Browse GamesGame JamsUpload GameDeveloper LogsCommunity
Indie game storeFree gamesFun gamesHorror games
Game developmentAssetsComics
SalesBundles
Jobs
Tags
Somdudewillson45 days ago

A quite enjoyable experience - carefully picking where and when to dip out of the light and navigating between objectives is a solid core loop. Adding occasional physical wandering monsters and chase scenes works well, thanks to the solid movement implementation. The bits of story are enjoyable too.

It would be great if you mentioned the inventory key somewhere, though, given how important it is. I also noticed that the stability flask is only healing 10 sanity, not 15, and I think the 'Defend' options aren't quite balanced - I did some Monte Carlo combat simulations and they appear to all be strictly worse on average than just repeatedly attempting to escape.

Here's my playthrough:

Reply
Volcanolord45 days ago(+1)

Thanks for the feedback! Interesting point about the combat. I kind of balanced the defend actions around being useful for fights that start at lower percentages as they give you a bit of protection from the attacks from longer fights, it might be better mathmatically to spam escape but you could also very likely just get killed before you actually succeed that way due to damage the enemy deals. 

My general strategy revolves around using guard first then attempting escape on the next turn with a slight defence boost and increased chance, not sure if that would still be mathmatically better to just escape twice but you'd also probably lose more health that way if it were to fail 2 times in a row. 

It was definitely an interesting system to try and balance but I think it was worth it for the theme.

I'll add the inventory key to the games main page, sorry about that I guess I assumed people would try escape anyway lol. I suppose putting it on "I" could have helped too.

Reply
Somdudewillson45 days ago (1 edit)

My analyses were based on average damage taken, to be clear - it is possible guard is worth doing in specific circumstances if taking standard deviations into account. And you might be underestimating just how strong "just try twice" is - if the chance is 50%, 'guard' will give you a 60% chance, but trying twice is a 75% chance. Plus, you have a chance to just not take any damage, and in a game with this many encounters attrition really matters.

Reply
Somdudewillson45 days ago

I redid the simulations, tracking extra statistics, and it seems that there are specific circumstances where 'guard' will give you somewhat better survival rates, though you still take more damage per fight on average. 'Pray' also seems to be basically the best option if you have immediate access to the autoheal after the fight and are okay sitting at 60 hp. 'Swipe' seems useful only against low-damage enemies like the Deer Reflections and the Eyes.

Reply
Volcanolord45 days ago(+1)

The results are interesting. I definitely was aware the pray was designed to basically be a get-out of jail free card if you were aware of your surroundings and near a light, this is why it's got quite a high cost so if you use it poorly and end up in another fight you'll be in trouble.

I was definitely thinking of adding a stronger guard at some point midway through the game but couldn't really think of a good reason to suddenly give you a new ability. I want the player to feel quite weak and, well, defenceless so giving an upgrade midway through the game just felt wrong. 

Swipe honestly started out as just being an ability that added more escape chance at the cost of more sanity, but I found I never really used it so I kind of redesigned it be a bit of a gamble to add some escape chance while also having its own chance to instantly escape the battle. I actually added a mechanic quite close to the end of development to it where if you use it at 80% then you can instantly guarantee your escape by using it but i'm not sure how useful that was. maybe if it came with a reduced cost it would be worth doing.

Thanks for doing this analysis! It's interesting to see as I was unaware of a way of doing some raw data testing on the combat before.

Reply
Somdudewillson44 days ago

It's only really possible because player choice is so limited and enemy behavior is nice and simple. I just threw together a tiny JS fiddle that tested different scenarios and tracked statistics in a loop. I think 'Swipe' suffers heavily from the fact that taking an action other than attempting escape more than once is virtually always a terrible idea - and you need to do so to get escape chance to 100%

Reply
AboutFAQBlogContact us
Copyright © 2024 itch corp · Directory · Terms · Privacy · Cookies