Simplifying gluttony

[iotuegli]

i like the idea of invariance like Learning Ability for inspirationals, that is the way.
but i don't like the change to glutton the same bar half in a session and half in another, not matter what you do in the meanwhile, that's not good

however id prefer some super gluttony foods made from different ingredients with an hidden recipe like: meal system
that will help a lot to reach some soft caps like 50-75 biles but then you'll need to eat different foods, like allways.

and so on, i agree with the raise of multiple bars in the same time, like inspirationals where when you raise S&C you'll gain 2 points and also gaining 1 point for each other bar you have full.

that will help the spread between humors to narrow

[twram]

guide that said keep your humours close together

Yes, that concept is not difficult. But to accomplish that you need to calculate what needs to be eaten in order to raise your lower stats without accidentally raising the highest stat again. There are several more complex factors that have an impact on being able to do that reliably, and on being able to continue doing that as the stats increase and food requirements are changing.

From OP

the system is too complex for newbies, which significantly lowers player retention rate

I don't want the game dumbed down for the masses. I want a complex and thought provoking system and game. So if players are quitting because of complexity, figure out some way to break people in more gradually besides requiring that they read all of the guides and wiki first. Yes, they could read a gluttony guide before they start playing. As well as an agriculture guide, trading guide, base building guide, survival guide, hunting guide, village guide, cash shop guide, artifice guide, tracking guide, wiki and forums.

Bocage

I see your point, could be a good idea. I've often thought complicated software should have a limited menu/UI option that introduces you gradually into its functions. Anything that gets salem more players and keeps them is worth considering.

[Potjeh]

The problem with all the randomness is that you basically have to grind luck late-game, which is the most frustrating kind of grind. Something along the lines of a quest to fetch 20 direwolf paws when they only have a 5% drop rate.

[martinuzz]

The problem with all the randomness is that you basically have to grind luck late-game, which is the most frustrating kind of grind. Something along the lines of a quest to fetch 20 direwolf paws when they only have a 5% drop rate.

Yeah, I think you pin down a big problem there. There is too much chance for failure, even with very high skill, on certain things. Planed boards. Grrrrrr. Oiled boards. Hrhrhrhrhrrrrrgrr. Slotting. Hrmblgrmpht.
Trees. [censored]grrrr[censored]hmbl[censored].

I hate gambling and casino luck games, there's just a bit too much of that in Salem. Expecially with things like trees and boards, where the process involved in making them already takes so much time and or effort.

[jorb]

Seeing how devs are currently rethinking gluttony, it's perfect time to discuss how to improve it. IMO the problems with it are:
- the system is too complex for newbies, which significantly lowers player retention rate
- randomness is too high, which makes it especially frustrating at low purities (which also happen to be the only purities attainable without legacy humus)
- the difference between high purity and low purity food is too high, mainly due to how the purity multiplier is calculated
- timer-based drain is frustrating to work with (hiding how long it takes to eat something doesn't help), and ineffective at promoting food diversity which makes gluttony a very grindy experience

I think the fundamental problem is that the rather high-factored outward appearance of complexity (If by complexity one means something good, which is another discussion entirely) actually merely obscures a rather trivial reality. The game doesn't reward interesting series of gobbling -- which was the intention originally, and the reason for the intricate system in the first place -- but rather the spamming of one single, optimal dish at the near exclusion of all others. The issue thus being that the game de-facto has four optimal dishes -- one per bile -- and that outside of those four there is no real reason to make anything else. This makes most foods and the industries related to them feel extremely unrewarding and boring in their payouts. Stuff needs to be exciting and worth collecting, and food isn't right now, as the only meaningful question one can really ask of it is "Is this the optimal food for this bile?" if not, then there is little need to ever bother making it, and the industries related to them can be considered a waste of time.

The entire batch of numbers presented by a dish could essentially be broken down to "Food points awarded per unit of time spent gobbling this item" (The randomness element being considered a mere slight distributive variance.) The food items thus have essentially four stats, only one of which is relevant at any given time, as you can only level one bile at a time. Compare this to Haven, in which food has some nine dimensions (Type and one for each of the stats), all of which can theoretically be considered relevant at all times. The variance bonus in Haven also has more integrity, and the one here feels rather artificial, and, more importantly, doesn't matter a whole lot.

Thus: If I wanted to be harsh I could argue that deciding what food to gobble is in Salem a matter of linear comparison along one dimension. There is very little beauty in that.

I don't think the randomness per se is an issue. Haven has a similar element of randomness in its food system.

I agree that randomness should be used with care (And it can be argued that we've used it too much in Salem. Using it the way we have has been an intentional experiment on our part.) , but over larger batches it can, again, be considered a matter of distribution, which I think is fine.

The actual minigame in gluttony of working against the clock and whatever is the one thing I enjoy most about the system. I always enjoy failing at it, and I'd be fine with that part of if it weren't for the above.

The appearance of complexity does indeed turn people off, especially when many of them can probably intuitively sense that the underlying math isn't particularly interesting.

[Tamasin]

Well i've always wanted to be able to add herbs and spices to recipes which would make them better for different purposes. I don't think its a solution to fixing gluttony, but it would be fun

[Potjeh]

I don't think you can really make a comparison between FEP randomness and gluttony randomness. In the FEP system you can opt out of dice rolling by eating just single stat food, or you can take your chances and eat just one of any food type to max out the requirement reduction. Ie, there's this whole reliability vs cost problem, and the optimal solution to it lies in between the two extremes, the exact position on the scale depending on a number of factors such as current stats and raw ingredients availability. In gluttony you can't really choose anything. The only way to stack the odds is raising purity, which is a system that doesn't really exist yet. But more importantly, there'd still be no actual decisions to be made even if we did have purity raising, since higher purities are simply superior in every possible way. Sure, you can make the purification process full of decision making, but the actual consumption of it's products would still remain a mindless grind.

[jorb]

At the very least the randomness is the smaller problem. The food system would not become much more fun if that particular element were determined instead, say, by minutes on the clock, or by some other trivial deterministic pattern. If anything the numbness of its fundamentals would become even more apparent with the only unknown variable removed.

[Potjeh]

Well no, just removing the randomness wouldn't fix the whole problem, that's what a better invariance penalty model is for. Just removing the huge randomness would still be an improvement, though, because it's nothing but a pointless obstacle.

[jorb]

Well no, just removing the randomness wouldn't fix the whole problem, that's what a better invariance penalty model is for. Just removing the huge randomness would still be an improvement, though, because it's nothing but a pointless obstacle.

I would argue that the system would most likely feel even more numb if there was no problem of calculation there at all. At least now there's some skill involved in having a feel for predicting randomized outcomes.

Again: In Haven the FEP rainbow given by a food item matters, gives the food some identity and makes you kind of care about it. Food in Haven can be memorable. Not so much here, and I think that has to be considered the real problem.