong one) I’m using to determine their completion rates from Raptr / Steam / BioWare, the percentage who completed the game and (in the grey bar the bottom) the average playing time to complete the main story on normal difficulty for the first time.

You'll need to click to enlarge the image to make it legible.

So what do we see?

The average completion rate for these selected titles – supposedly the best, or most recognised in recent memory – is 35%, requiring an average of 11 hours to complete. The most completed game, with 7 out of 10 players who started it finishing it, was Heavy Rain, a very narrative-heavy title.

Well, Mass Effect 2 is a cover shooter of sorts...

Not too far behind Heavy Rain is the cover shooter favourite Gears of War 3, which arguably has less of a narrative drive and more a series of notepad sketches that string sequences of assault rifle fire and grenade throwing together. A higher proportion of players finished Gears of War 3’s main campaign than finished Mass Effect 2 – I’ll leave it to you to make your own judgement on that one.

At the bottom end of completion appears to be Red Dead Redemption – supposedly a narrative driven and popular title, but one that 9 out of 10 players didn’t make all the way through – and Super Meat Boy, a reputedly very hard platformer.

At least one very interesting gap appears in the data – look at LA Noire’s completion rate among Raptr users versus Steam users. Raptr, which includes more console owners in its user base, says that 2 out of 5 players who started LA Noire finished it, but Steam and its PC players indicate that only 1 in 10 made it all the way through. Also, roughly double the number of Raptr users completed Braid (45%) versus Steam users (at 27%).

In contrast, more Steam users played Batman: Arkham City all the way through to the end (32% versus 16% among Raptr users).

So, What Does This Tell Us?

My initial take-out was that game completion rates aren’t necessarily as dire as the original article indicates, but they still aren’t good. It could be considered a sad indictment on gamers that only a third of players who started the short and evocative platformer Limbo actually played it to the end.

... or maybe those Limbo players got up to the spider chase bit and were too creeped out to continue.

There’s also an argument to be made that even in titles seen to be heavily multiplayer oriented, at least 1 in 2 gamers still complete the campaigns at least once to see what goes down. In fact, these titles have a higher ‘main story’ completion rate than games that are meant to be narratively driven!

It’s also apparent that games with reputations for being hard – Super Meat Boy, Bayonetta, Dark Souls – aren’t completed by a significant proportion of players who buy them. It’s evidence that backs up expectations.

I’m open to criticism on the methodology, so let me know if you think I’ve missed something or got something wrong. I’m going to look at this data again and examine time taken to complete versus completion rates, but that will be in a later post.