So one thing I'm already noticing is that colonists with the same memories are reporting to have those memories influence their emotions in different ways, and this variation seems to change over time as the memory gets older... but in unusual and rather bizarre ways. First Example Day 1 in a colony and two colonists have the "Had a nice trip here. What a landscape!" memory. Both in the same faded state. 'Colonist A' reports the memory is subtracting from all emotion values. 'Colonist B' reports the memory is adding to the happiness value, subtracting from every other state. (Checking again shortly after that same in-game day, now the memory for both colonists reports to be adding to happiness and subtracting from every other state. Potential display bug?) Second Example Another two colonists on Day 1 have the "Comforted by having a good cower" memory. One memory is fresh, another is faded now (So presumably it's from pre-colony arrival). The fresh memory reports to be subtracting from all emotional states. The old faded memory reports to be adding to happiness and subtracting from all other emotional states. Which seems significantly backwards, as the feeling of being comforted from the cower wears off the memories influence should taper off until it's having no influence at all. Whilst when it's fresh it would increase happiness and lower sadness. Not sure why it would really lower anger or fear, but the same goes with all memories which equally seem to influence all 4 states and a memory that just influences a single emotion appears to be impossible. So perhaps it's more simple to balance if every memory in the game influences all 4 rather than having memories that only impact one or two emotional states. Potential Anomaly Influences Some other unusual memory influences currently observed: * "Stomped feet and felt better about it" (Faded) - Subtract all states. * "Ate something recently" (Faded) - Subtract all states. * "Comforted by the sight of brave soldiers keeping the colony safe" (Fresh memory) - Subtract all states. (Interestingly another colonist has the same memories and those instances boost happiness) * "Killed some wild game and felt quite good about it" (Fresh memory) - Subtract all states. * "Bid farewell to loved ones, this is going to be an adventure of a lifetime!" (Faded) - Happiness boost only. Even if eager for the adventure, leaving loved ones should give some sadness. Initial Thoughts Putting aside the possibility that memory influences on emotional states are just out right broken and has been for some time it just wasn't so easy to see that was the case, or that the new interface is suffering from presentation issues and is misreporting memory influences, for now.... It would seem that memories influence a colonists emotional state in a variable way that changes from colonist to colonist even if there's no obvious personality trait to explain the discrepancy, if that's the case then just spewing out behaviour file figures means very little as it would appear the game during runtime is applying other factors against the initial base figures to reach new results for each colonist. Time almost certainly seems to be one of the factors that alters the emotional impact as it gets older, albeit in rather unexpected ways (Such as a fresh happy memory reporting as subtracting from all states, but then reporting as a happiness boost once it's old). It would make sense that a colonists personality traits would factor in to how much to alter a memories influences not just in terms of by how much it changes the emotional states but also in what direction so that way colonists can have some sense of having their own likes/dislikes and other little quirks. But some of the observed variation between colonists seems to go beyond that as well. For example the "Felt comforted by seeing soldiers keep colony safe"... the colonist that reported it as a full subtraction memory didn't have a "I hate feeling comforted" trait or anything else that would explain the difference from the colonist who got a positive happiness influence from the same memories. So what other possible influence factors could there be that could factor in to what would on the surface appear to be nothing but a positive memory, reporting itself as having a negative influence on all emotions?
There are a few things going on here, some are bugs, and a few are issues with how we are representing the data. I don't want to tell you exactly what is what because the interpretations are valuable to see how we are doing, but this is my primary focus right now so I appreciate the feedback ☺
You tease. Not even a little indication? hehe Not had too much chance to poke around since 42d, but I'm currently considering the possibility of some kind of reactive traits where colonists personalities are actually evolving as events occur and moods change. Where a colonist might meet a fishperson, find them interesting, develop the fishy behaviour trait. Which then serves as a additional factor for changing how memories to do with fishpeople relate to mood. Also the possibility of some interrelation between emotions at certain thresholds. So where a colonist leading emotion may be happy, if their anger emotion is at a certain threshold then memories will have different influence outcomes. That way there is actually more to colonist behaviour than just "What's highest?" and actually controlling secondary emotions is equally important for trying to act as a passive influence on how colonists behave. And that's why different colonists are reacting to the same memories in varying ways beyond just traits.. But so much needed to properly put it through its paces, get the right situations together and see how it plays out, and not enough time! What's really needed is the ability to isolate a single colonist in a 4x4 room and just throw all sorts of situations at them. Then add a different 2 colonists in a different room and repeat.... it wouldn't really help in terms of feedback with how it plays out in an live environment but would certainly help with compulsive curiosity! Debug dev commands! Debug dev commands! My kingdom for debug dev commands!
