Hey everyone! This week Team Stonehearth is working hard on combat tuning and balance. We’ve been refining this process throughout the release, but it’s only now, that we’ve got all our new classes (Knight, Cleric, Archer), all our new monsters (Kobolds, Ogres, Orcs), our two modes (Normal and Hard) and a few weeks of feedback from all of you, that we really feel like we can make a definitive Alpha 16 experience. So what does our process now look like?
Combat Tuning
Recap from the video:
- To tune combat we create environments (town_survival_test) that contain beds and equipment and hearthlings. We use them as an area where we pit groups of hearthlings against groups of monsters and adjust hearthling and monster stats until the monsters are JUST hard enough to provide a decent challenge.
- The real question is: how do we batch hearthlings into groups that appropriately represent your towns? And what does “decent challenge” really mean?
- Let’s take a look at TownModel, which is a spreadsheet where we guestimate how many hearthlings are in a player’s town, how many of them are combat units, and what level those units are.
- We also use TownModel to keep track of when enemies spawn. Over the course of our tuning, we’ve based this on number of hearthlings and number of elapsed days, and are currently experimenting with making military score a factor also.
- Our monster groups are divided up into seven groups: easy, medium, hard, and epic 1 through epic 4. Each monster group targets towns of certain sizes and scores, and, in hard mode, towns that have existed in the world for a certain number of days.
- Based on this model, we populate the test world with people of appropriate number, equipment, and level, and run the raids at them. Each raid in normal mode should knock off roughly a quarter of one person’s heart, so that the player knows that after a few such raids, they might lose a person. We hope this encourages them to think about making an herbalist, and later, a cleric.
- One thing we learned back at the beginning of A16 is that hearthling HP had too much variance. People with low body scores died immediately, which was no fun, and people with high body scores never seemed to take any damage, which was no fun either. To fix, we took the average hearthling health score, which was about 50, and made this the floor, and reduced the impact of the body score by about half. This means that even low body score hearthlings can survive a few hits from low level enemies, giving you time to react tactically to an encounter by hitting the Town Alert button.
- Monsters in general got more HP and less damage per attack, so combats would last longer, giving you more time to make tactical decisions.
- We also learned early on that equipment had a huge impact on combat. Low level people with good equipment could survive raids that killed high level people with bad equipment. This amount of variance made it very hard to estimate monster difficulty, so we gated equipment based on character level. Now only people of high level can equip high level equipment, making it easier to estimate how much punishment a character will be able to take from a monster before dying.
- Halfway through A16, we also added Hard Mode! After a lot of experimentation, we decided that hard mode players: would be expected to be better at the game, and therefore fulfill the hearthling immigration requirements faster; would probably need to allocate 40% of their hearthlings to combat, instead of the 20% number we had for normal mode; and use tactics occasionally in combat, to optimize damage output and soak for survivability.
- In the end, we decided that normal mode should be a game from which you can recover from a mistake in a few days, but the same mistake, in hard mode, may start a downwards spiral that may eventually cost you your whole town.
- There’s more to monster tuning for modders too: monster stats are now easily accessible in monster tuning .json files, and the town model is a .json file for easy editing.
Image may be NSFW.
Clik here to view.
We’ve tuned a bunch even since I made this video, so some things in the video may not be accurate when we do our next update. It’s an evolving process, and we’d love to hear how you’re finding the results. If you can think of any optimizations, we’d love to hear those too!
Other Announcements!
Streams will happen as usual this week! Amazing! Image may be NSFW.
Clik here to view.
Image may be NSFW.
Clik here to view. Image may be NSFW.
Clik here to view.