Making the game harder, take two

[Philip]

I'd say before diminishing returns are instituted on monster experience, we should know where players are getting their experience from. I bet fizzix's stats generator could do this for us with reasonable accuracy by recording experience from the monsters killed. Obviously there'll be some inaccuracies because most characters don't kill every monster they meet, but I doubt they'll be statistically significant.

I suspect that the way this'll break down is that group monsters are responsible for the bulk of the player's experience gains in the 500'-2500' range, after which single big targets start to take over. However, each individual species of monster only provides a noticeable amount of experience in a very narrow band, after which the character isn't really killing them for the experience. That said, reducing the experience to zero basically means you're placing the monster in the dungeon just so it's in the way. It's not interesting to fight, and there's no rewards for killing it, so why is it here?

Drops have already been changed to max(average(dlvl, mlvl), mlvl), so no worries there.

I find it a bit odd that you pinpoint artifact rings. I don't scum for rings (generally don't scum at all, in fact), so your concern is basically a non-starter for me. Of course, enlightenment showing all objects on the level is something I make use of all the time, but not for ringarts; more for targeting my dungeon exploration. Is that so bad?

I got my experience with my latest winner from dragons I was immune to, later demon pits. Pots of exp also had a say in my experience. I liked dragons because of low risk good reward(I wanted items), later demons in ASC's because they summon and give exp and items. A pack of Barbazu's gave me two exp pots. I don't remember what I did before that. I think players will value drops more for exp pots and kill uniques more in risky situations. I like the second thing. My information is based on iVannila 3.1.2v2, so not new. Also, the scumming stuff is impossible in iVannila because clairvoyance only detects object in certain radius.

[Derakon]

I'd say before diminishing returns are instituted on monster experience, we should know where players are getting their experience from. I bet fizzix's stats generator could do this for us with reasonable accuracy by recording experience from the monsters killed. Obviously there'll be some inaccuracies because most characters don't kill every monster they meet, but I doubt they'll be statistically significant.

I suspect that the way this'll break down is that group monsters are responsible for the bulk of the player's experience gains in the 500'-2500' range, after which single big targets start to take over. However, each individual species of monster only provides a noticeable amount of experience in a very narrow band, after which the character isn't really killing them for the experience. That said, reducing the experience to zero basically means you're placing the monster in the dungeon just so it's in the way. It's not interesting to fight, and there's no rewards for killing it, so why is it here?

Drops have already been changed to max(average(dlvl, mlvl), mlvl), so no worries there.

I find it a bit odd that you pinpoint artifact rings. I don't scum for rings (generally don't scum at all, in fact), so your concern is basically a non-starter for me. Of course, enlightenment showing all objects on the level is something I make use of all the time, but not for ringarts; more for targeting my dungeon exploration. Is that so bad?

[Londorth]

One reason Angband is so easy (tongue in cheek you understand) is that risk is not proportional to reward. The first time a character meets an orc, it's a challenge. The next couple of times, probably too. But then he slaughters hundreds of them, levelling up etc. And the same is repeated thoughout - choose who to fight (Vrocks are a mid-game favourite for me) and avoid many of the foes.

Suggestion - scale experience by how many of the same monsters you have already killed. First orc - x. Second orc - 0.9x. Third orc - 0.81x. Fourth orc - 0.729x. That means that however many orcs you kill, you can only get 10x experience, forcing you to engage other opponents. Done right, it would also make the experience multipliers for the better races far greater penalties too - it might literally be impossible to get a high elf mage to 50th level.

(For the realists among us, I think this is good too - you really don't get as much experience in the real life sense from killing the same monster over and over as you would from seeking out new things to kill)

It's already been suggested that we need to return to drops being average(monster level, dungeon level) which I support. Also we need to make all cheap escapes remove items - keep teleport as a beam but make sure all treasure in the beam is removed (not teleported), banishment should remove all treasure under the creatures banished, mass banishment should remove all treasure in the affected area and destruction must destroy everything including artifacts.

(Of course you can still 'cheat' in a GV, tempt the monsters out of the vault or to cleared areas and then tele them, but that's a lot more challenging (and risky) than just MBing or destructing, and I think a good tactic to make the game interesting.)

The final area that to my mind needs fixed is the ease of finding artifact rings. Identify all normal rings, goto dlvl 98, wait for superb level, enlighten, etc. I probably support the idea that enlightenment should not show treasure, period. At least that way you have to go round the level enough to detect treasure.

Paul

[Timo Pietilä]

I agree that x2 brands are a good idea. I'll put it on my to-do list. We'll need different names, so I propose that we steal from sangband (though it only has fire and poison doubled):

fire brand x2, flame brand x3
frost brand x2, ice brand x3
poison brand x2, venom brand x3

I am inclined to simply reduce acid and elec brands to x2 only, and not have a x3 version.

