It's a pretty fun game and gives me a lot of nostalgia for the tower defense games I used to play on Andkon.com.
This isn't really a bullet hell game, but there are a lot of bullets, at least. It's not really appropriate for this jam but it did fit the theme.
In general I'd discourage the use of AI as it will encourage you to pursue and settle into inefficient workflows. In this game it probably did not detract from the quality, but try to figure out how you would go about solving problems you'd usually mentally offload onto AI -- you'll find that if you understand the investigative and troubleshooting process firsthand you can actually work MUCH faster and more efficiently.
Looking forward to future releases from you! Contact me if you ever need fast-paced music (but please note that I don't work on projects that use significant AI assistance).
Play game
Flow Shrink's itch.io pageResults
| Criteria | Rank | Score* | Raw Score |
| Audio | #22 | 3.444 | 3.444 |
| Fun | #33 | 3.278 | 3.278 |
| Overall | #44 | 3.000 | 3.000 |
| Visual | #44 | 3.222 | 3.222 |
| Theme | #63 | 2.278 | 2.278 |
Ranked from 18 ratings. Score is adjusted from raw score by the median number of ratings per game in the jam.
Are you using any third party assets?
It's barebones HTML + CSS + JS, no engine no frameworks. I started working on a TD concept a few days before the JAM. Did most of the work during the JAM and tried my best to adapt to the theme and setup.
Are you using any AI generated assets?
I have used an AI coding assistant. It helped for general coding, helping with procedural rules for organic shapes and sounds. It also helped when creating some Python scripts to simulate meta, the game is still unbalanced.
Comments
Really cool game, I love the idea and the vibe! Like you've said, there are some balancing issues (I had a volt and green guy combo at the start that demolished everything almost instantly). Really appreciate you disclosing your use of AI, I recognise that it does make me feel worse about the game, and I think it's admirable that you did so (seemingly) knowing how it may effect your score. ty for the cool game and being a cool person!
No, you are breathtaking! Thank you!
As I was saying to Youssy too, a major update is inc that I've been working on during voting lockdown. Besides the big AUDIO overhaul (that one took around 3 days), a lot of the bugs have been fixed, meta update, the curve has been softened (easier early and super late game).
Revamped the rarity system, well we had one but it was glitchy. With a bit of luck you could highroll a legendary 2-3 ability early game and watch your game be broken for the first 10 minutes, now you cant even get epics/legendaries early on unless you spend some LVL up points into increasing your rarity, but you don't want to spend those points early on, you need towers and basic upgrades.
A lot of performance improvements too.
------------
Now I have one more dream, around 200-300 wave, tbd, to unlock an Ascendance system where you start over but it's harder, though you have some perma (rogueLITE) points to invest in a tree. Scaling to infinite with more and more units and more projectiles doesn't go well performance wise and music / clarity wise either. Time will tell if this will get implemented, but if it does that would make it the final vertical slice I had in mind.
very fun game. the sound effects and music were so good, and i like how you implemented the theme i mean shrinking those big circles into bubbles was so fun lmao :D , and the art and UI for the game was so well made.
great job :)
Thank you very much <3!
I actually just finished a major update that's been in the works the entire week while the uploads were locked. And the music was fully revamped, there's a general metronome beat and each tower can hit every other beat, or every beat, or 1/2, or 1/4 or 1/8. But the most important thing is that a tower will hit if it has a unit in range AND when it's the turn (sound wise) to shoot.
This means I had to remove ALL ATK SPEED modifiers from all card upgrades and support towers. And each tower on it's upgrade life from Common1 to legendary 4 goes up in clear musical incremental steps common1 to rare4, then reset (makes up with dmg buff for the reset), then again epic 1 to lege 4.
The results is something that sounds a bit more like music to the ear, and all instruments (the towers are being rebranded) can only hit the 4 most known chords (G Major (I) D Major (V) E minor (vi) C Major (IV)). It's not there where I want it to be, but it got way closer; I'll upload new build tomorrow after lockdown is over.
Whenever I check an entry, I play it first, THEN I check the description after for more context about the game. With this in mind, I play your game, get really confused as to how it works, I figure it out and enjoy the soundscapes (as a music guy), then I realize something is VERY off. Everything feels very out of place, from the text to the game as whole. It was as if you didn't follow the theme and did something else entirely. Then I go check the description and everything makes sense. The way you implemented the theme to a prototype you were already working on is kinda lazy ngl. I didn't even know about the shrinking part of the game before I checked the description. Also, Ik you said in the your disclaimer that you used AI for the coding and everything else was your personal ideas, but I'm a bit skeptical cause it feels like AI was used in other parts of the game as well (Just a hunch tho). Very disappointed that a passionate person with lots of experience and cool ideas ended up giving in to AI ):
Hej hej and thank you, it shows that you care and that is very important for the gaming community overall, so keep it up.
I am grateful for being part of this JAM, having my entry accepted and pushing myself with all the other devs to deliver a full experience.
My background and passion comes from a different angle and maybe sharing a PoV helps. I do not try to normalize AI but I need to be up to date in order to survive (not as a gamedev, this is a side hobby activty but in my field).
