v5.0 Demo Revamp Post-Mortem

I recently released a series of very large updates for the Isles of Sea and Sky demo. I wanted to share a post-mortem to talk about what went into the updates, the feedback and revision process, and the final state of the current demo, ending with opinions from our community about the end result.

I should start by mentioning that I have a very interesting and probably unusual situation with my demo on Steam:

1. It is very large. It usually takes most people about 2 to 4 hours to complete from what my team and I have observed.

2. It has a very high average play time. The average on Steam is about 1 hour and 30 minutes with around 30k downloads.

3. We have an extremely passionate community surrounding the demo (and the future full game of course). So much so that we even have speedrunners who have continuously competed on demo completion runs in different categories like any% and 100%, improving their times and techniques throughout the history of demo updates.

That being said, each update to the demo has been a big deal for our community, and I am always so excited to see what they have to say about each one.

Intro

What is Isles of Sea and Sky?

A fantastical, oceanic, and open world puzzle adventure game. Solve novel block-pushing puzzles while unearthing a mystifying narrative, awakening delightful new friends, gathering collectables, and unleashing extraordinary powers that will blow your mind!

Who Are We?

Cicada Games is my small studio of 3 people. This is our first major title. Our team consists of myself developing the game (art, coding, design, etc.), Dan my producer who carries the brunt of so much non-game dev work so I can focus more on game dev (marketing, social media, feedback, QA, reminding my dumb ass that we have meetings -_-, etc.), and Craig our extremely talented composer who also helps greatly with feedback and QA. Dan, Craig, and myself all grew up on the same street and were childhood best friends if you can believe it. We have a long history of playing and making games together starting with adventures I used to create for my buddies with sidewalk chalk.

v5.0 Series of Demo Updates - Post Mortem

Why a New Update?

Last month we took part in the Cerebral Puzzle Showcase. This event has been a BIG DEAL for us for the past 2 years since its inception. The reason being that puzzle games have it very rough when it comes to marketing (plenty of bad puzzle games out there), so a curated event like this results in great visibility and wishlists. The first event was so successful that Valve basically copied it immediately after with their own “Steam Puzzle Fest.” However, that event was completely uncurated, featuring hundreds (thousands?) of games, many of which were not puzzle games at all. The result was a “rich get richer” event in which games that were already wildly popular and favored by Steam algorithms received a disproportionate amount of traffic, but this is not what I’m here to talk about and I could go on about this at length, so let me get back on track.

I wanted to create a big update for the demo for the Cerebral Puzzle Showcase in order to boost excitement and visibility during the event. Also, conventional wisdom says that having big updates for your demo is a good way to foster excitement for your game, and we have experienced as much with each update so far. Another piece of conventional wisdom is that “signal boosting” (timing multiple events, news, updates, etc. all at once) works well for marketing purposes, so what better time to release a big new update than during an event that has been historically very successful for us? The new update would include some new content from the full game, as well as many new features and QoL improvements which I will dive into before talking about the feedback, revisions, and results.

I’ve chosen a couple important pieces from each update to talk about:

v5.0

Sokoblock “Meta Puzzle”

A big new, mysterious puzzle that takes place across multiple rooms. The nature of this puzzle is hard to describe, but basically it entails 4 blocks which must be moved onto corresponding symbols on the floor (Fig.1a). However, the puzzle takes place across 4 different rooms (Fig. 1b) where only one of the blocks can be moved in each room. So the player has to go to each room and move the block there into the correct position (Fig. 1c), which changes their positions in the other rooms, until all 4 blocks are in the correct positions. This opens up the 5th “goal room” which leads to the reward for solving the puzzle.

After getting it into the hands of the players, some pain points became obvious with the design. Namely the fact that a player that has no idea what they are doing with the puzzle yet who is simply experimenting, can create painful situations in which they have blocked their moves in the other rooms, requiring tedious backtracking to the first room where the mistake was made in order to correct it. One of my main philosophies is that players should be able to experiment safely, and this did not match that philosophy. The layout of the puzzle also sometimes required the use of a holding spot (Fig. 1d) to keep a block in place out of the way and return later to move it into the correct position. It also facilitated the need for a special reset button. This layout could cause tedious backtracking even if the player knew what they were doing, but also two new arbitrary mechanics needed to be learned for this puzzle and this puzzle alone. You will see below in the notes on v5.2 how I addressed these issues.