I would be OK with weaker acid-brand (or even removed), but elec brand is the brand that makes most sense of the lot (ask yourself, what is the thing that can be hold safely in real life, and still can kill by touch to the wrong end?) and it used to be x5 brand. You would be going in opposite direction to my liking. Make it rare but powerful.

[Derakon]

Poor elec brand, down to x2 from its heady days as x5 way back when. How times have changed.

More seriously, why treat them differently? It's not like on-weapon acid/elec brands are called out as being particularly overpowered. Just call them acidic/corrosive and sparking/shocking.

[Magnate]

So let's bring this back to the front page.

Generally speaking, I see two ways to make Angband harder: reduce the margin for error, and make the tactical game harder to optimize. The margin for error is already pretty dang slim in the late game, but maybe we could step it up earlier by introducing some nastier monsters in the 1500'-2000' range.

As for hardening the tactical game, the way I'd see doing that is to identify current "default" strategies -- ones that practically everyone makes use of because they are clearly dominant -- and weakening them so that less dominant strategies become worth considering. One example of this is archery: previously archery massively out-damaged melee and spells, so it was a no-brainer. We've identified that, weakened the multipliers, and now it seems to still be worth using, but not as your sole source of damage. This is good.

Another one I've seen mentioned is branding rings. In particular, acid and electricity brands are useful since relatively few monsters resist those elements, but they don't show up often on weapons. Having the ability to brand any weapon with those elements, even at the cost of losing a ring slot, is pretty powerful; assuming you have, say, a 3d4 weapon, you get an increase of 15 damage on average against vulnerable monsters. That's pretty awesome, but many players have even stronger weapons in mind. A 4d5 weapon gains 24 damage on average, and an 8d4 gains 40!

What I'd suggest here isn't removing the off-weapon brands outright, since they're a neat concept. But they do need to be brought down to size. I suggest introducing "strong" and "weak" elemental brands at x3 and x2, with the branding rings only giving the latter. That should weaken the brands enough that they aren't an obvious play, while still making them useful in some situations.

I'm sure we were separated at birth. Last weekend I pushed to staging a refactor of slays and brands which enables adding additional brands with one line in src/object/list-slays.h (and an extra flag in src/list-object-flags.h).

There is some more work to do to remove the clunky defines OF_BRAND_MASK and friends, but we're nearly there.

I agree that x2 brands are a good idea. I'll put it on my to-do list. We'll need different names, so I propose that we steal from sangband (though it only has fire and poison doubled):

fire brand x2, flame brand x3
frost brand x2, ice brand x3
poison brand x2, venom brand x3

I am inclined to simply reduce acid and elec brands to x2 only, and not have a x3 version.

[Derakon]

So let's bring this back to the front page.

Generally speaking, I see two ways to make Angband harder: reduce the margin for error, and make the tactical game harder to optimize. The margin for error is already pretty dang slim in the late game, but maybe we could step it up earlier by introducing some nastier monsters in the 1500'-2000' range.

As for hardening the tactical game, the way I'd see doing that is to identify current "default" strategies -- ones that practically everyone makes use of because they are clearly dominant -- and weakening them so that less dominant strategies become worth considering. One example of this is archery: previously archery massively out-damaged melee and spells, so it was a no-brainer. We've identified that, weakened the multipliers, and now it seems to still be worth using, but not as your sole source of damage. This is good.

Another one I've seen mentioned is branding rings. In particular, acid and electricity brands are useful since relatively few monsters resist those elements, but they don't show up often on weapons. Having the ability to brand any weapon with those elements, even at the cost of losing a ring slot, is pretty powerful; assuming you have, say, a 3d4 weapon, you get an increase of 15 damage on average against vulnerable monsters. That's pretty awesome, but many players have even stronger weapons in mind. A 4d5 weapon gains 24 damage on average, and an 8d4 gains 40!

What I'd suggest here isn't removing the off-weapon brands outright, since they're a neat concept. But they do need to be brought down to size. I suggest introducing "strong" and "weak" elemental brands at x3 and x2, with the branding rings only giving the latter. That should weaken the brands enough that they aren't an obvious play, while still making them useful in some situations.

Also, patch Sauron's acid vulnerability.

What other dominant strategies are you guys aware of, that should be made less overpowering?

[fbas]

here's an idea to make the game instantly harder. have an alternate monsters.txt file with much higher monster hit points, combine with altered file for reduced probability of objects and artifacts, etc.

if you wanted to mod code for it, introduce a "difficulty" multiplier for the above scenarios.

[Adley]

"The critical hits system has been redone for most classes. While mages have the same ol' system as before, other classes can now get crits more frequently and with more power, depending on the weight of their weapons (the heavier the better). For warriors and paladins, this factor could now make large axes and such a more attractive proposition than the extra-blow generating whips and such that get used in Vanilla."

A solution?

