It’s postmortem time!
If you haven’t yet played Grow Your Love, you can do so here:
You can also download the official soundtrack here:
This was my first time doing a jam entry with my trusty partner-in-crime Kat Jia since waayyy back in Ludum Dare 28 when we made Match Girl (and took 2nd place!). That was two years ago, and since then, the number of jam entries has more than doubled, going from 780 to 1,638. Wow!
This time around, we managed to take 3rd place in Audio, 13th place in Fun, and 13th place Overall. Not bad at all! 😀
My personal goals going into the project were:
- Don’t make another rhythm game
- Work on pixel art together
- Have fun!
- (Maybe) treat iOS and Android porting as first-class
Humorously enough, I actually did very poorly on most of these. xD While I didn’t REALLY make another rhythm game, one of our seven minigames was a simple dance game (haha). Kat ended up doing 90% of the art as I had no time at all to work on any of the graphics…I drew the shooting stars and some of the UI elements like the arrows in the kniting minigame, but that was it =X. And iOS and Android porting, pfffttt, come on, we both saw it coming from a mile away that that would get dropped.
Fortunately, the most important goal of having fun was a huge success; this game was definitely very enjoyable to make!
As always, let’s see what went well and what didn’t go so well.
What went well
Concept, mood, feel, and overall vision
Probably the most common comment we got about our game was that it was really cute and adorable, which was great! When we set out to decide what kind of game we were going to make, we weren’t trying to make something that was super innovative or challenging or anything — we really just wanted to make something fun and cute; something that would fit in with our Cocoa Moss label. We had a really good and light-hearted energy bouncing ideas back and forth for the different parts of the game, and I think it reflects in the overall end product. Kat did an especially great job with the dancing animations, and I tried to write out the Letter and Texting dialogue in a way to match the same style as well. It all ended up coming together very well, and I think the game is much more about enjoying a cute story than about the actual scoring of the minigames itself, which is perfect!
Making smart code architecture decisions
Having SEVEN different minigames to make meant that code reuse was at a premium, and I found myself super-grateful to myself that I had enough foresight to plan well for it. There was no way I would have finished all seven games in time if I had to create entirely separate logic for each of them. Sharing most of the common logic for the tutorial demos, starting and ending the game, etc. saved me heaps of time and even though there were lots of hacks and messiness throughout the codebase (as always), it ended up working out really well.
Slamming out music faster than ever before
Holy crap. I know I’m known in some circles for my speed-composing abilities, but I even outdid myself this time by writing the entire 10-song soundtrack in less than 5 hours total. O_O FL Studio saves your project working times automatically; here was the breakdown of how long I spent on each track:
Lovers’ Rave: 32 minutes
Watch Carefully: 15 minutes
Hope You’re Not Asleep: 34 minutes
Stargazing: 20 minutes
Love Sprout: 3 minutes
Grow Your Love: 1 hour 6 minutes
Wiggly: 31 minutes
Our Love Has Grown: 46 minutes
From Me to You: 17 minutes
Rainbows of Yarn: 20 minutes
Total: 4 hours 44 minutes
Aside from the title theme I basically wrote each song in 20-30 minutes, which was kind of insane. Of course, these were shorter songs, so it makes sense that I was able to churn them out at a faster pace, but I was surprised at how well I was able to work under pressure here while still remaining creative, especially given that some of the songs are not of my usual style. I was under MASSIVE time pressure as I started doing the bulk of the soundtrack work, so I really had no choice but to do it quickly. Ironically, this might have =helped= my creativity by giving me no choice but to go with my first gut instinct. I’m especially happy with “Wiggly”, which was ridiculously fun to write.
You can read more of my comments on the individual songs in the in-game jukebox if you’d like — there are some interesting notes in there since a lot of the songs were repurposed from what they were originally intended for. To be honest, neither me nor Kat were planning for the mood of the game to play out exactly how it did; it sort of evolved as things went. I had to rethink how I was doing the music to reflect the shift in mood that ended up happening, but luckily I didn’t have any wasted work as a result. I do have to say that the full hour I spent on the main Grow Your Love theme was well worth it; it’s now my favorite track of the entire soundtrack and I think it really gives a lovely first impression. You’ll notice that I centered all of the songs around a shared motif — it’s a great technique I’ve used over and over again that really brings cohesion to a soundtrack.
