Bart wrote:I can see that Crawlers enjoy risk/luck factor far more than I do. I would like Crawl to be a bit more chess-like and not so much lottery-like - especially if we're talking about getting a losing ticket, not a winning one.
Randomness is a pretty fundamental part of the game. Enough so that it is, I believe, specifically mentioned in the game's design philosophy. Other roguelikes do set out to be much more predictable. Brogue is the example that always comes up in this forum as a roguelike that does a fantastic job with predictability and transparency. But the conclusion that gets brought up every time Brogue comes up is that DCSS is too far away from Brogue to ever be Brogue-ified.
That said, I do think there's an interesting issue here to discuss, but it's much more general and complicated than "sometimes monsters use their most powerful spell multiple times in a row":
some monsters have a much bigger difference between expected threat and maximum threat. It is theoretically optimal to always play the game around every monster's maximum threat, and this can feel frustrating when there is a monster with a very low expected threat but a very high maximum threat.
The thing is, playing around maximum threat is a core part of DCSS strategy. There are tons of encounters that could go wrong but probably won't. I would say a major difference between a good player and a bad player is their ability to play around maximum threat. A good player will recognize the incredibly high potential danger of an encounter early and flee or use consumables pre-emptively, a bad player waits until a monster has already proven to be a threat, by which time their options are much more limited and it may be too late.
On the other hand, there are times when having a threat variance that is too high can be tedious. The best example of this, I think, is distortion-branded weapons before enemy weapon brands were already revealed. Back then, the maximum threat of any enemy with a glowing weapon was extremely high, especially early game, even if the expected threat was very low, since distortion weapons were very rare. It was theoretically optimal to avoid ever fighting a hill goblin or plain orc with a glowing weapon in melee range, but that was incredibly tedious and not doing that was okay the vast majority of the time, so even many good players didn't actually follow that advice. That's why things were changed so that you could see the enemy's weapon brand - because that particular case of very high threat variance was incredibly tedious.
Another good example that is currently being discussed is Neqoxecs - the fact that they're generally unthreatening, but you have to play around the fact that any given turn in LoS of them could give you Frail or Blurry Vision, makes them a very high threat variance for such a common enemy, which many find annoying. Neqoxecs and distortion brands also both highlight a particular feature that goes poorly with high threat variance - they're common. Common enemies with very high threat variance are much more annoying than rare ones. People don't mind playing around a very high threat variance unique, because you encounter them at most once per game, and so playing around the maximum potential danger isn't tedious, it's just the obvious way to react.
Overall, I think this is something that needs to get handled on a case by case basis. There's no good way to ensure things work out accross the board, we can just look at particular cases of enemies with a poor balance of expected threat, maximum threat, and common-ness and determine if things need to change (like people are currently looking at Neqoxecs). Looking at the four specific cases cited in the original post, here are my own opinions on each of them:
- Orc Priests: Debatable. While an orc priest getting particularly smite-happy is problematic, they're fairly rare and have a high expected threat already in the early game. They become a lot more common in the mines, but by then the threat of a single orc priest is much lower, and I think it's okay for a pack to still be dangerous.
- Hellions: Hellions spam Hellfire burst frequently enough that I would say their expected threat is already very high, so I don't consider them a huge problem. They become a much bigger threat in packs, but packs of hellions are just straight up incredibly dangerous, not cases that are usually harmless and unless you get unlucky, so you can act accordngly.
- Deep Elf Annihilators: I'm conflicted on this one. They're fairly rare and have a fairly high expected threat, so on paper I think they're fine, but my personal experience has been that LCS his a lot harder than anything else they do, and I have resulted in a situation where a low health DEA took out half my health with a single LCS and I was forced to flee even though the odds of it LCS-ing again were low and I knew I could kill it if it didn't.
- Orb of Fire: Very rare, and already have incredibly high expected threat. There is nothing wrong with being required to play as if an OoF might double-fireball you on any given turn. There's a reason some people consider rF to be the only mandatory resistance in the game.