Against the concept of balance

[Skeletor]

Balance is overrated.
I could easily argue that  good game design shall not be afraid of unfairness and unbalance but rather deal with them dynamically and reward players who are able to find order in chaos.

Have a look at what happened to an extremely varied and fun game, when the devs decided to pursue above all balance: https://www.reddit.com/r/roguelikes/comments/axyr0l/do_you_guys_think_balancing_dcss_kind_of_ruined/
Or the various "level matching" enemies of all those mainstream games which feel really like movies rather than games (Witcher 3, Skyrim..).

As a roguelike lover, I like the deus ex machina, original crazy solutions to apparently unsolvable situations.
The scarecrow of an unbeatable tank-like PC shall be embraced and addressed with further degree of chaos, not by limiting the latter.

[Krice]

You need balancing to make sure the player can survive. For example you need to create enough food items to balance the food clock, so completely random creation is "unbalanced" and can create situations where food runs out which is somewhat poor game design. But it seems that the right amount of unexpected is better for roguelikes, rather than trying to balance too much. In player's selectable races/classes balancing is a strange thing to do, because it's the "difficulty level" of the game. It's supposed to be unbalanced and some classes are easier, some harder to play.

[AgingMinotaur]

To be sure, unpredictability is the name of the game. And single player games need to take balance less into account than competitive games. Of course, there are games like Brogue, designed around being very balanced from the outset. That has a distinct coolness, and allows for clever use of bottlenecks and other map features. But to discuss the other end of the scale, RLs do seem to lend themselves well to unbalanced design. Random generation itself tends to lead down this road, perhaps, since the outcomes are less predictable to the designer, as well But if you have enough moving parts, the odd unbalanced situations may even each other out in the long run

It's a tried and tested recipe in RLs, and Rogue itself had this: Randomness decides that some positions are unwinnable (like starting in the middle of a monster zoo), whilst others give rare advantages. It becomes very much like a game of chance: You keep rolling the dice, hoping for what you need to show up. It can work if the turnaround of characters is quick/interesting enough, or other factors make the game well rounded. Typical of many classics is that leveling up (aka grinding) can give a more stable mid-to-end game. Or features like nonpersistent levels and random loot provide various ways out of most tight spots. Some RLs are just open ended enough that even locally unwinnable positions are possible to overcome: by backtracking, taking a detour, running for your life, or somehow mac gyvering your way out Some very expansive games like Caves of Qud excel in this department.

Done well, random unbalance provide some very exquisite pleasures, on both ends of the power scale – from beating a very difficult position by spending your resources wisely, to reaching the godlike level where you're slicing through previously feared foes as were they butter. The funny thing is that even with lucky characters, odds are you will make a wrong move at the exact worst moment and still end up bulldozed by some random mob.

As always,
Minotauros

[Troubler]

Have a look at what happened to an extremely varied and fun game, when the devs decided to pursue above all balance

The game has ended up oriented towards competative play, and as a result is relatively bland outside of that context.
Anything that isn't well balanced is going to end up being exploited by the target demographic.
If a few people exploit a certain strategy, then everyone is driven to do something similar to keep up.
And that means how much of the game competitive players actually have access to shrinks considerably.
Or if there is a superior strategy, though unpleasant, then playing with a competitive mindset forces you to do it anyway.
If you care about competitive players, then you will want to do as much balancing as possible so they don't have either of these two issues.

So it's not such a bad thing in that context, but I don't agree when the Crawl fanbase tries to pressure the genre in general to go down this route.
I feel pretty bad for people who enjoyed the game before it had to have so much flavor sucked out to satisfy munchkins.
I would not want that to happen to a roguelike I had really invested myself into either.

The developers steadfast adherance to their erudite zero-sum design philosophy is very interesting though.
In that sense it makes for a truely unique experience to go through at least. One that sets it apart from most roguelikes.
As someone who goes from game to game, I probably appreciate it more for this perspective than the original design choices.
But beating it felt more like doing my taxes than an adventure. I didn't hate it, but I probably wont be coming back for seconds either.

But also on the other end of the spectrum are games like Rogue itself, which often feel more like gambling than a challenge.
Or rather, the gamble is if a round is beatable at all, and the challenge is figuring out when it is and what to do then.
So it's not all good, and not all bad. There are countless different approaches to make an engaging roguelike.

I am pretty conflicted about this greater conflict between Lawful and Chaotic roguelike design in general.
Chaos is often unfair, but then Law is often boring. Even Neutrality often ends up sacrificing too much of what makes either desirable.
I would just say to keep looking through the games, and hope for developers with good intuition, rather than the dominance of any specific design philosophy.