What didn’t go so well
People not getting the menu controls
Sadness. Perhaps I was going too deep with the “two button controls” theme, but I thought that in order to really be complete the menu of the game should use two button controls as well. Of course, it’s hard to make menu controls using only two buttons, so I looked to DiveKick for how they did it and figured that I would do the same thing. Apparently it wasn’t obvious to most people as they just gave the screen a blank stare and wondered why pressing left or right separately didn’t do anything, so in the post-compo version I made it painfully obvious by putting big flashing indicators on the screen and highlighting the text, press LEFT + RIGHT TOGETHER. Sigh. It’s always the little things that you assume that bite you in the foot later on…
Not testing on the release platform until later on
This one was a smaller point, but for most of development I was testing on Haxe’s Neko VM instead of actually testing the Flash compile. This was fine, except halfway through development when I tried the flash build and ran into a crisis. You see, most of the graphics in the game are upscaled by 4x, so I had set the camera zoom at 4. But the text was way too big at 4x, so I had the text sized 50% smaller, so overall the text would be upscaled by 2x. Setting the text scales to 50% worked fine in neko and everything was perfectly happy. Well, as it turns out, in Flash with the buffer rendering flow, you actually can’t draw pixels with a size less than the current camera zoom, so instead of the text being nice and crispy at a 50% scale compared to everything else it was actually just unreadable. Oops. Luckily I was able to jury-rig a fix by hacking at the haxepunk code, so that was a major crisis averted. The hack was pretty ugly — I set the camera zoom to 2 instead of 4 and then edited the image class so that every image in the game would be scaled up by 2, etc — but hey, you gotta do what you gotta do. Managed to dodge the bullet on that one, phew!
Balance and scoring
Ugh, this is probably my number one regret with the game. The scoring formulas and balancing of the minigames is pretty terrible and I really didn’t test them very much. In particular the dance game seems to get people a lot — like rhythm tengoku/rhythm heaven, it’s grading you not only on whether you memorize the sequence correctly, but how on-beat you are. Of course, I had no time to make sure the lag calibration was spot-on, so it’s not perfect at all. =( In the compo version, some players also misunderstood the goal of the Letter writing game (do it as fast as possible), and players also tried to figure out which word made grammatical sense in the Texting minigame, instead of realizing that you just needed to tap on the YELLOW highlighted word. In the post-compo version I tried to be much more explicit with the directions for those games, and made the tutorial overlay display across the entire screen because people were probably skipping the text due to being distracted. Some people also got confused with the Stargazing game in the compo version by pointing to where the star was going rather than which side the star was on. In the post-compo version I solved this issue by getting rid of the stars that start on one side and travel towards the opposite side, to make it more obvious. We also made some tweaks to the dance minigame during development to make it more clear. Clarity in instruction is always super hard to get right, I guess…
Initially I had planned on rebalancing the game and even adding an “Expert Mode” where each minigame gets harder (e.g. harder rhythms for dancing, longer letter and text, more arrows for knitting), but in the end I realized that Grow Your Love really isn’t =about= the challenge and the scoring anyways, so it’s not that important to have Expert Mode or anything like that.
Not enough time (as always)
Some of the issues with balance and scoring were a direct result of me having almost zero time to properly test and play through the game, as I was still working on core features like the main menu.
Several things accounted for the time pressure:
- Making seven separate minigames just takes a lot of work, period.
- There was a nontrivial amount of work put into making a horizontally-scrolling menu workable; I couldn’t just copy the vertical menus that I usually put into all of my other games.
- Certain aspects such as Free Play mode ended up not being that important to the overall experience but still took time to implement and get right.
- Did I mention making seven different games?
Overall there was just a LOT of stuff to get done; not only did I have the 7 games to program, but there was the menu, jukebox, liner notes in the jukebox, free play mode, scoring algorithms, the intro scene, record saving, tutorials, tutorial skipping, and don’t forget the ending scene! This was probably the most hours I ever worked for an LD, and I felt =exhausted= afterwards, jeez! Even after submission, I was still finding stupid bugs that I needed to patch up (at this point I still hadn’t ever had time to actually play through the entire game). I don’t know if there was really anything that could have been done about this, aside from maybe not worrying about Free Play mode…perhaps we could have gone down to 5 minigames, but I really think 7 is a good number and that we like all of the ones that managed to make it in. Honestly I think we just did the best that we could have! =P
All in all, Grow Your Love was great fun to make and I hope you guys enjoyed playing through it as well. As always, we learned a lot, and in the future we probably won’t try doing a warioware-style collection of minigames again. I know that it worked well this time around, but it was just too much work! It also caused an issue in that it took a long while before we had a single one of the minigames completely finished, as we were sort of working towards a lot of them at the same time. One of these days, I’d like to have a LD where I actually finish the work for the main game EARLY instead of scrambling at the last minute! (Melody Muncher was sort of close…).
Thanks for playing and reading about our game!