Sandbox Onboarding
I recently wrote a blog post on Tutorials in general, as well one chapter of a blog post on specifically the new tutorial area of Stars Reach (an in-development social sandbox MMORPG by Raph Koster). And the feedback I received from both players and even a few developers, made me strongly doubt whether I managed to express myself coherently enough to get my point across.
There was a decent amount of agreement with my post - and while there was very little direct disagreement, there definitely was a quite significant amount of: “I don’t see the issue.” The (seemingly surprised) developer responses can probably be summed up as:
“We are not going to turn this game into a themepark. That on-rails treadmill experience we are introducing in the tutorial, will later be taken away again from the player. What could possibly go wrong?”
Since I’ve already written about tutorials in general, as well as the Stars Reach tutorial specifically - this post is going to be about the onboarding and introduction of players to a social sandbox MMORPG (with most of it likely being applicable to non-MMORPGs as well). That’s a lot more specific than “any tutorial” - but I will keep it general enough, to not become a Stars Reach post.
The Harsh Truth
Let’s get this out of the way first: No game is for everyone. No matter what a game does - there will always be some players who just are not into that.
Any onboarding process will always loose some players and therefore also act as a filter - not just as an introduction. There’s a reason why analysts call this a “funnel”: the stream of users is wider at the top, and gets ever narrower towards the bottom - loosing people at every step along the way. That is both normal, and unavoidable. You can try to reduce the number of losses (most effectively by removing steps and shortening the funnel), but more importantly you can try to capture just the right audience, and limit the losses to mostly those, who would never have stuck around anyway. Trying to chase everyone will always fail, and optimizing for just the right audience is the only sensible thing to do.
That funnel starts whenever the potential new player first hears about the game (sees a trailer, is told by a friend, reads about it on a games website, etc.) - and it ends when that new player commits to the game, mentally switches from “just trying out” to fully adopting it. That’s when the chances of loosing that player drop steeply.
Onboarding ends when the player is sold on the promise of the game.
Since we are unavoidably loosing people at every step along the onboarding funnel, it should be obvious that this promise needs to be presented to the player as early as possible. As I said above - the most effective way of reducing losses, is to decrease the number of steps and shorten the funnel. Here’s the twist though: in the case of a sandbox MMORPG, it’s the promise of freedom that sells the game - but the actual full freedom would completely overwhelm most players this early. It’s only the promise that has to be sold. And that promise has to be sold without the player encountering any hurdles, or feeling lost, or anything else that could make them bounce off the game. Another reason to sell the promise as early as possible. Every extra step will bounce off extra players.
The Challenge Of Sandbox Tutorials
Themeparks have it easy: they are mostly on rails anyway. Drop the player in, have an NPC tell them what to do (first quest), show them a quest arrow to follow, once at the right place, tell them onscreen to “kill 10 rats”, then have the arrow return them to the NPC to turn in the quest and get some rewards. And voila - the player already knows the core gameplay loop that will take them all the way to max level. Everything else, every other system or mechanic, can be introduced along the way in exactly the same fashion. And preferably that is done spread out over time, to allow players to internalize one thing before the next is brought up.
But at the same time, the player is also immediately introduced to the main promise of the genre: that quest loop that not only provides regular dopamine hits, XP, currency and item rewards - but also progresses the character (xp, level ups, power growth) and the story (main story quests, each zone a list of checkboxes to tick off). That’s the promise that sells players on a themepark: the promise of that convenient quest-loop, the ease of just following the arrow, the regular dopamine hits and level up dings, the fun of making numbers go up combined with the power fantasy of getting stronger, and eventually becoming the big hero that saves the world from all evil. There is a big audience for that. It’s also incredibly elegant how that first “kill 10 rats” quests teaches the player everything. Well, everything important - including both the core loop and the promise of the game.
A sandbox game simply cannot do that. It has to sell the promise of freedom - it can’t sell the player on the promise of that themepark dopamine-loop, because that would not only be a lie but also filter for the wrong audience. And once the player is sold on that dopamine-loop, it will not be possible to withdraw it again. At least not without definitely loosing that player and getting a well-earned negative review in return.
