On Zelda and Tutorials

by Carl

Recently, there was a reference on Kotaku to an earlier article on Gamasutra, which claimed that the problem with current Zelda titles is that it takes to long to get past the initial tutorials. Gamasutra had this to say:

The Zelda games have – at least since Link to The Past – followed the same structure of introducing the player as an average boy of little importance who is quickly called to action to stop an impending evil. Over the years however, greater emphasis has been placed on the games’ story, and consequently the point at which the player answers the call has been pushed back in each game to expound on the plot more. While the beginning portions are used to deliver the story, this time is also used as a tutorial for the player. After they have learned a bit through the tutorial they eventually come across a “tutorial dungeon”, which is an area that gives the player the chance to try out the puzzle-solving, combat, and navigational skills that are required in later dungeons, but without any real difficulty.

It’s at this point when they arrive at the “tutorial dungeon” that the player finally gets to try their hand at the core Zelda gameplay experience. Usually after a tutorial dungeon is completed, the player feels like they have finished the tutorial part of the game and finally gets to start playing the “real” game, so the time to complete a tutorial dungeon represents the time until a player truly begins playing the “real” game. For this reason, it’s important to note how long it takes for a player to experience this.

The author finds that not only is the tutorial length increasing, it is increasing at a dangerous rate!

Putting these numbers together, I found a trend that seemed relatively disturbing. The blue bar for each of the games is either equal to or greater than the green bar for the previous game; this means that it takes the same or more time for a player to arrive at a dungeon in one game, than it did for the player to complete a dungeon in the previous game. If this trend holds true–and really it has to end at some point–then the next Zelda game could very well require 100 minutes of playing before the player actually arrives at a dungeon!

Yes, and very soon after that, Zelda games will consist of 50 hours of game play before the first temple and then the game will be over, oh no!

In my day job, I sometimes teach logic and critical thinking, and the more I think about it, the more this feels to me like an example of the post-hoc fallacy:

  1. Early Zelda games were good.
  2. Early Zelda games had short tutorials.
  3. Therefore, short tutorials make Zelda games good!

The problem with the post-hoc fallacy is that it assumes that because two things were correlated in the past, the one thing must cause the other. It’s bad logic, which can be seen a couple of ways. For example, we could use the same reasoning pattern to make this obviously absurd argument:

  1. A Link to the Past was a great game.
  2. A Link to the Past begins on a rainy night.
  3. Therefore, beginning on a rainy night makes Zelda games great.

This reasoning pattern does not work, but that doesn’t prove that the Gamasutra author isn’t on to something. Perhaps there is something special about short tutorials that really does make Zelda games better. Let’s look at the evidence for that claim.

First of all, did all early Zelda games have short tutorials? How long was the tutorial in the original Legend of Zelda? Oh yeah, right, there’s no tutorial of any kind. Unless you get a manual with a map, you’re going to be wandering around forever in the first Zelda trying to figure out that there are such things as dungeons, let alone that you need to go through the dungeons in a certain order. That’s the reason why as great as the original Zelda was at the time, people who missed it then find it almost impossible to get into now.

This means that right off the bat we have an example of an early Zelda game that didn’t have a short tutorial. The original Zelda had no tutorial, and it was the worse for it. This means that the tutorial in A Link to the Past was infinity times longer than in the original Zelda! (OK, not really.) More seriously, what we learn from this is that there comes a point at which a tutorial can be too short. I feel that in retrospect, the original Zelda falls into that camp with its lack of a tutorial.

Next, let’s talk about A Link to the Past. I don’t think the initial castle really counts as a dungeon. Yes, there is a map, and you get the boomerang before you fight the guard in front of Zelda’s cell (the map – weapon – boss sequence basic to Zelda games), but there’s no compass, and the whole castle feels more like a special sequence than an ordinary dungeon. It seems to me that if you’re going to complain about the lengthy introductions to games, you should consider the whole castle sequence an introduction and time how long it takes to get the first of the 3 pendants the old man tells you about. LttP was the first Zelda game I ever played, and I recall being unfamiliar with the rhythm of map – weapon – boss until well into the game. When you complete the castle portion, you aren’t yet aware of how central that pattern is to the game as a whole.

That was me quibbling with the premises of the argument, but if we did accept the premises would we still see evidence that a shorter introductory is generally better than a longer one?