The memory stuff looks like it's going to take some Creative Programming to resolve so I think what we have now is going to be what is in the next stable, but I'll be experimenting with it more as soon as that's out the door. Generally, as I said in the blog post last week, I'm working toward actually showing the entire state of the character, even if it means some simplifications to make graphically representing it not a nightmare, so while some of the inconsistencies are by design and some aren't, either they'll be getting ironed out or i'll be tweaking the UI to show you why they look the way they do.
Right... new theory! The arrows aren't related to a memories immediate emotional state, but rather act as a "Points over time" indicator. Whilst a memory is showing a up arrow on the emotional state it doesn't mean "This memory has increased this emotion" but rather it means "This memory is currently applying a positive xx amount to this emotion over time". If it's showing a down arrow it means "This memory is currently applying a negative xx amount to this emotion over time". In the case of faded memories it means "This memories main effect to this emotion has worn off and now the points added/subtracted from this memory are gradually being cleared". So you could theoretically have a colonist with positive memories actually lose happiness over time if the gain 'per tick' is less than the total sum of other memories loss 'per tick'. And would explain why happy memories sometimes display a down arrow to their happiness emotional state if the 'buzz' from the event has worn off and how the net benefit is being removed.
Not really tried 43B much yet, but the new markers and how the difference seems to relate to the gap between highest and lowest contributors give the impression that the current emotional state is the middle point of all the various contributors. So if there was: Memory 1 Anger = 40 Memory 2 Anger = 5 Memory 3 Anger = 5 Total anger state would be 16. And for Memory 1 Anger = 40 Memory 2 Anger = 40 Memory 3 Anger = 40 Total anger state would be 40. Which wouldn't really make much sense, as if someone was pissed off then the various additional annoyances would accumulate to aggravate that anger, not mitigate it. If I was in a bad mood in the morning and opened to fridge to find someone had drunk the last of the milk I was going to use for my morning coffee, I wouldn't become LESS angry. lol Same with happiness, having positive memories of something amazing happening wouldn't end up making me less happy because I also had a good nights sleep. Still, just initial reaction to 43B and the new markers having only been playing the build for 5 minutes.
You're actually pretty close! But we don't have those problems unless someone messed up on the writing of some memory data.
The initial 43B reaction post was mostly based on looking at early colonist memories that were meant to represent events prior to arriving, was noticing things like 3 positive value memories to happiness, none seeming to subtract (or null out) happiness, yet the colonists actual happiness state was around about what I would call the middling point of the 3 positive values (going purely on visual interpretation of rapid hovering between the three and making a judgement call ) After observing it longer as memories entirely sourced from colony events took over from the seeded memories of colonists pre-colony times, I noticed this oddity where multiple happy memories of differing values caused a significantly lesser amount of happiness overall not so much... it's still very much comes across as still happening but not to the same degree so perhaps I'm just adjusting how I make sense to the little markers. I do however get the distinct impression some memories are meant to detract from a emotional state rather than add to it, but as all the directional arrows are relative to the current state of the memory (E.G. A anger = 30 memory would show as a downwards arrow to the 30 mark on the bar if current colonist anger = 50), it can be somewhat difficult to tell what memories are serving as a nulling influence, and which are serving as a positive influence but of very insignificant amount relative to the current value of that emotion. I'm assuming now that memories that serve as a detraction do so by offering up a <emotion> = 0 factor, which means the calculated total for the emotion ends up balancing out lower than it would otherwise. So I'm still very much leaning toward the direction of current emotional state is relative to some kind of middling point of all factors contributing toward it, as it certainly doesn't seem to be accumulative, otherwise eating a sausage, having a little dance and sleeping in a bed with no reported happiness = 0 memories should skyrocket colonist happiness beyond the values of any one of the memories, whilst presently the 3 combined seem to lead to happiness state somewhat under the value of the greatest memory but significantly above that of the lower end of the influential memories. Side note: Doing a little dance of joy seems to have a very low happiness influence footprint. Not sure if that's intended.
In terms of traits and such, my impression at the moment is they have no direct influence over emotional states. Instead they serve as a controlling factor for deciding what memories a colonist will and will not generate when faced with certain situations (and by also dictating what actions a colonist will consider performing to generate new memories). So a 'Brutish' colonist won't inherently be more angry, but rather will be more open to performing acts which will then cause them to create anger contributing memories. And a doomed colonist won't always be more sad, but rather will be more open to reacting to events by picking variant memories that contribute more sadness versus the memories picked from the valid reaction lists created for a colonist who isn't 'doomed'.
The traits part is very accurate. The guesses about how the memories impact the state indicate to me that a thing i thought might be necessary (but was going to be a pain to write) might indeed be necessary, as i think it's the bit you're missing in order to form an accurate premise. Thanks for indulging me with your guesses, this is actually really helpful in figuring out what parts of the system are clear and which are opaque.
