Variant with diverse character combos and great item system?

[stabwound]

Yeah, I really like the barebones design philosophy of Sil. I may be off here, but it seems like it's pretty anti-grinding, and that's one of the things I really like about Nethack and Crawl. In these games, you have persistent, limited levels, and you're forced to make due with what you get.

Sadly, what turns me off about Angband in general is how its designed for you to have to grind a ton for specific resistances, items, etc. I can see the appeal for some people, but its on the completely opposite spectrum of something like Crawl.

I'll definitely give some of these suggestions a shot. I do get some fun out of PosChengband, but I also get the impression that it's another grind-heavy game. I love the varied loot system and character combos, though.

[Derakon]

Sadly, what turns me off about Angband in general is how its designed for you to have to grind a ton for specific resistances, items, etc. I can see the appeal for some people, but its on the completely opposite spectrum of something like Crawl.

I think Angband gets a lot of flak for allowing grinding, but it's also not necessary to anywhere near the extent that many people think it is. I mean, to beat Sauron and Morgoth, you need:

* Enough speed (+20 and a temporary source is fine)
* Enough damage (~350/round on a melee weapon, less on spells/archery since they're more reliable/safer)
* Enough hitpoints (600 minimum)
* Some healing items and a few panic buttons (scrolls of Destruction/Mass Genocide)

These requirements are not, usually, hard to meet. There's lots of other things that will make winning easier, sometimes significantly so. And there's some abilities that are more or less required to make it to Sauron and Morgoth (basic 4 resists, free action, see invisible). But the vast majority of abilities, including every single high resistance/status protection with the possible exception of poison (and Free Action, of course) are not actually necessary. And I'm not talking "well, the game has been beaten by ego-less characters, so none of this stuff is necessary", I mean "I routinely beat the game with major gaps in the character sheet". And I don't consider myself to be some godly player either; decent, but nothing amazing.

[ekolis]

And then there are players like me who suck so bad they DIE while grinding...

[half]

Yeah, I really like the barebones design philosophy of Sil. I may be off here, but it seems like it's pretty anti-grinding, and that's one of the things I really like about Nethack and Crawl.

Yes, that was part of the design of Sil. Scatha and I feel quite strongly that a game should be designed so that optimal play is fun play. Having a superior but boring option goes against this principle, and the worst is having a continuum of strategies between these as then you can't simply delineate the degenerate play and decide not to do it. Most people around here don't mind having degenerate gameplay options though, and some see it as a positive advantage to be able to choose and adjust the difficulty level during the game. I can see some sense in this, but not sufficient to tempt me to design (or play) games like that.

So I suppose Sil is an interesting example of a non-persistent dungeon game (i.e. a moria-like) with the arbitrarily slow descent removed, making it much more like a persistent dungeon game in some respects. I think the really interesting thing about it are the abilities. Some are just bonuses to rolls, but many change the basic tactical gameplay in ways that I hope many other roguelikes take up (such as Flanking, Charge, and Sprinting).

[Pete Mack]

Half, fortunately there are multiple ways to define optimal in angband.
Minimal likelihood of dying certainly means grinding.
But minimal player turns to win means something else.
Cliff Stamp showed NPP (small levels) can be one in as few as 20,000 player turns (< 100K game turns, at speed.)

So there's room for fun ~ optimal in angband, too.

[Patashu]

Half, fortunately there are multiple ways to define optimal in angband.
Minimal likelihood of dying certainly means grinding.
But minimal player turns to win means something else.

How many roguelike players do you think speedrun/aim for low turn count? I'd guess at least 90% are just focused on winning and so are highly risk averse.

[Estie]

The part about vanilla that I find the most "grindy" is that where I am about ready to fight Morgoth but there are still most evil uniques alive, and I rather want them dead before engaging M.

Getting the necessary items is challenging and fun, as the character is still weak whilst missing parts of the endgame set. Getting the level wanted for the endfight can be a bit of a grind too with classes that have weak sustained offense, mages rangers come to mind.

When all that is done, I feel compelled to kill at least some of the more unpleasant high end uniques. That in itself wouldnt be so bad except the motivation for it is "so they cant be summoned by M". I think I would enjoy the fights more if uniques were the major source of artifacts in the game rather than an unpleasant chore to complete before being eligible for the final fight.

So leave them alive and dont bother ? This works but I want a better rig for the fight with all uniques alive. Either way, the endgame is somewhat grindy if all you want is to do away with M.
Some of my characters I played past the point of being able to win, wanting to see what randarts were rolled. I would by no means call that a grind though, more of a relaxed chilling while having autopilot on.

Imho Sils "if you grind, I kill you" approach completely misses the point. Grinding too much in vanilla is punishment in itself; trying to "save" people from that fate by penalizing the grind artifically is like closing a shop rather than letting people queue up. The trick is to spice up the queuing time.

One of the main features of roguelikes for me is the metric of keypresses. While they would be called "turnbased" in gamers lingo, that isnt the whole story. Not each turn is the same; for example resting up requires 1 keypress if macrod, and assuming no interruption it can contain a few or many hundred turns.

The basic element that sets the rhythm (and there is a rhythm to angband, only the bot can play without that for hours on end) is not the game turn like in chess, but the keypressed action that can contain 1 or many game turns. Playing angband is going through a sequence of such keypresses. I have suspended disbelief when it comes to rationalizing the connection between the amount of keypresses and game turn duration long ago. I dont want to re-establish any such dependency again, neither by playing turn-minimizing vanilla games nor by playing Sil where you have to watch out for game turns when playing.