Do not try to sell the player on a promise you have no intentions of keeping.
The big challenge of sandbox tutorials is that they have to sell the promise of freedom - while not overwhelming the player with said freedom. And they have to do that, without relying on methods that would inadvertently present a different promise. And that is not an easy thing to do.
The Reward Problem
A themepark tells the player who they are (the chosen one), what to do (kill 10 rats) and why it matters (to save the world from evil) - and then showers them with extrinsic rewards like XP, currency and items - but also story beats. In a themepark, the entire game world and story exist to validate the player and fuel their power fantasy.
In a sandbox though, the player has to decide who they want to be, what they want to do, and why it matters to them - because in this game world, they are not special. This heavily relies on intrinsic motivation - and the validation has to come from other players. That’s why sandboxes are designed around inter-dependency, so that what one player does, actually matters to other players. Players may not be special - but they do matter. And that is because everyone sits somewhere within the supply chain of anyone else. Once this clicks for a player, it’s incredibly powerful at retention. But getting this to work can be hard - especially with new players who don’t yet have the mental models, social ties or confidence.
The problem is, that handing out extrinsic rewards undermines and devalues the intrinsic ones. In other words: the best way to stop players from caring about forming their own goals for intrinsic motivation, is to motivate them with extrinsic rewards for doing what they’re told to do. Once validation is derived from those extrinsic rewards, seeking validation from other players becomes less necessary. Other players become less necessary. The whole foundation of player-supplied motivation that a social sandbox relies upon begins to crumble.
That does not mean, that a sandbox absolutely can not tell players what to do, and then reward them for doing so. But it has to be very careful to do this in a way that does not cheapen intrinsic rewards - and in a way that doesn’t cause the player to take away the wrong learnings. It probably would be best, to not do this at all during the initial learning and teaching phase, when the player is still forming and confirming their expectations on how this game works. After the player is sold on the promise of freedom, and has their social ties and self-motivated future plans - offering such quests and rewards might just be fine. But if the game leans too heavily into this too early, the player is trained to develop a mindset of:
“If the game doesn’t tell me what to do, there is nothing to do.”
That mindset is lethal in a sandbox. Once it develops in a player, that player is basically already lost. In order to retain players, the game has to first teach them to supply their own motivation and define their own goals.
Do not teach the player that the game will tell them what to do.
What A Sandbox Tutorial Should Do
Make Other Players The Tutorial
Promote, support and reward newbie-friendly guilds, mentorship programs, newbie help channels, shared projects (like player cities), player-run academies and similar undertakings. This not only helps forming early social bonds - but it also turns knowledge into a valuable resource and confers social status upon the teacher.
Letting new players see and interact with veterans also helps to give a first view of the possibility space. New players can get glimpses of what others players are able to achieve - see their space ships, their houses, their cities, their farms, their pets… all of which can serve as aspirational goals for the new player.
This absolutely requires that the newbie zone is accessible to veterans - and not some sequestered off island that’s impossible to ever get back to.
In some Asian MMO I once played (long forgot the name), you’d start out mostly grinding some easy creatures that dropped a few trash items which sold for mere coppers. But max-level characters on their huge mounts would appear in the starter zone, and offer newbies 2 gold per stack of a certain trash item. Turns out it could be used for some late game crafting recipe, and stopped dropping once you hit level 5 - so farming it by creating new characters all the time was a bit tedious. Those 2 gold (peanuts for the veteran) were huge, for a newbie like me. And seeing those veterans showing off all their fancy gear was impressive. But the single greatest aspect of all of that, was to feel useful and accepted as a newbie. It’s just a huge morale boost.
Of course, other players should not be the sole and only source of teaching - and it should fully be possible for a shy player to get started without having to talk to anyone. But the option of getting taught by other players should not only be there - it should be at least equally promoted and visible as solo NPC options are.
