Honey Rose post-mortem, and the rules of engagement

For the project's third anniversary, I wanted to write a post-mortem about the game as is now tradition, something objective and funny, insighful and pointed, with clear "what went right",
"what went wrong", lessons learned, and hopeful promises for the future.

Instead, I find I lack the words and perspective to properly convey everything about what this experience has been and meant.

In retrospect, most things that could went wrong, and I made a number of errors and decisions that ultimately made the game's experience unengaging for many. This write-up aims to look into my own expectations and motivations for making this game in contrast to what I now understand of some player's.

Where to start? The game being a life-management sim, it seems fitting to attempt to sum it up as a series of stats:

-The game has taken 2 years and 10 months to make.

-It's comprised of 3056 individual frames of hand drawn animation: 2080 for the battle characters, and 976 for the visual novel ones, plus about 250 more for background animation.

-It also uses about 50 backgrounds (some of them with a few variations) and 60 illustrations for specific individual events.

-The game's Construct 2 project file has 121409 events, and the event sheet for the game's script has by itself over 39000 of them.

-I had fully exported the project about 100 times before release, and about 50 more since then for patches and hotfixes.

-A single playthrough takes in average 8 hours.

-It has been licensed on Steam over 13000 times, actually downloaded and played 2045, and bought 141 times (plus 3 on itch.io and 2 via direct download). (For all information regarding the pay-what-you-liked business model and it's design and intentions, you can read here )

-More than half the Steam players have played less than 10 minutes, and the average time played is 1h30, with an average play session of 22mn.

-Big Blue has been defeated by around 6% of the Steam players.

-It stands at a 78% "mostly positive" player score on Steam with 19 reviews, has a 4/10 and 6/10 on Sens Critique, and has been reviewed at 7 on Orange Bison (it has not been reviewed on Metacritic).

-By all metrics, it is a commercial bust despite slightly higher than average player numbers.

But really, the game purposefully eschews numbers in favor of context, and that is what I want to focus on.

I am both elated by the positive response to the game, and deeply saddened by the negative feedback.

The game has received fantastic support in the form of patronage, let's plays, reviews, fan-art and articles. Reviews have been fair, and in pointing the game's shortcomings, I have heard about a number of missteps I hope to learn from. (I won't expand on them here, as they would require a series of articles on their own, and are most likely relative to elements too individual and specific to the game to be of use to many others).

To all of you, I say: I am thankful, grateful and humbled by each and every one of your shows of support and positive thoughts about your experience! I have loved interacting with each of you and discussing your experiences with the game (both positive and negative), and those moments have been my greatest success in making this game at all.

As is natural, it has also been sharply critiqued, both for technical and aesthetical reasons. Since the game is so personal, simply hearing about a player's dissatisfaction is enough to overlook all the positive feedback.
The sense I get from a number of players' reactions is one of disappointment, which I now share, and I feel I have let down too many people who came in looking for something that I either failed to achieve, or wasn't aiming to offer in the first place.

To all of them, I say: I'm sorry. My aim here, is now to address a few of the more common design criticism about the game relative to the notion of player engagement (I've addressed all technical issues I could in the form of patches), in the hopes to shed some light about the thoughts that drove the whole design:

About pacing, and why Little Sun's match happens in March: some have expressed the idea that the first match in the game happens "too late". In truth, matches start to happen very early, in the form of dreams: from a purely mechanical standpoint, you can even "fight the last boss" in the first few days of the game. It underlines that presentation and context matters most: players get lost in their power fantasy, and expect it to play out as it traditionally would. The fact that Little Sun's tournament fight happens later than they expect is understood as a letdown, and not a meaningful event, even though they have the chance to play out a mechanical rendition of her match much earlier. It shows it's not the act of fighting her they're after: it's the opportunity to overcome her in the context of the game's setting, which is both a sign of engagement, and a preconception built
from expectations.

About the calendar: it displays the date of events players have received information about, but not the precise events that are meant to take place.
This is to challenge the player's information retention and their attention to the world's details. A month takes about an hour to play through, and most calendar marks will refer to events that are 15 to 30mn of play away. Players never have to retain information for too long, just enough to risk to forget it, and be in a situation where they have to wonder if they had made plans for a meetup, a match, or a test. The variance of possible events is very low to limit the margin of error, and most failing scenarios are accounted for and offer chances for catching up, but the very idea of playing beyond a perceived failing seems to be a primary factor for disengagement.

About writing: The character's plots, from their casual interactions to their more intricate subplots are an important portion of the game's script. The characters were intentionally designed to have simple (simplistic) outwards facades that the players would have to get past and get to know, as they would actual people. Each character's subplot is also designed to have thematic relevance and offer a basis of discussion over a specific viewpoint. Some players have critiqued the writing in a general sense, even though by their own admission they often didn't engage with the characters and skipped the text. While this is a clear sign of the script's failing as it couldn't engage them, I was hoping for critiques to evoke how and why the object of criticism failed, either through comparison or reference to the objectives it was aiming to achieve, rather than a simple declarative statement of opinion. As it stands, I do not know the issue(s) some had: whether it be pacing, phrasing, heavy-handedness, characterization, or the game's structure and lack of twists - and what to address first.
Overall, I was hoping more players would engage and discuss the game's different perspectives, to critique and challenge them with their own. That it failed to generate this discussion is to me the biggest misfire of the game.

