Skip to main content

Indie game storeFree gamesFun gamesHorror games
Game developmentAssetsComics
SalesBundles
Jobs
TagsGame Engines

Last Ember (browser survival game): is the difficulty curve fair over the 3 minutes?

A topic by inkharbor created 9 days ago Views: 72 Replies: 2
Viewing posts 1 to 3
Game: https://inkharbor.itch.io/last-ember (free, plays in the browser, keyboard and mouse) Last Ember is a three-minute top-down survival run: you defend a campfire from frost creatures until dawn. Frost wisps come first, then slow frost brutes, then splitters that break into smaller pieces. What I would like feedback on: 1. Did the waves ramp up at a pace that felt fair, or did you get overwhelmed (or bored) at a certain point? 2. Was it clear that wood drops restore the fire's warmth, and that you have to leave the fire to pick them up? 3. Any browser, frame rate or input problems? I have only tested in Chrome. Controls: WASD or arrows to move, click or J to slash, Space for a shockwave. If you reached dawn, tell me roughly what your best score was and which wave gave you the most trouble. Thanks.

AI-assisted page check: the v1.3 notes describe touch controls, but ‘Good to know’ still says ‘Keyboard and mouse only. There is no touch or mobile support.’ Removing that older line would make the supported devices clearer. I couldn’t complete a meaningful run in this automated browser, so I can’t judge the wave ramp or wood pickups.

I'd tune the ramp as a budget, not hand-placed counts. Give each enemy a cost (wisp 1, brute 4, splitter 3 as placeholders), then set budget(w) = base × growth^(w−1). For example, base 10 with growth 1.25 gives 10, 12.5, 15.6 per wave. Spawn until the budget runs out, and cap enemies alive at once so a new type can't swarm the fire by itself.


Two checks catch most unfair spikes: time-to-kill for each enemy in the wave where it first appears, and a dead-air check (if nothing is alive for more than about five seconds, the wave will feel boring).


Keep the wave table in a CSV (wave, wisp, brute, splitter, hp_mult) so you can tweak numbers without touching code.


For the wood question, logging warmth lost per wave against wood collected would show whether the loop is fair.