Otherworld - a game I'm working on

Yeah, foraging and social.

Recently I've been pretty frustrated working on this on my own. I have so many testers, yet they aren't doing anything. I added a thing that keeps track of when people log in, so I can tell absolutely no one but me has logged in since it was implemented. It's been two days. Nobody comments on my plans, so how am I supposed to know if I'm headed in the right direction, or if everybody here just says "great job, keep it up" without ever criticizing anything. I know it's not all great. I'm a perfectionist so often I set my goals too high. I could use someone who can see the core of mechanics and what is absolutely the bare bones that are needed to make them run.


One of my new ideas to simplify the hunting system is danger ratings. Each animal will have a number from 0 to 100 that defines how likely they are to cause damage. Being in a group reduces the chances of being hit, but it's not linear. So combat will be resolved by statistics instead of individual rounds. If the animal hurts someone, there will be a damage/wound pool that will be divided between hurt people. It can all go to one person if that person is unlucky. It should check how much damage the characters can inflict with the weapons they have, how long it takes them to take down the animal, and if the animal kills someone before that then they most likely fail.

 

At this point, I'm not sure if testers are so useful because big features are coming out so slowly that I can mainly test them myself. I would benefit more from people who actually commented on ideas. I've been getting my testers from Cantr forums.

I have the status page that marks the order in which I'm planning to implement things, but I've been skipping past the top three for ages and instead kept coming up with new ones. My idea on healing currently is that people will be able to do basic first aid without materials, assuming that they have access to some natural materials in any environment, but more advanced healing will require cloth bandages and stitches.

 

Would someone be willing to help me code? It's really not that complicated, I just have trouble getting started. Most of the game is just about making queries to the database and nested if conditions sometimes affected by randomly generated numbers. It's done in php. What complicates it is checking that data is valid before proceeding. For example if a character does something, you need to check that they have enough AP, that they're in the same location where the target of the action is and so on.

What's more complicated is coming up with values and mathematical formulas for calculating chances.


This is what I'm currently working on.

 

I made a lot of progress yesterday and this morning, but now I ran into a roadblock. I seriously need to simplify the combat system. Each weapon should only have one damage value, but there can be multiple attack types that use a different multiplier. When a group attacks an animal, their damage potential is counted together and compared to the animal danger value, which ranges from 0 to 100. If there are multiple animals, the combined power probably shouldn't exceed 100 as the chances of winning quickly turn negative.

The chance of an animal dying is (1 - dangerousness/110)*pow(grouppower;0.1)/2
The chance of a group member getting hurt by the animal is dangerousness/110*(1/pow(grouppower;0.1))
For each case, minimum is capped at 1% and maximum at 99%.

Instead of binding the damage solely to weapons, it could be affected by skills. A skillful fighter with a bone knife is likely to do more damage than an inexperienced fighter with a sword. But we can think about that later.

If an animal has a danger rating of 50, the group needs a grouppower of 500 to have a 50% chance of a kill. If the danger rating is 40, mere 100 power has a 50% chance to kill. So it escalates rather quickly. We can tweak which ratings we give animals. It no longer works as linearly as in the mindmap. So we should think about scenarios, how many people with knives or spears does it take to kill a rabbit, or a boar, or a mammoth? Based on these, we can figure out the strength of each weapon.

We can also tweak the exponent, but for example if it's 0.2, it already drastically increases the chances of winning, so the best value would be somewhere between 0.1 and 0.2.

 

I did some thinking and I had this idea that will make everything easier. Instead of tallying up weapon efficiency points, the chances of success will depend on the number of people in the group, and later skills and weapons will work as multipliers, making certain people count as fractional or multiple people, depending on whether their modifiers are smaller or greater than 100%. The beauty of this system is that the number of people should matter more than the weapons they use. If you sick 20 people on a mammoth, they have a high chance of killing it even if their weapons sucked. Try that alone and you're more likely to end up as a pancake.