About repetitiveness: the game's core genre, the life-management sim, is traditionally tied to a sense of statistical progress following player action. Games of this genre often use a simple presentation and favor numbers over text as the game's flow will inherently be a succession of repetitive feedback loops: make a decision, see a mathematical consequence. Since the game here is about uncertainty in the face of the unknown and the value of perseverance, I chose to mask the numbers and replace them by context. In addition, I made the events the player encounter non-deterministic to challenge the sense of artificiality that comes from having a preset linear course to a story.
This unfortunately results in having the repetitive nature of the mechanics be more apparent than in most games (which are all repetitive by nature, but often rely on much more simple and short loops as a distraction and a source of iteration). The challenge of having reactive events presented in a non-linear fashion was most likely too great to tackle as a first narrative-driven project of this scope. After a point, I couldn't handle the increasing size and scope of the script and had to pare down on the number and variance of events. This leads to all experiences with the game not being equally eventful or diverse, both by nature of the randomness, and because many events are tied to the player's choices and situation. Indulging in routine will unfortunately lead only to routine, some of the alternatives being directly tied to the consequences of failure.

About reactivity to player choice, consequences, and engagement: the game's entire structure is built around the player's failures and the possibility to learn and come back from them. However, most player reactions show they have no desire to engage with the consequences of their actions if they don't lead to success, and choose instead to reload/try again if the story doesn't proceed as they had planned. This is a direct consequence of our culture of glorification of individual success and instant gratification: we tolerate no room for error, and any deviation from a plan is considered a shortcoming. This also ties in to the lack of perceived immediate consequence: some have complained they didn't have a sense of progress in February, meaning the game didn't properly track their choices in their eyes, when in fact the consequences to most player's actions are simply longer-reaching to reflect on the actual real sense of unknown when making day-to-day decisions. The slower pace is meant to challenge the notion of instant gratification, and offer consequences without tying them to rewards. There are short term reactions to indicate the game is indeed tracking the player's actions, but most players simply dismissed them as inconsequential as they didn't translate to statistical rewards. This resulted in player frustration: after the initial rush of the opening, they eventually abandon the experience out of a sense of short-term unfulfillment.

The game also makes an enormous amount of errors. Among the more explicit missteps: there aren't enough events, and those that are there aren't diverse enough. The game being reliant and randomness for its narration, this directly heightens the feeling of repetitiveness if the player encounters the same events too many times in a row. There are also a number of technical issues and presentation oversights, most of which due to my inexperience and the origins of the project as a learning platform: the menus animate for too long, the intro sequence to Honey's Room cannot be skipped, some of the information conveyed to the player can be too easily overlooked or outright missed. Many of those, I'd now do differently, as they are relics of my learning experience, and negatively affect the feeling of a polished, finished release. The game could and should have used more time for polish and for playtesting among a bigger tester base, with more varied backgrounds and player experiences.

Over the course of development, I've attempted several times to make the game more accessible (through aesthetics, inspirations and references, presentation, and even sometimes mechanics). Looking back, that may have been a source to my trouble, as the game can now be mistakenly seen as something it's not: lighthearted and friendly, casual and entertaining. While the outlook of the game attempts to be positive, my main aim wasn't to entertain the audience in a traditional sense: it was to challenge them, offer them a certain perspective, and parallel my own experiences and doubts. By trying to appeal to more players, I may have lost the game's values and core themes.

 
Parts of the original design and presentation still remains in the game, notably through the "diegetic" mode.

One of the game's biggest risks was to attempt to subvert traditional "gaming" conventions and design processes to add meaning. However, the act of subvertion demands a mastery of the conventions themselves, and I now see how the design wasn't solid enough overall to support the meaning I was attempting to convey.

The biggest divergence with the original pitch was the decision to make all player choices happen during the day, rather than use separate day and night halves.
Having two separate choice instances made it possible to work on both aspects of the game at once, when the whole point was to prioritize one over the other.

I had also considered shortening the calendar, and rushing the first match to satisfy the player's fantasy, but I felt this would betray the game's themes, and lessen the eventual impact and emotional value of perservering and overcoming the game's challenges.

In the end, I may have attempted to finish it too soon: the game should have held more, and what it does should have been done better. One of my (many) aims with this project, as a personal training work, was to complete and release it before a set deadline (my 30th birthday). While I did accomplish this goal to my initial satisfaction, this probably hurt the game in the end and I should likely have spent more time polishing and reviewing its content.

In some conversations, the game has been likened to an "art project". This is to me a pleonasm: games *are* art, in that they convey meaning and a view of the world individual to each author, and challenge their audience to share or reflect on their own perspectives.

In our feverish haste to consume games as disposable entertainment products, we often forget to seek out their meaning. That Honey can't properly convey hers is my biggest failing, and the main "thing that went wrong".

As to what went right?

That the game, despite its many faults, exists at all!

Happy birthday, Honey!

EDIT: I've rephrased some of this post to better reflect its aim, and to include some of the discussion that came from the comments below!