IMPORTANT EDIT: StarPeanut (aka The Hero) did some code diving and discovered a bug that was destroying the intended correlation between animal mood and product quality. The data presented here suggest that there is no relationship between the two, when in fact, the intended result is that there is a strong correlation. However, there is a bug that wrecks this. Discussion of this bug, and how to work around it can be found later in this thread. As of this writing, SDV is in version 1.11. I am hopeful that this bug will be patched out in future versions. I will update this header if the bug is resolved. In my main thread, I've been doing a detailed playthrough that tries to be grounded in game mechanics. I've gotten to the point where I need to start working with animals, and quickly discovered that no one knows anything about that. The amount of speculation presented as fact out there on the internet is staggering. Lots and lots of arcane sounding rules about what affects what, with no evidence or data to back anything up. People use reasoning like "I left the barn door open once, and my goat gave me a small milk, so open doors means no large milk." This is my attempt to make actual measurements and try to understand how the mechanics work. To my knowledge, this post will be the only place in the entire internet where any real data has been presented. If you know of anyone who has done any code-diving or empirical testing, I'm interested. Some ground rules: If you have any actual data, I am interested. If you have anecdotes, I am not. The signal to noise on this subject is terrible, and talking about how you did complicated action x and got result y only creates work for everyone else, unless you have solid data showing a clear pattern (and no, "I did it three times and it seemed to work" doesn't count). I am assuming that Concerned Ape is a clinically sane man trying to create a game that could be enjoyed and understood. I expect that if two things are intended to be related, then the effect will be strong enough to be noticeable. If an effect is so weak that you would have to do thousands of trials and carefully record your data to see it, then it probably isn't real, because why would CA code something like that? The tldr summary: Animal mood is essentially random, and has no clear correlation with either animal love or product quality. It is mostly only useful when it reports a specific problem (e.g. "Bessie was not fed yesterday"). There appears to be no mechanical difference between "fine" and "really happy" and I suspect that these are just there for flavor. Animal love is correlated with product quality. The effect is strong enough to measure, with zero-heart animals giving close to zero percent large products and 5-heart animals clocking in at around 30%. I tested three scenarios, one where I went to bed right after tending my animals and not waiting for them to go inside, another where I waited but left the door open, and a third where I closed the door behind them. I found no correlation between any of these behaviors and animal mood/love/products. In fact, the most work intensive behavior (closing the door behind them) actually had a negative correlation with product quality, which I write off as random noise. Measurements were only made on cows, but as a first order assumption, I am assuming that CA did not write parallel but unique mechanics for each animal. At most, I would expect different drop rates, but not actual changes to the game rules. If you have data that contradicts this, I am interested. Remember, anecdotes are not data. Mechanics Love Each heart takes 200 points to fill. Here are some things that affect points: Petting +5 Milking +5 Eating grass +8 Eating Hay +10 Left outside -10 Not Fed -10 No heat in winter -10 The negative effects aren't cumulative. You lose ten points, regardless if all three bad effects are present or only one. Animals are only "left outside" if they can't get to the barn before they fall asleep. Going to bed before they reach the door does not cause them to be counted as being trapped. I don't know what happens if an animal is physically blocked from getting in but you go to bed before they fall asleep. I didn't bother testing this because that's just not something that comes up during natural play. Mood The game uses an internal index number for mood, which is translated into text when the player checks the animal. Some text corresponds to more than one index number. My list so far: 2: sad (rare, not sure what causes this) 13: fine 23: not fed 24: left outside 31: really happy 49: fine (only happens on rain days, and even then, only rarely. No obvious difference from 13) 100: really happy (all animals have this by default until the player interacts with them, and then the "real" mood is selected. On very rare occasions, this number ends up being selected as the real mood. No obvious effect) Data Here is the case where I went to bed immediately after tending my animals. I tracked the behavior of two separate cows, one with zero hearts and one with 5. I had to periodically use a save game editor to re-zero the first cow. I didn't do this every day, so the zero-heart cow is probably scoring slightly better than it should. The correlation between love and product size is probably real. Mood is all over the place and has no obvious connection to product quality. Love doesn't seem to affect mood either. Here is when I waited for the animals to go inside, but left the door open There is no obvious change. The effect between love and quality is still present, so that's probably real. Here is where I closed the door behind them. I did not bother tracking the zero-heart cow, because at this point I was convinced that the love-product relationship is real, and maintaining the zero-heart cow was more than doubling my work. The weakest performance yet, which I attribute to random noise rather than actual code. None of these data are strong enough to draw rigorously scientific conclusions, but they are plenty strong enough for my own casual play. I'm going to stop caring about animal mood (except for when it indicates a specific problem) and I'm certainly not going to bother with manipulating the doors.