An animal with the rating 100 would be like the ultimate super beast, like a dragon or something. We might never have one of those, but it's good to decide in advance what the bounds are.


I slightly tweaked the formulas because of course they need to be different when they are based on people rather than arbitrary values. In the beginning, I can implement the system without taking into account which weapons are used, as long as each hunter has a weapon of some sort.

Later on I can post a list of possible weapon types and let people brainstorm what their multipliers should be.

Edit: Oh, and what I forgot to mention is, the danger ratings can potentially be used for human opponents and packs as well. If you have two animals with danger rating 50 and they hunt as a couple, that doesn't mean their combined rating is 100 but something in between. I will figure out a formula for it later on.

 

I checked it out but it sounds like that sort of a thing is for a story driven game, where as Otherworld is more like an open sandbox. I think they are too different to work together.

I'm currently in the middle of changing the resources system to rely on main types (such as birch, camphor, mustard as opposed to birch bark, birch wood, birch sap, camphor, camphor wood, mustard seeds and mustard greens all being separate resources). The new system defines which subtypes are possible and when you find a deposit, you can harvest any of the subtypes. In the future it will be bound to seasons, so that you only get flowers in the spring and fruit in the fall, roughly. Also I think in the future, I make it so that rocks and ores can exist as blocks, gravel, sand etc. and be used for different things, and possibly also require different tools to harvest, but I'll do that later.

 

I think festivals should be up to the locals. I can make the system support hosting events that NPCs can attend and boost your reputation. My idea was that it would randomly define how successful the event was, and there would be a chance that it would end up in a disaster. You could serve alcohol to make the range of success or disaster wider. More experienced party planners should have a bonus and be less likely to fail, but it should never be completely risk free. When you invite dozens or hundreds of people, you increase the risk of some of them not getting along. This draft I have has lines like "x people made up their differences", "x people had an argument", "x people woke up with a headache the next morning" and so on.

I finished the resource system update and I'm currently working on travel groups. A travel group is an invisible object in your inventory, that other people can enter and their coordinates are updated as you move. I still need to make it consume AP of the group members and not just the group owner. Also authorized group members should be able to move the group as if they were the owner. I also need to implement a mechanic for kicking people, because currently if you have people in your group and they don't click the button to rest, you could be stuck in a location unable to move. That's if it was checking their AP in the first place. It doesn't yet. It would be useful to bind resting to cron. I think it's about time to test that, don't you think?

 

I want the different cultures to be designed by the player community as much as possible opposed to being dictated from above. I have a lot of great ideas (or at least great in my opinion) but for example, I think playing in a culture that supports slavery is a cool setting, but people who want to rp heroes and hate having other people dictate what they can do wouldn't like that. That's why there are going to be multiple civilizations with different morals. Or at least that's the plan. Worst case scenario is, everybody is scattered around the map, running past each other while exploring and never seeing each other. Currently the only way people can generally hook up is if they know outside the game where to find someone. In the future, you can request roles to be filled in your community, and make children that will be claimed by players and you can then mold them into the kind of citizens you expect them to be.

I had feedback from a playtester that she would prefer if in the beginning there was no PvP combat at all, until the community is large enough to sustain violence and death. This made me think of a system where a game merely tells you who won and who got hurt, but the results would be up to roleplaying. At this point, people dying doesn't serve anyone. Down the line, permadeath will be an important aspect of the game, but currently it defeats the purpose of testing if testers die while hunting because healing hasn't been implemented yet. Once we get cron working, we can have natural healing over time.

 

Permission denied in cron jobs comand If it was working like it's supposed to, it should be sending an email once an hour. If it's less frequent, something is off. Anyway this topic I found explains that you need to separately tell it to run it through php because it's a php script, so you need to find the address of the parser. It's usually usr/local/bin/php but sometimes it's different.

 

Anyway, it's obviously not doing what it's supposed to do or all the characters would have 1000 AP by now. The command needs to be edited.