Before it gets any better on ASSET and AUDIO generation (there are crazy tools, we just don't have acces to) - so ART, first it came for all the classic development. And this is SAD and I would never try to defend it. The coding layoffs became the norm, at way way way higher level than in ART, because coding tools not only are accepted in the industry, but now it's required you are confy with them as part of getting a job or project.
And while I'm older (close to 40) and I code less these days and plan/organize projects more, I'm a bit worried about the future of DEV since this area I know and understand a bit better. I was having recently a talk with a friend that's teaching at a tech university and the conclusion was that the workforce does not need JUNIORS anymore. Unless you need to grow 3-4 super good students to know the ins and outs around and carry on the legacy (and don't end up with a magic black box source code), AIs in the hand of a few experience developers can cover up for a dozen of juniors. And this is a tragedy because youngsters feel the need to also jump with Claude and stuff and end up not understanding wtf they are building, but they are just trying to survive. This is a recipe for disaster but then again we've seen it a few times in the past.
At some point webdev used to be pure (back in my age granpa), but then all sorts of layers and frameworks arrived so new devs only learned those tools, right and it looked like a disaster too. But before that there were assemblers, and layers and layers of tools and compliated libraries were added so the code looks and gets structure to make more sense. I don't know assemblers, I used to make it at uni but then never used it ever again.
So from a dev perspective, I have at times this "WE're DOOMED" feeling, nobody will know to fix our AI code but AI, the end is near. But I think humanity will thrive and adapt, we always do even if sometimes a bit too late. It's just we don't know how to transition and what elements will leave or stay from our current pipelines.
To touch the ART part, my late father was a guitarist, my brother is a sound engineer, my wife is an artist, my mentor was my art teacher from 10 -> 16. I tried to learn the piano (big fail w/o musical education) and draw and stuff, at least ART has always been part of my life and I have learned to appreciate it and taste it in my own ways. I think AIs are coming after this too, it's just a bit behind since coding was more abstract and easier to automate. And this is PURE SADNESS. Let's not talk about content on platforms, I'm sick and tired having to filter each youtube that I open - is this a super low effort full AI made video that's been spit into my feed or it's actually good content?
And everything I've been saying about DEV will be valid (if its not already) for art too.
---
Now that hopefully we're best buddies and in the same boat, let's talk about games.
I have TENS OF THOUSANDS of hours in games, I am not kidding, it has been a blessing and a curse, from Commodore 64 and NES, to all console generations and PC, I have seen them all. I Have all consoles, switch, steamdeck (LOVING IT), Oculus (meh) and around 50 board games too. When I open a game, I judge it based on how good of a game it is or what does it tries to be. I buy all possible early access titles, but super early and love to compare their 0.1 with 1.0 builds and usually good games have the DNA in them from that prototype.
There are so, so many ways that can make a game bad. AI slops is the newest ones and itch and steam are full of it. But we also had years of asset flip, another plague with often the same results. But hey, what about cloning games? Is DOTA a nice custom map in w3, let's do hon and league and dota2 and hots and smite. Did PUBG exploded (well it was more ARMA with Dayz), let's make fortnite, apex, warzone.
But the worst clone case for me is Vampire Survivor (we're in Bullet Hell anyway, and I know this is not the first entry in genre but for many it was and I love it personally) - everyone had to make a clone for it, and it's still happening years later. Yes, sometimes a good game comes out of it, I have hundred of hours in Brotato and Danger 5 cleared on half of the characters - maybe death must die was decent too and megabonk... but there's also HUNDREDS of clones, I DON'T LIKE IT.
What about monetization? Oh, 10k hours in dota2 over 16 years span, I COULD HAVE BEEN A PLANE PILOT they say with 10k hours in. I loved the first battlepass that they added, and the way the funneled the price into The International, it was organic, it was magic. I detest how everything today is a battlepass and why not loot boxes. GOTCHA! This is another plague.
What about putting up content on a conveyor belt where coders and artists are consumed and have to release a new game (with or w/o AI) every year. Do you need a new Fifa game and ulitmate team, you don't. Yet they do that, they fight for your wallet and attention.
Bad AI use has been the tip of an iceberg in an industry that turned creativity and joy into a cash grabbing machine, with MBAs and Psychology at the top on how to squeeze you more. And this is the real tragedy, not only AI slops, but ALL OF THEM COMBINED.
---
For me AI is just a tool, I adapt and learn to use them right, to combine the best out of all to create a good experience and learn whats all the fuss about.
Did I create a good experience, maybe not 'but I tried, didn't I? Goddamit, at least I did that" (good movie). In the last week of the event I have pushed around 12-14h a day, till 4-5AM, just to continue from 9AM, a bit of work for main job, then back to my game since we had deadline. Hell the cards balance, and they are still off, took me 2 days of straight wor and those godamn sounds, trying to make the user to 'make the sound' while playing the game was so hard and I'm so far far away of what I want to do here. I'm in love with everything procedural generated, and I don't mean some map in PoE or Dead Cells, Minecraft or Rimworld (love all of them btw), but rather No Man's Sky procedural level. I want to use AI to learn and iterate faster on the procedural area, and learn what can be done. This is my 2nd game ever as a hobby, the first with AI - there could be a a 3rd, time will tell.
