Making the game harder, take two

[Magnate]

This started as a post in the "junk" thread but deserves a new thread as it's not about junk. It started off as being a reply to d_m about the ID proposals for 3.3:

So are we doing both instant strong pseudo at clev 20 (as per ticket #1094), *and* instant-ID at clev 40 (as per ewert)? For all classes?

That's a massive change. I'm not saying I'm against it, but we shouldn't underestimate how much it will "make the game easier" by effectively removing the ID minigame. It is possible to reach cl20 in minutes of play (I play slowly and even I can do it in under an hour), and instant pseudo will automatically reveal any cursed items (unless we introduce Sangband-style "hidden curses", which people will hate). Given the sophistication and speed of ID-by-use now there's not much difference in impact between the pseudo at cl20 and the full ID at cl40.

Why do I describe it as making the game easier? I think it essentially equates to giving all chars an extra inv slot.

I don't mind trying it, but I think we are going to have to think quite hard about how to make the game harder again, especially the mid-game. Fewer good/great drops? More monster hp or speed? Giving a whole bunch of monsters +3-5 speed might help. Undetectable traps?

Angband has always been a game you could win if you were slow and careful enough (c.f. Moria with AMHD offscreen instadeaths). The "difficulty" in Angband was simply that it severely punished risk-taking. The last two years of development have really toned down that punishment a lot, and I'm not sure that simply cranking it back up is what we want.

But I am sure that most people don't want instadeaths either. There was an interesting debate about OOD monsters a while back, when we increased the range of OOD monsters that could be generated on a level. I remember Timo saying that the occasional instadeath, to a massively OOD monster, is actually part of the excitement of the game. But a lot of people don't agree with that, and think that instadeaths are inherently wrong.

NPP is often held up as a positive example. It seems to have retained the "balance and danger of the old Angband", while delivering a better UI, better information to the player and more interesting features. I think a lot of its success is based on more severe punishments for diving/risk-taking: monsters are tougher, and you really can't fight hydra pits if you dive too fast etc. Nor can you do as much sneaking around out of your depth as you can in V, because of monsters with TELE_SELF_TO etc.

Personally I think it's probably a good thing if V is "easier" than most variants - much better than the other way round. But I'm just thinking aloud about the tension between making the game more challenging and making it a grind-fest where we all play like Neo and win every game in 5M turns.

[Nick]

Just on the ID topic, I think any radical changes should wait until changes to curses have been made.

[Estie]

I like the idea but I think levels 20/30 is too low. There is a point where "oh, a leather armor, lets id and see if it is good!" turns to "should I id this leather armor ? its unlikely to be an upgrade." That point is past level 35. Maybe it depends on playstyle, but I´d start with auto-id at level 40 and see where it leads.

Making the game harder ? Increase monster damage.

[LostTemplar]

Monster damage will not help, monsters can be avoided, and gear can be obtained without fighting and without risk. Imho to make game harder just remove destruction and teleport other, or seriously remake monster abilities and AI, to make monsters more likely to attack first. Something like stronger and faster never sleeping ghosts, etc.

[ewert]

iVanilla has:

Near instant pseudo-id at clvl 30 for warriors, bit higher for others, and IIRC mages never get instant pseudo-id before instant-id at clvl40 anyways.

I agree, clvl 20 is way too early for instant pseudo. By that time the pseudo should be reliable (meaning you lug the item around for a while), not instant. However by clvl30 id is no longer a problem, and all that id stuff is just annoyance and reduces game fun. So really fast pseudo (or even instant) about there, and by clvl40 you are so freaking annoyed by id it does nothing but annoy, it has no gameplay effect anymore.

I've played a couple of games (1 winner, 1 near-so) with current pseudo/auto-id in iVanilla, and they feel okay. Id is part of the game till it no longer matters as a money cost, then poof it disappears.

[ewert]

Oh and to make the point even stronger about id being just an annoyance very soon in the game, I may often just "waste" id charges/scrolls by midgame, if pseudo doesn't kick in soon enough. Because the charges/scrolls don't mean anything anymore, there is no reason not to just waste one on a heavy piece of gear if pseudo didn't go off yet ...

