|
MissMarple posted:People have blindspots around probability in general, and even with fair results would have messed up expectations. That actually makes me wonder, to what degree are such things mathed out by the developers beforehand (let's say for non-monetised systems for simplicity's sake)? Is there some designer whose job it is to decide "We want X item to drop after an average of Y hours playtime" or "We want X% of players to see this happen during a single average playthrough", and who then sits down to calculate the appropriate distributions and plugs in those drop chances? Or is it more a matter of people plugging in some roughly estimated default value and then adjusting that based on playtesting? The answer is probably "depends on the studio", but I guess with the huge role that randomisation takes in a whole lot of games some sort of industry standard might have developed.
|
# ¿ Oct 2, 2017 14:29 |
|
|
# ¿ Apr 20, 2024 01:39 |
|
KRILLIN IN THE NAME posted:It's already possible, but kind of unfeasible if you're using more than one storefront. Steam keys only come from the developer (or publisher) in the first place - in order to have a key sold on G2A the developer or publisher had to have generated that key in the first place (for reviewers/streamers or to be sold elsewhere like itch, their own website, presumably physical box copies). That leads me to a weird question: If they're generated independently by the publisher, is there any mechanism that checks for a key's uniqueness? IIRC two keys ending up identical by simple accident should be so improbably as to be virtually impossible, but could there be a chance of human error where two sources end up using the same algorithm/keyphrase?
|
# ¿ Jul 17, 2019 15:44 |
|
So recently an indie game I'm keeping an eye on mentioned that they're using photogrammetry to make some of their models, including very prominent ones like character clothes: https://www.youtube.com/watch?v=6zRdjB8s41M&t=30s Now, I've hosed about with photogrammetry back in uni some years ago, and while it could produce reasonable results even back then, I remember it being kind of a pain in the rear end to set up that also required a fair bit of cleanup afterwards. I'm surprised to that it's now in a state where it can apparently be successfully used even by tiny dev studios (this game in particular is made by 2 people). Now I wonder, how common is this technology in gamedev these days overall, as opposed to making assets by hand?
|
# ¿ Jan 25, 2021 12:09 |
|
For the most part it's really just a meme. It was one of the first 3D representations of a fairly complex object, and people coming after used it as sort of an unofficial benchmark to show off what their own implementations could do. It's got a few aspects useful for that purpose like smooth alternating curves, hollow parts, and casting shadows on itself but really a lot of shapes could do for that purpose. The teapot just happened to be first and people ran with it. Hell, in my university 3D course we recreated the thing from scratch by slapping coordinates into matlab to demonstrate the basic principles of 3D object representations.
Perestroika fucked around with this message at 10:52 on Apr 13, 2022 |
# ¿ Apr 13, 2022 10:46 |
|
Another factor is also that realistically heavy ragdolls usually don't behave in a way that feels right. In most environments, a model can be controlled either by pre-rigged animation or by ragdolls, but not a combination of both. So with "heavy" ragdolls, the moment you switch from the former to the latter, the model will usually simply slump straight down into a heap. That tends to feel off to players, since usually people don't just fold together like that when they're hurt. They flinch away, they stumble and fall lengthwise, they throw out their arms for balance. By making the ragdoll lighter and/or exaggerating the impact forces involved, you end up with a result that, while still not realistic, still tends to feel a bit less wrong than the alternative. Plus it's usually just more viscerally fun to send people flying. Now, there are some other ways around that. IIRC Call of Duty does a hybrid thing where they have pre-rigged death animations that switch into ragdolls at a preset moment, which makes it feel more natural and also helps to hide the transition between animation and ragdoll. But of course that requires you to make a whole new set of animations by hand, and quite a lot of them to avoid overly frequent repeats. Another option is using procedural animations like GTA does, but that's usually overkill unless you either have more money than god or specifically build your entire game around them.
|
# ¿ May 15, 2022 11:32 |