So this is the only part where I say your detective work was a tiny bit insulting to the effort end ethics (you can still be agains AI, you can score me with 0 stars etc)! But I didn't lie, I never tried to hide what I was doing, got huge disclaimer day 1, I never said to be doing anything else than what I was doing. You don't have to be disappointed nor to assume that I was kinda lazy or deceiving or that the stupid story and concept were not my personal ideas (I was proud of it oups haha). Maybe if I wasn't so upfront with everything, you would have said, yeah, a bit off but I had fun clicking the belt and that would have been it.
I can tell you hate AI work and that is great. But NOTHING about this game was low effort on my side. Yes I used AI tools, I used claude, I used gemini but they are SO INCREDIBLE STUPID (especially on art and sound, or maybe I don't know how to use them, or maybe I didn't had access to more advanced tools).
I took the time to answer you because I do think you deserve a clear answer not just haha FU, survive, adapt (I don't think that's true btw, I've seen that on reddit a lot). I care about gaming, I'm passionate about it and I think navigating type of AI use will become just as comparing asset use, monetization, copy cats and the list goes on - is the product trying to be any good, maybe. And I don't say my game is good, I'm saying we have to be open that not everything that says "I HAVE USED AI or ASSETS" is a bad product or deceiving.
And now I'm behind 1h with main job, it took me 1hour for this reply (AND THIS IS ALL YOUR FAULT hehe) on top of last week half asing my work just to finish my game.
Much love and keep the good games flowing. Cheers,
J
PS: OMG. HUGE wall of text, english not first language; sorry if it goes places, I hope I managed to share my POV; I don't want you convince you of anything, but just show how it looks from my side too.
You can totally look down on it and punish the use of AI here or on others game, I'm not trying to get you on my boat; I'm not even sure I wanna be on this boat either. And vote with your wallet too.
Just don't assume everyone is out there trying to be a cheater, give them a chance. But I think you got that already and I'm grateful for that. I honestly think I got some good reception considering the use of AI, so Ty Bullet Hell Jam too.
Thanks again for carrying and GL HF developing awesome games. And sorry for my wall of text, I cringed a lil bit after I re-read it but the main idea was there, just TMI :))
Cheers,
j
Thank you! I have worked really hard on that part and it had 4-5 iterations.
What I'd like next is to move even more in this direction and maybe add a global metronome to the entire game where the towers don't just hit when something is in range but rather they hit if something is in range and use the length of a note.
This means I'd have to fully remove ATTACK SPEED from any cards and modifiers and have a tower just have options to be full, 1/2, 1/4, 1/8. This way it's a little bit less random. That plus the already working it - 'all music use the same 4 chords' part, and hopefully the audio experience will be even better!
The design here is really interesting, but the walls of text make it hard to parse what exactly is happening or what i should be doing with my upgrades. I don't really know what a SYNTH is and my attention span isn't going to let me read the handbook to figure it out, you know? Upgrades were interesting, though i wish i could have picked up and moved or sold my towers, not sure if i could or not but i never figured out how.
I could see me wasting a lot of time on a game like this, it's visually interesting with a good gameplay depth, great entry! i gotta take a star off for ai use personally, but this is strong and you should run with it c:
Thank you! Selling tower is one of the top requested features among my RL friends that tested the game too. I was trying to be a bit more hardcore (there used to be a card that removes a tower but it's disabled atm).
A lot of more work need to be done but for now I'm happy the builds are locked for review, I can take a little breather. Also, remember you have a CryptRat to beat xD
The visuals are charismatic and attractive, but without a doubt the best part is the sound...
I won't deny that I felt a bit out of place with the story, but all the other peripherals were amazing...
I just got past Sector 100.
Game's fun! It's cool and relaxing, the presentation is perfect, the UI is perfect and I love the number of different types of upgrades you can apply to a single tower.
Thanks for sharing!
NO NO, THANK YOU! Your project was nothing short but amazing.
I got some of my friends hooked on it too, past 3-400 wave and I'm considering a roguelike part where you reset but it's harder, since scaling gets harder and harder 250+. But then again, as soon as I get some time, I have some more ideas in the AUDIO part that I've been putting on paper and that's going to be my first post JAM major update.
This was a very interesting idea, and really impressive that it was made with vanilla web coding. I honestly couldn't tell while playing because it felt very polished. The presentation was great too, especially the sound. It was really fun to click and hear the notes constantly changing within the scale. It gave the game a relaxing, chill vibe. I also thought the different genre elements blended together really well. Very cool and unique entry.
Yeah, I saw what you were going for and it definitely works. It gave it a cool identity, and I always appreciate when a jam game tries to mix ideas instead of playing it completely safe.
Totally fair on having a bunch of games left to play. I'm in the same boat lol. If you do come back for the leaderboard later, I hope you have fun chasing it!
Leave a comment
Log in with itch.io to leave a comment.