And frankly, going back to dungeon, I'm never at a loss of inventory spots. Coming up, I just drop the rest of the scrolls if I need the slot, or drop the staff. Hence, after a while in the game id and the mechanics of id provide zilch, nada, nothing of playgame value. I also play pretty much always under burden limit, so not even that, nope.

[Magnate]

And frankly, going back to dungeon, I'm never at a loss of inventory spots. Coming up, I just drop the rest of the scrolls if I need the slot, or drop the staff. Hence, after a while in the game id and the mechanics of id provide zilch, nada, nothing of playgame value. I also play pretty much always under burden limit, so not even that, nope.

Well I think you are atypical - a lot of people deal very often with both burden issues and lack of inv slots.

I agree with Nick that we should do some creative thinking about curses before making changes to ID.

I also don't really like such a huge breakpoint feature as instant ID at X clev. I'd prefer to see things accelerate gradually by clev - so for instance pseudo accelerates until it's instant at (say) cl35, at which point you auto-ID things after a certain time. That time then reduces until ID is instant some levels later.

Removing *destruct* is a possibility, but IMO removing -tOther isn't ... but maybe that's just my subjective view based on lots of experiences where tO was the only solution. It has been suggested before that some monsters could resist tO, or monsters could get a save against it. We could also give monsters a chance to resist *destruction* too ...

[buzzkill]

I remember Timo saying that the occasional instadeath, to a massively OOD monster, is actually part of the excitement of the game. But a lot of people don't agree with that, and think that instadeaths are inherently wrong.

Instadeaths are an inherently part of Angband and most any RL. Removing or minimizing them just seems wrong.

Removing *destruct* is a possibility, but IMO removing -tOther isn't ... but maybe that's just my subjective view based on lots of experiences where tO was the only solution. It has been suggested before that some monsters could resist tO, or monsters could get a save against it. We could also give monsters a chance to resist *destruction* too ...

TO is a luxury, cheap and exploitable. To call it the only solution to any problem in a game where escapes are so easy and plentiful is ill-conceived at best.

My 2 cents (once again) is that Vanilla is straying from it's roots, in danger of becoming a variant, a variant of itself. The Fangbnad comp has reminded me of what Vanilla should be. Slap a fancy UI (and tile support) on Fang and I'll play it, but I've got no real desire to play V (and rarely do), and that bothers me a little, because I don't think that the problem is with me.

It seems that there is a lot of change happening just for the sake of change. Things are being added and removed constantly (in the dev), at times it would seem without very much thought about long term goals. I see it as a 'let's throw it up against the wall and see what sticks' approach. Vanilla should have better direction, a more defined path that it follows. I fear Eddie's (?) arms race metaphor is permeating not only the the weapons and armours, but the whole of the game. Changes are made, then further changes are made in order to accommodate the first changes...

[ewert]

Well I think you are atypical - a lot of people deal very often with both burden issues and lack of inv slots.

Probably, I just never manage to have so much really useful stuff before ID is meaningless. And when coming up slot limited (cleared a vault or something), burden/slot from ID source is meaningless as you just dump the id item (and often spellbooks, phase doors, et cetera) ...

I also don't really like such a huge breakpoint feature as instant ID at X clev. I'd prefer to see things accelerate gradually by clev - so for instance pseudo accelerates until it's instant at (say) cl35, at which point you auto-ID things after a certain time. That time then reduces until ID is instant some levels later.

Hmmh that's a good idea in essence, I guess could do it within current pseudo-id code. However, the current breakpoint is clvl 40 for instant id in iVanilla. That is pretty darn high level. You'd actually only make id easier if I make that change (delayed full ID after instant pseudo id) compared to what iVanilla has now. Overall pretty minimal gaming experience change, because if you are burden limited, you want to id stuff right away so you can chuck away bad ones to get +1 +2 speed back, so the delayed full id could just as well not be there (only warrior does not have class given ID at that point, only because he never gets one).

So in essence, yes you are right. In game mechanics and gameplay experience, the time (classlevel-wise) between delayed ID and instant ID is for many irrelevant, as other mechanics (burden, slots) get precedence in requiring instant ID to maximize performance.

Removing *destruct* is a possibility, but IMO removing -tOther isn't ... but maybe that's just my subjective view based on lots of experiences where tO was the only solution. It has been suggested before that some monsters could resist tO, or monsters could get a save against it. We could also give monsters a chance to resist *destruction* too ...