[fizzix]

Vanilla combat needs to go. It is silly (from an immersion point of view) and counter-intuitive, and a better system exists (o-combat). It won't make V Oangband, Oangband has levelling perks, four spell schools, and other changes.

Fractional blows was recently added to V, and this makes significant changes to the acceptability of various weapons at various points. It might in fact satisfy most of your complaints. Give it a shot first and then see if you think additional changes are necessary.

[EpicMan]

Lord Fell: you are correct. Yes, V combat is broken. We all know this, and have done for decades. This is why Leon re-wrote combat entirely when he wrote O, and why "O-combat" is used in many variants, and is cited a lot as a "more realistic" combat model. This is all good - but there are no plans to introduce it to V, or otherwise move V combat in that direction. I'd love to bring it in, but then V wouldn't be V, it would be more like O ...

So we won't change something that we think is "broken"? It's been in V for decades, but so was the ID mechanic, and that was changed - in Vanilla first, wasn't it?

I think changing the combat system to something more intuitive would have a smaller effect on V than the ID changes have (which I point out merely to illustrate that V can change a fixed tradition without becoming not-Vanilla). This would make Angband more newbie-friendly in that their natural inclination to buy that big weapon is beneficial rather than detrimental. This would make things easier for new players without reducing real difficulty (after balancing the new system, of course).

Vanilla combat needs to go. It is silly (from an immersion point of view) and counter-intuitive, and a better system exists (o-combat). It won't make V Oangband, Oangband has levelling perks, four spell schools, and other changes.

[Lord Fell]

Lord Fell: you are correct. Yes, V combat is broken. We all know this, and have done for decades. This is why Leon re-wrote combat entirely when he wrote O, and why "O-combat" is used in many variants, and is cited a lot as a "more realistic" combat model. This is all good - but there are no plans to introduce it to V, or otherwise move V combat in that direction. I'd love to bring it in, but then V wouldn't be V, it would be more like O ...

This is why people's advice is to play variants if you want more realistic combat, ok? No need to go round the houses again.

I can definitely be sympathetic about keeping Vanilla separate and distinct from Variants. If someone adds wilderness regions, creates secondary dungeon areas, switches from a Class Base to Skill Based character system... this is all Variant, I don't want it in my vanilla.

On the other hand, I think that if a coder/maintainer of a Variant produced code that was an improvement on Vanilla, that Vanilla would adopt this code. Refusing to adopt superior code just because it originated in a Variant seems needlessly elitist to me.

I think this is more a case of the 2nd sort.

[Djabanete]

One possible addition to combat: allow certain weapons to be wielded two-handed. You gain combat bonuses (damage, hit, +1 blow, whatever seems appropriate) when you wield the weapon two-handed. You can't use a shield while wielding a weapon in both hands.

This is how Hengband does it. I like it a lot, because it balances out the inherent "light weapon bias" in the current model with the option of getting more punch out of a heavy weapon if you're willing to forgo the shield. Depending on what kinds of shields and weapons you have, sometimes there are interesting strategic choices to be made.

[Magnate]

A normal dagger is better than a longsword as long as your strength and dexterity are in the right range. Too weak/clumsy, and you don't get multiple blows, so you opt for a big weapon instead -- my priests often end up with flails or morningstars early on, for example. Too strong and you can get multiple blows even with the heavier weapons, so daggers again lose their shine. It's only in the very early game, for non-caster classes, that daggers are remotely automatically better.

But Lord Fell's point is that V's blows system, and in particular the way it fails to treat weapon weight properly, is fundamentally broken with respect to any sort of "fantasy realism".

Lord Fell: you are correct. Yes, V combat is broken. We all know this, and have done for decades. This is why Leon re-wrote combat entirely when he wrote O, and why "O-combat" is used in many variants, and is cited a lot as a "more realistic" combat model. This is all good - but there are no plans to introduce it to V, or otherwise move V combat in that direction. I'd love to bring it in, but then V wouldn't be V, it would be more like O ...

This is why people's advice is to play variants if you want more realistic combat, ok? No need to go round the houses again.

[Tiburon Silverflame]

It doesn't seem right to me that end-game mages can carry as much weight as warriors. Part of the early fun of mages is in having to keep your weight down and use the lightest armour/weapons you can find, but by the end you can use pretty much what you like.

Instead of capping carrying capacity at 18/50 you could cap it at 18/220 instead so only warriors can normally get to the full amount at the end.

What does class have to do with carrying capacity? If it's a question of "let's make heavier armor harder for mages"...is raw carrying capacity the right way to go about this, even assuming it's a desirable goal?

I wouldn't necessarily mind an extension to the capacity table...say, a fairly large plateau where things remain as is, up to around 18/150, then some increase up to 18/220. I don't think I'd like reducing the carrying capacity at around 18/50; sure, inventory management is an essential aspect to the game, but one has to balance things.