Finally, the reward for solving this meta-puzzle was a big one:

Gopher Gloves

A new concept not available in previous demo versions: A permanent, Metroidvania style item (Fig. 2). You may be aware of the Mole Mitts from the Zelda Minish Cap game, but like the other power ups in this game, there is always some twist which expands upon the puzzle aspect of it. Rather than simply provide an automatic way to bypass “gates” (in this case piles of dirt), they open up new puzzles and new possible solutions. So while these gloves allow one to dig through these piles of dirt like in Minish Cap, there are a couple of added layers that create puzzle mechanics:

1. The player can not stop once they begin digging in a direction until there is no dirt left to dig (Fig. 3).

2. Dirt piles are not annihilated, but instead shift behind the player.

3. Dirt piles are solid (blocks can not be pushed through them, and they act as weights on pressure plate switches).

The end result is somewhat like an amalgamation of sliding puzzles (where one must shift pieces around to create a desired layout) and ice sliding puzzles (where the player must move fully in one direction until the end of a path). On paper, this sounds incredibly boring, as it combines two types of puzzles most people dislike, but the result is actually something that is much more interesting and fun than the sum of its parts.

With the addition of this item to the demo, several new puzzles and pathways were added which leverage or require the item.

v5.1

Not much to this patch beyond bug fixes. There are always tons of bugs that players find once you get a build in their hands!

V5.2

Big Keys -> Rune Stones

A gatekeeping mechanic in the initial part of each main island is a big key which is obtained by visiting the main room where the god of the island resides. After multiple years of watching people play the game, it had become apparent that the classic Zelda (A Link to the Past) trope of small disposable keys vs. a large permanent key was not universally known. A great deal of players would acquire the big key and think that they could only use it once, which caused a lot of decision making problems and confusion. There were two issues on my end at play here:

1. I establish immediately and reinforced often that keys are expendable items which break after use before introducing the big key.

2. I was not even sticking to the Zelda big key trope faithfully, since in A Link to the Past, the big key in a dungeon is only used to open the boss door and the big treasure chest. In my game, there are several locks scattered about the island that the key is used for, which I think removes some of the understanding of how the key works if someone is aware of the Zelda trope.

I had received feedback from players about this confusion, as well as suggestions from fellow game designers to try changing up the metaphor of the big key and lock. There were only two arguments against the change: Big keys are cool and our team and players were used to the current design. Both of these are really non-issues when compared to the confusion that some new players faced if they weren’t aware of the trope.

So, I changed Big Keys to Rune Stones (Fig. 4), which meant:

• The artwork of both key and lock changed.

• The “Rune Locks” appear more like a somewhat grayed out stone before the Rune Stone is obtained.

• When the Rune Stone is obtained, an effect is created on the Rune Lock in the room which draws the player’s attention to it. All of the corresponding Rune Locks are then brightly colored with a glowing emblem to drive home that they are now unlocked (Fig. 5).

Sokoblock Puzzle Revamp

After a lot of thinking on this one, I was able to come up with a few changes that not only help to avoid tedium, excessive confusion, and frustration, but also serve to make the puzzle more interesting and fun!

What was changed:

• The layout of blocks was changed (Fig. 6a) which killed four birds with one stone!

  • The player could no longer block themselves in other rooms, and could adjust the position of a block in any room regardless of the positions of the other blocks. This means experimentation is not punished.

  • It removed the need for back tracking if the player understands the puzzle.

  • It removed the need for the temporary holding spots.

  • It removed the need for the special reset room button.

• Symbols on blocks removed (Fig. 6a), but visible in the goal room (Fig. 6b). Now a player can view the layout of blocks and where they need to go in the goal room for this puzzle, and then head out to each room to arrange the blocks. This creates more mystery in the rooms themselves, while consolidating a lot of information into one room. This room also shows a green glow on blocks in the correct position (Fig. 6b), and a red glow on blocks in the incorrect positions. A confused player can always return to this room for more clues as they experiment.

• As the player moves blocks into the correct positions, their symbols are revealed on the other screens (Fig. 6a, bottom left block), further reinforcing the player’s correct moves. E.g. “I think this is the right thing to do” *upon entering the next Sokoblock room:* “Ok, I definitely see now that my instinct was correct.”

Pattern Puzzle Revamped

Another major source of confusion throughout the years of feedback on this demo has been this puzzle in which a clue for solving a puzzle on a small 1 room island is hidden in plain sight on another island. This has been the most consistent badly stuck point for players, i.e. players often become so stuck that they reach out through social media or to our community asking for hints. I have to guess that for every player that needs to reach out for hints at a point like this, there must be at least dozens of other players who give up and say nothing. I find these rounds of iterative design serve to enhance the player experience, and at worst, do not affect the spirit, interest, or fun of the puzzle if done correctly.

So how was this issue addressed?

1. The small island with the puzzle is now blocked off with a lock / stone bearing a unique symbol (Fig. 7).

2. The clue for the puzzle now appears, but fades to semi-transparent when the player enters the room. Drawing attention but also indicating that it is not “activated” (Fig. 8a-b, before and after).