Teach Verbs, Not Objectives
Show players what they can do, and how systems work - but don’t tell them what they must accomplish. In this context, “verbs” are player actions, the things that players can do: Move, attack, craft, trade, build, harvest, heal, form parties, join guilds, etc. Objectives on the other hand, are specific outcomes the game tells you to achieve, like “kill 10 rats”, “craft a sword”, “deliver this letter”. These guide the player towards pre-defined paths - while sticking to verbs builds player agency, encourages experimentation and prepares players for self-directed goals.
Where objectives teach task completion and train the player to think: “The game will tell me what to do next” - verbs do teach players about the possibility space and train them to ask: “What can I do with the tools I’ve been given?”. In a sandbox MMORPG, the core appeal is freedom. The “game” is not the quest-chain, it’s the simulated ecosystem and all player interactions. Therefore it is very important to reduce the theme-park mindset, and not reinforce it. This is especially important for new players who might not be familiar with sandboxes.
Objective-heavy onboarding creates reliance on extrinsic motivation, dependency on quest markers, and the wrong mindset. Verb-focused onboarding emphasizes player autonomy (which is not only one of three pillars of Self-Determination Theory - but also what sandbox games thrive on). It lets players build competence gradually, by teaching small, atomic actions, demonstrating cause and effect and allowing players to experiment safely. This small scope step-by-step approach helps to not overwhelm players. It doesn’t tell the player to “bring me 10 rat furs” - but offers them to “show you how to hunt and skin creatures”.
You might think this distinction is rather moot if the player then ends up killing and skinning a few rats anyway - but do not underestimate the psychological effect this can have. Especially when it comes to shaping a player’s sense of agency.
If a sandbox tutorial offers too much external structure, players will feel lost and directionless, once that structure falls away. Less structure requires more self-direction, and thus preps the player to choose their own goals. And letting the player choose whatever kind of animal they’d like to try hunting (and how many of them), is a very easy and subtle way to coax the player into making such choices for themselves (and getting used to doing so) without even noticing.
If the tutorial offers too much freedom, it will create anxiety in the player - but if it offers too little freedom it will create dependency (as in “lack of being independent”). Give players too much structure and they’ll come out of the tutorial not even able to think for themselves. Hard-scripted quest-chains are the prime example of “too much structure”.
Give The Player Identity Before Mastery
Let players feel like a meaningful “someone” in the world immediately, even while they are still mechanically weak. Allow them to carry a title early like for example “scout” or “apprentice blacksmith” or “trader”. Let them start with a small plot of land that they can later build on, expand and improve. Give them something to do early on, that matters and contributes to the economy in some small form. Make them part of a faction right away. Let them identify with a role and self-concept, long before they achieve mastery with it.
Avoid making the player feel generic and insignificant until they have hit max level, or achieved mastery. Themeparks do that - technically players are completely insignificant to others, until they reach endgame and join up with a raid group. That then is a position that matters - but for the entire leveling phase before that, they are practically insignificant. Not to the NPCs and the story though - yeah that’s fake, but that doesn’t matter. To the NPCs and the story they are the chosen one, the big hero, the savior of the world. And that’s how themepark can get away with having the player be technically insignificant. As long as they don’t feel insignificant because the power fantasy provides the illusion of being useful, everything is perfectly fine. We players are most willingly suspending disbelief after all.
But in a sandbox where validation always comes from other players, and not from the story and power fantasy - it’s hugely important to get a little bit of that early on, even as a newbie, long before achieving mastery. Mastery is the long journey - and identity provides the motivation to continue on that journey. Instead of first having to grind to max to then become a blacksmith - you can already be a blacksmith, while slowly getting better at it. What helps is that what a newbie does really doesn’t have to much impact or importance. Even if the contribution is absolutely tiny, it still will feel like everything to the newbie. Remember my example further up? To newbie me those 2 gold I got were huge - to the veteran it was insignificant peanuts.
But there has to be something - no matter how small - some tiny thing that a newbie player can provide, that’s still useful, if only a little, even to a maxed-out veteran. And that little something should let the newbie player identify as being a “blacksmith” or “tamer” or “farmer” or “scout” or whatever.
Once the player does identify that way, they will gain motivation from that assumed identity. They will have the aspirational goal to become a master of what they are doing. They will feel some early social recognition as a “future master blacksmith”. Identity is belonging, and belonging keeps people.