Let’s look at Majora’s Mask. I’ll admit that the beginning of Majora’s Mask is stressful. You’re a Deku Scrub, and you keep wondering how to change back into a human being, and there are no save points. But that’s what make it such a great beginning to the game: you’re faced with this impossible situation with your back up against the wall and the pressure mounting and then suddenly, you’re given the ocarina and you have a chance to go back and try it all again from the start. Time traveling back to the first day wouldn’t have been the narrative release it was without the long build up before it. As a result, I would argue that while the initial Clocktown sequence might be trimmed slightly, it’s important for it to be reasonably long and stressful for the game to have impact. The first real dungeon in Majora doesn’t come until you figure out how to get into the swamp, which will probably take at least one three day cycle if not more. The time to dungeon in Majora is very long, but the length of it is what gives Majora its wonderful character. Because you have to figure out what’s going on in town first, you get to know the town’s inhabitants, and that sets in motion the player’s drive to resolve the mysteries of the town residents and complete the game.

On the other hand, Phantom Hourglass is a divisive game (I love it), but you can’t say it was bad because of a lengthy tutorial. You wash up on shore, talk to a fairy, get a sword, and go to the Temple of the Ocean King all in pretty short order. If tutorial length was a sign of a good game, this should be unanimously loved game, not one subjected to a mixed critical reaction.

I suspect that the Gamasutra article was inspired by an earlier article by Tim Rogers that complained about Skyward Sword:

Around the time where the game sends me on the third fetch quest revolving around teaching the player how to use the “Dowsing” ability to search for some laundry list of objects, I told a friend I was bored.

“How long have you been playing?” he asked. I told him I’d been playing for three hours, at which point he said “It gets really good about six hours in.”

You know what else you can do in six hours? You can watch There Will Be Blood twice, and then sit and think about it in the dark and silence for 44 minutes.

I agree with Tim that six hours is too long to wait for a game to get fun. I disagree that this has anything to do with “tutorials” or “time to first dungeon” per se. I blogged about it at the time:

For me, the initial part of Skyward was great. I loved the interaction between Link and the other school kids and especially Zelda. Then getting to the first temple kind of dragged and once you get inside you keep having to go through the same four rooms over and over, which is boring. But then you get the dungeon item, and the pace picks up again. After that I got stuck in a couple of places, but for the most part once you make it past the first three temples the game is consistently good. So, I would say it does get better at six hours in, but there’s still plenty of good stuff in hour one as well and some decent stuff after that.

Changing gears for a second, I think we all can agree that the introduction to Twilight Princess wasn’t very good. However, the problem wasn’t that it took a long time to get to the first temple. It’s that trying to get a cat to take your fish or whatever was counterintuitive and not very fun. Similarly, whether or not you like the initial part of Skyward will vary greatly depending on what you think of the relationship between Link and Zelda, the bird flying competition, the hunt for Fi, and the sequence at the Sealed Grounds. I personally got stuck just outside of the first temple. That wasn’t fun! But the not-fun part of it was separate from how long it had taken me to get to that temple.

I see an interesting contradiction in the way some of these articles are written. Look at this part of the Kotaku article:

I asked Miyamoto if he noticed that Nintendo’s franchise games feel formulaic and how his teams are dealing with the problem of being too formulaic with their longest-running series.

What’s the contradiction? The beginning of a Zelda game is its least formulaic part. Each of the games begins with similar themes (a boy chosen by destiny to blah, blah, blah), but very different specifics (washes up on an island, gets lost in the forest, goes to see his grandmother…). It’s only after you beat the first real temple that things fall into that familiar map – weapon – boss rhythm. The rhythm is important because it makes the game manageable mentally and in terms of game design, but it also means the end of the novelty and unpredictability of the pre-dungeon part. As veteran players, this initial portion makes us anxious (“When am I going to stop doing these weird things and get to the real game?”), but it’s also the only time when we’re back with the novices and faced with the unknown.

Kotaku also writes,

Basically, I asked him [Miyamoto]… well I’ll just quote my own question: “The theory that began to be developed in this particular article was that these games can become a lot of fun—I finished Skyward Sword and thought it was fantastic—but there might be an issue there that a lot of video games from Nintendo and others take much longer to be fun and to get into than a lot of others. And then I think about Angry Birds. That’s fun within seconds. Have you noticed that is at all an issue? Is that something you’ve noticed that some games take a long time to become fun? Is that a problem with modern games?”

There are two sides to this. First, obviously a Zelda game is much a bigger time commitment than Angry Birds. You have to set aside an hour or so if you want to make progress in the game. But if the beginning of a Zelda game is good then it’s also good “in seconds” because watching the initial cut scene is fun, going through the pre-dungeon interactions is fun, and entering the tutorial dungeon is fun.

To conclude, what I’m saying is that overall there’s no real correlation between time to first temple and Zelda game quality. The parts before your first temple are actually the least formulaic part of the whole series, so it’s weird to complain about it unless there’s something specifically wrong with it (as there was in Twilight Princess). A good tutorial should be fun in its own right, and the beginning of a Zelda game is often its least formulaic aspect.