Psieye wrote:sanka wrote:crate wrote:So instead of giving you lots of advice and overwhelming you with information, I will provide instead this single, not-as-simple-as-it-sounds piece of information I use to tell apart good players from not-good players: You should almost never move toward an enemy that is in your line-of-sight. Once you understand this you will improve.
This one helped me.
Crate's one-line advice is profound. The link in my sig goes into detailed advice but it all does boil down to that one line: there is almost always something better to do than move towards an enemy. One of the subtler points about that: noise (AI behaviour differences to visual vs audio cues).
This is the most important advice and is one the reasons that the very first thing I make sure to pick up is stones when starting. However I often don't follow this advice rigorously (I always do it to some extent) even if I know its a good idea because I get tired of being super careful. Usually I only rigorously follow it when I am feeling really weak or there is something I really want to be careful of, for example an open Lair 6 and I want to be very careful about death yak packs or even worse pull two death yak packs at once. I have found I play considerably better on octopodes because my constant feeling of being flimsy basically shifts my mind into this method of thinking. Whereas when I play a Formicid, even though there is no tele backup plan, since I generally build them tanky I go a bit more ham on things.
Edit:
I think the main other thing to realize, especially if you background is normal non-roguelike rpgs, is you very rarely are entirely safe. 20 AC with no GDR and 20 EV is pretty good defense for the lair but against something like a komodo dragon at that level with a low HP species it only takes one bad roll to ruin your day. You roll 1 on AC and don't evade and the komodo rolls high damage that can be very bad. Even if the chance for that is only 1/20 or whatever that is still 5% chance since you wind up killing thousands of monsters with many attack that scenario will play out quite a few times.
You may have a 90% chance of not dying and in a normal RPG that is good since you can just reload a save and you will win "easy". But a 10% of dying in a roguelike in general is actually basically close to a guaranteed loss if you push that envelope 100 times. The wrinkle with crawl compared to many other roguelikes is that everything is very variable. You can always hit weakly. You can have an Executioner Axe and 27 skill and you can still roll a 1 do the same crappy damage a hand axe would, it just happens less and you can hit alot harder.
Even coming from other rogouelikes its not just knowing that something is a big or low threat. You need to shift your threat assessment thinking into "What happens if I roll bad and he rolls good?". Unlike many roguelikes, there can be radical swings and fairly bad/abrupt consequences. Many times I opt to simply run from a D4 or 5 Ogre even on a pretty strong meleedude even if I "should" win simply because there have been too many cases where I rolled wrong once or twice and died. I distinctly remember getting one shotted by an ogre on a guy I thought was doing pretty well and I was like "That complete BS! I had good AC!" but then I was like "Ok, well that is the number. He can hit that hard. Its a fact. Its rare but it can happen. If I don't want to be one shotted do not fight that thing without HP greater than that number. Even if its only 5% of the time with those defenses." This is espcially true in the earlier levels where even with GDR your AC is probably not enough to matter for that GDR so you basically have no guarantees at all.
Everything is a risk, everything is a sort of decision threat assessment but Ithink some newer players get thrown by the math of Crawl since it purposely has considerably more variance than other RPG systems, specifically to keep you on your toes.