If mastery comes first, then players are encouraged to optimize the power curve, to grind, to meet expectations (meta-gaming) and to think in numbers (how much DPS do you deal?) - but if identity comes first, that encourages roleplay, social status, building reputation and that leads to increased diversity of roles. Sandboxes benefit a lot from the latter, therefore they absolutely should nudge players towards building a virtual identity for themselves, already in the tutorial.
And that’s pretty easy to do actually - players will automatically start forming an identity in their head, if you just provide them with some examples of the possibilities that they have. Many people will indeed already start doing that, when watching trailers or reading the list of available professions on website. Providing something as simple as early job-titles already goes a long way in fostering that.
What A Sandbox Tutorial Should Not Do
Teach Systems First
Designers see the complexity of the game and think: “Players will quit unless we explain everything”. As a result they build long tutorials, tooltips everywhere, isolated training areas and step-by-step system walkthroughs.
This typically backfires, because it teaches players to ask “what is the correct way to play?” and “what is the intended use of this tool?”. That’s a themepark mindset and discourages experimentation, emergent play and expressive play.
Use Quest Treadmills To Ease Players In
The idea is to just use these treadmills to help teach mechanics, and then remove them later.
This typically backfires, because players don’t experience the treadmill as “temporary”. They think “this is how the game is”, form the wrong habits, the wrong expectations and the wrong mindset. When the rails then vanish, they feel abandoned, lost or cheated. At the same time, the treadmill might scare away actual sandbox enthusiasts.
Use Fake Constraints
The idea is to use disconnected tutorial islands, invisible walls, “not-allowed-yet” gates and scripted NPC behavior for teaching the player.
This typically backfires because it severely undermines the promise of freedom. Players will conclude that the freedom is fake, pure marketing copy - and that the game doesn’t actually offer anything that’s not staged. When the game later changes, it’s often too late to change people’s opinions on that.
Try To Appeal To Everyone
Because the game is loosing players early, clearer objectives and more rewards get added, and friction is removed.
Any game that tries to do that, ultimately ends up appealing to no one. The changes might cause less early loss of players outside the core audience - but in the long term it causes more losses across the entire audience. A game that clearly knows it is for a specific audience, will also exude stronger appeal to exactly that audience. If you chase two rabbits, you’ll catch neither.
Overprotect New Players
You want to protect new players from everything that could feel bad to them, so you isolate them from veterans, turn off all PvP, turn off world modification, etc.
This typically backfires because it does not remove the risk, only delay it. And when the protections finally are turned off, the player might feel that the game suddenly changed, or that the tutorial lied to them or misrepresented the actual game. The pain is then amplified because the player was taught to feel safe. If the game has rougher sides, it should be honest about that, and use the tutorial to warn and prepare the player, rather than sweep it under the rug.
Give Players Direction Instead Of Clarity
This is when the game tells the player what to do, instead of teaching them how things work.
If the mechanics and systems are made clear to the player, they should immediately have their own ideas on what they want to try and do with that. When clarity turns into direction, then exploration dies, discovery becomes consumption and players stop forming theories. They just do as told. But that just doesn’t work out in a sandbox.
The Litmus Test
Here’s a simple, straight-forward question that you can check any part or feature of any sandbox tutorial against:
If this feature was kept in for the entirety of the game - would the game still be a sandbox?
If the answer is no, then this tutorial is training/appealing to the wrong player. It is targeting the wrong audience. It’s a feature that is not meant for the same audience who the game is meant for.
Conclusion
Teaching a new player to “sandbox” is hard. The best chance we have is to dive right in, be honest, sell the promise of freedom, and make everything in the tutorial support the right mindset of self-directed, self-motivated play. Foster early social interactions, help new players build an identity, and make them feel useful.
If there’s one thing not to do, then that’s to make the tutorial a themepark that teaches some game systems/mechanics through objectives, while at the same time inadvertently also teaching how to “themepark” rather than how to “sandbox”, and leaving the player with the wrong mindset and a promise we don’t intend to keep.