Obviously, others milage may (and does) vary; to each their own and have fun playing.

edit: minor wording correction

[BlueFish]

Sil can certainly be frustrating, but the short duration of the games is its saving grace. You never lose *that* much of an investment. Those short durations are a side-effect of the forced diving.

Sil would be a very easy game without forced diving - far easier than Vanilla, I think. If you removed that restriction, you'd have to re-balance all over again, to introduce some challenge. It would be a very different game.

[Pete Mack]

How many roguelike players do you think speedrun/aim for low turn count? I'd guess at least 90% are just focused on winning and so are highly risk averse.

But I was assuming a prior that the player wins regularly in the first place...
It's pretty clear that OP (stabwound) is in the other 10%.

[half]

Half, fortunately there are multiple ways to define optimal in angband.
Minimal likelihood of dying certainly means grinding.
But minimal player turns to win means something else.
Cliff Stamp showed NPP (small levels) can be one in as few as 20,000 player turns (< 100K game turns, at speed.)

So there's room for fun ~ optimal in angband, too.

I think this is true if you (i) are predicating on winning anyway, and (ii) turn off connected stairs. It is a great way for experienced people to play, though it is not the way the game presents itself to beginners (since connected stairs is on by default and players won't win for ages if ever). It is pretty good advice for the OP though.

[stabwound]

While we're on the subject of discussing various variants, I have another question. Are there any prominent variants that feature the "auto-explore" function of games like Crawl, Brogue and ToME4? I know some people are against such a feature, but it really makes it hard to play a roguelike that doesn't have it once you've gotten used to it.

Let's face it: Angband levels are huge and tend to be pretty tedious to explore. What I like about autoexplore is that it cuts down on a ton of keypresses and meaningless exploration and gets you to the next decision or encounter you should come across. It also makes it convenient to see something like a pile of coins, then hit the autoexplore key to run over and pick them up.

I feel like Angband in general could really benefit from such a feature. Of course, you're not going to use it much on levels that are dangerous, because autoexplore doesn't take any tactics into account; it just goes for the shortest routes. But it is still very convenient and takes away most of the tedious parts of games like Crawl.

[Derakon]

I'm not aware of any Angband variants that have autoexplore. The big argument against having it is that Angband can spring very nasty surprises on you that you really want to have detected in advance. Sometimes you'll be running along a corridor, hit a screen transition, stop to cast detect monsters, and decide you're really best off turning around and going the other way. But "screen transition" is an arbitrary distance these days when windows can be resized to just about anything.

In other words, it's much harder to define an "obvious decision point" in Angband than in many other roguelikes. There are monsters that you don't want to ever wake up -- the entire region of the dungeon around them effectively becomes a no-fly zone. And there are even more monsters that you just don't want to get into line of sight of (zephyr hounds, for example), but the only practical way to avoid them is to detect them and then avoid the area they.

The difficulty with auto-explore is, how do you decide when the player needs to detect again? Some players are more reckless than others. New players aren't going to understand when auto-explore gets them into trouble, and (with some justification) may claim that the game just decided to capriciously kill them.

Arguably this is a flaw in Angband's game design (that you can effectively lose the game before ever seeing the monster that kills you). However, it's a really deep-seated one, and digging it out would require completely changing the game to the extent that it wouldn't really be Angband any more. It's part and parcel of having a very high-stakes, "swingy" balance system where you can die with a single wrong action.

[NotMorgoth]

Arguably this is a flaw in Angband's game design (that you can effectively lose the game before ever seeing the monster that kills you). However, it's a really deep-seated one, and digging it out would require completely changing the game to the extent that it wouldn't really be Angband any more. It's part and parcel of having a very high-stakes, "swingy" balance system where you can die with a single wrong action.

Angband is not the only roguelike where you can die with a single wrong action - maybe the part of the game that actually causes this problem is the ease and availability of detection.

Crawl sidesteps the problem by not having detection or telepathy (even Ashenzari, the god of detection, only gives a general idea of the power level of the detected monsters.)

[Derakon]

Angband is not the only roguelike where you can die with a single wrong action - maybe the part of the game that actually causes this problem is the ease and availability of detection.

Crawl sidesteps the problem by not having detection or telepathy (even Ashenzari, the god of detection, only gives a general idea of the power level of the detected monsters.)

The issue with removing detection is that then you also have to remove all the monsters that will kill you if you let them get into LOS of you. I mean, if I'm walking down a corridor and my detection reveals that there's a pack of gravity hounds in that direction, I turn right around and go somewhere else. If I don't have that detection then I'd continue going, run into the hounds, and quite possibly die. The action that "killed" me wasn't when I actually got into LOS of the hounds, it's when I failed to detect them.

That's what I meant when I said that making auto-explore feasible in Angband would require significant rebalancing.

[Philip]

Yeah, you'd have to change the core of angband. Be prepared, be ready and be aware. You'd also have to remove a lot of monsters, since without either no insta-kill monsters(where instakill is to kill a fully able character without a chance to react) or something similar, many characters don't have a chance. No insta-kill monsters makes it hard to have status effects and burst damage. Few characters are worried about anything else.