3. The clue now has a special symbol on the floor beneath it which matches the lock / stone mentioned above (Fig. 8b arrow). When the player steps on this symbol, the clue becomes fully opaque, and a very prominent effect and sound are triggered on the clue to ensure the player notes that something important has been triggered / activated.

4. When the player arrives at the island with the puzzle, the special lock is removed in a flourish. In addition, the same prominent special effect and sound played when activating the clue, now triggers on the puzzle in this room, hopefully linking the two events in the player’s mind so that they can connect the dots and attempt to solve it.

5. The puzzle is set up so that speedrunners do not actually need to activate the clue, which would otherwise be more time consuming. Instead, they can simply see the clue which is always visible when they pass through that room, and use a piece of speedrunning tech to skip past the lock. When designing this game, the best player experience for as many people as possible is my goal, but speedrunners are definitely not an afterthought. I feel it is possible to design for a general audience and speedrunners simultaneously, but the general audience does take priority if there is ever a conflict (So far conflicts here have been easily avoidable).

End Results

Wishlists / Cerebral Puzzle Showcase

The Cerebral Puzzle Showcase was a pretty big deal for us, garnering about 1,400 wishlists! After the event until the 30th of the month we gained about 300 more wishlists (Fig. 9).

So was all the work on a big new demo update worth it? I believe so for 3 main reasons:

1. It likely acted as a good signal boost during the event.

2. It included a lot of content which I had already created / had been working on for the full game, so it wasn’t like totally derailing development.

3. Last but certainly not least, it stoked a lot of excitement, discussion, feedback, and new speedrunning developments in our community!

Design Changes

We’ve received a lot of positive feedback about these changes and the clarity they bring directly, and so far we haven’t seen any playthroughs where someone was confused or unreasonably frustrated by the Sokoblock puzzle, the pattern puzzle, or the small key vs. big key metaphor. We HAVE received only a couple of comments about missing the old big key designs / preferring that metaphor. However, I feel this will be helped as those existing players get used to the new design over time. It’s also simply a matter of preference for something familiar, which is ultimately outweighed by the need to address issues caused by the old design. There may be room to polish the rune stone design over time as well.

Player Feedback

The feedback has been overwhelmingly positive. Here are some quotes we got from our community about the final v5.2 version of the demo:

“The previous demo made me excited for the full game. This new demo increased that excitement tenfold!” -Tonite

“Isles of Sea and Sky is a game I cannot sing enough praises about. It is an absolute masterclass of design on every level that I highly recommend to anyone.” -SapphireDrew

"After speedrunning the game for so long, Isles of Sea and Sky's latest demo reminded me of why I fell in love with this game to begin with. It truly refills the sense of mystery and wonder that keeps you exploring to learn more about the world." -Adventium_

"I love when a game surprises me. There's so many mysteries and secrets in the game that a surprise awaits you every turn. The demo is a good reminder that this game is very special and I absolutely cannot wait for the release! " -dackid

"When I first played..., I was reminded of my childhood. Playing this new Demo, the same feeling remained if it didn't get stronger… [it] added a new layer of mystery... that has me really excited for the full game!" -Pendragon

"Every demo update... continues to iterate, evolve, and improve the game, bringing me a sense of joy and wonder each time. The team's passion and dedication has been evident to me from the first moment I first washed up on the beach in-game. I can't wait to experience the masterpiece that will be the full game!" -CyberSe7en

“[Isles of Sea and Sky] manages to take everything I love about gaming and condenses it into one awesome package, it's not even out yet but I feel like I've gotten more from it than even the highest budget titles.” -MartianBlobfish

“I really liked how there were no instructions so that you learned by just butting your head against things. I didn't need any outside help but it still felt challenging. I think the cleverest the game made me feel was figuring out when things were connected that weren't on the same screen.” -archcorenth

Thank you so much for reading!

Welcome to The Player Experience.

One of my favorite things about game design is the player experience. How does the player interface with the game through controls, art, UI, music, etc? How does that interface “feel?” How best to make that interface smooth and seamless so that the player can become fully emersed in a game? I’ve always been interested in languages, and in many ways, games are a universal language now. Not only does great UX design allow us to create experiences and stories in a visual language that everyone can understand and enjoy, but it is also the secret sauce of game feel and quality. It is one of the most important aspects of game polish in my mind.

So I’ve started this blog, “The Player Experience,” in order to put down my thoughts on the subject. As the title implies, my posts will focus mainly on creating and refining great player experiences in games through UI, UX, input feedback, and general game feel design philosophy, as well as any other relevant topics. I hope it will become a worthwhile endeavor.