I vote for giving uniques resist chance to *destruct*. tO should remain infallible IMHO (well, if you don't fail it).

[bulian]

Both TO and destruction remove immediate threats with nearly the same effectiveness, as most monsters when TOd do not make it back to the player. What if the range of TO was inversely proportional to the monster level (or size)? The player would still get a break but there would be a higher probability that higher level monsters would find a path back to the player.

IMO instadeaths are OK. If you can win every time with proper play/risk minimization, its not really a game anymore.

[dos350]

ee

eek hi i have been playing angband for not long at all and i have never won but i think that this is serious changes ----~~~~~ please to not do it!

destruct is must
auto id is not need

please to lmk

[Estie]

[QUOTE=buzzkill;43860]Instadeaths are an inherently part of Angband and most any RL. Removing or minimizing them just seems wrong.



TO is a luxury, cheap and exploitable. <>


How do you exploit TO ?

[Bandobras]

Instadeaths are an inherently part of Angband and most any RL. Removing or minimizing them just seems wrong.

I hope this is a misunderstanding. I think you mean instant death as a consequence of taking a significant risk, while others mean random instant deaths, not related to taking risks or avoiding risks. An extreme manifestation of randomness in a game. Agreed, Dungeon Crawl has these and some more games, but I don't think it's universal for roguelikes and unintentional in some that have it.

but I've got no real desire to play V (and rarely do), and that bothers me a little, because I don't think that the problem is with me.

Just to clarify: did you have that desire before V started being tampered with by the current maintainers?

[fizzix]

Some thoughts on the OP.

instant ID at clevel 40 is highly recommended. I would even propose an ID on walkover or ID everything in LOS. But not before clevel 40. instant pseudo at clevel 20 is a bad idea. ewert's pseudo seemed to work pretty well.

As far as making the game harder:

I've long been a proponent of reducing destruction's effectiveness. Mainly by allowing monsters to save and if they do, teleporting them (as in TO) or moving them directly outside the destruction zone. This keeps destruction as an escape, a turn buying maneuver. But it doesn't make the level completely safe anymore. Destruction should also remove artifacts. I hate that it doesn't.

If TO changes, summoning needs to change also. Summon hi_dragon is common enough that a player needs a common counter. Right now that's TO. Reducing TO is possible. But for that to be done Tele-self needs to be a lot safer or summoning needs to be weakened. Making tele-self safer actually means that hound packs need to be about 1/3 the size that they currently are. The game then turns into teleport away and lure monsters away, which may be more fun than teleport monsters away and loot the place. I do know that the game has more interesting decisions before coming across -TO and _destruct (change done by reducing monster pack size to 1/3 what it currently is, maybe Z specific, removing -TO rods and making wands as rare as annihilation. adding -teleport and giving it an easy activation.)

DaJ allowed monsters to blink even if they were out of LoS. I like this also.

I also agree with Timo in that there needs to be greater variation on dungeon levels. Monsters and items in certain rooms (like the center of cross rooms) should be further out of depth than they currently are. There already is some of this in V since my medium vault changes are there.

Lastly, I don't understand buzzkill's complaint, or rather, I think I understand it but I don't think of it as bad. Angband development seems to me a lot closer to say, evolution, than to intelligent design. I don't think this is a bad thing. Moving towards a specific goal is impossible when no one can agree on what that goal should be. If anything threads like this are focused on creating a clearer goal. Namely, making the game more difficult.

[Nomad]

As somebody who tends to descend quite steadily, exploring most levels and taking on almost everything that's not going to result in insta-death, I actually find the difficulty level quite well-balanced. I think the issue is lack of association between risk and reward: it's too easy for experienced divers to collect high-quality gear without being forced to take on challenging battles to get it. "Evade anything tough until you've got the perfect kit to make killing it easy" is just as rewarding a play style as slogging it out in epic battles with monsters you can only just handle.

So, two suggestions I would make would be:

1. Restrict artefact generation so they're only ever found in vaults or dropped by uniques, never just lying on the floor or dropped by random monsters.

2. Get rid of wands and spells of teleport other, but leave the rods. That way, you still have it as a tool, but you're restricted to a limited number of shots at a time so you can't wade in and clear an entire vault with it.