Making the game harder

[Tiburon Silverflame]

Removing, or greatly limiting the availability, of object detection might be too extreme, tho. If object detection is made much more vague, it's possible that will be enough.

Or perhaps go with a middle ground. Remove the rod, scroll, and spell of treasure detection (or make the spell rogue-only). For !Enlightenment, !*Enlight, and Clairvoyance, fuzz things as I suggested.

Hmm.

To go with Derakon's comment, how about this...when objects (but not money) are created on the floor, there's an immediate 50% chance to create a monster right there. That way, you don't have free run to most things; much more of the time, you'll at least have to fight *something.* So, creating a level might proceed through the following steps:

a) Create vaults/pits
b) create rest of level
c) add objects, including 'guardians'
d) add monsters

So levels with lots on the ground, will also tend to be levels with lots of monsters.

[Hariolor]

I do like Timo's suggestion that detection/ESP not reveal the exact monster type. I suggest that everything you "see" that's not in LOS be rendered in grayscale and described with generic terms when you 'l'ook at it. The suggestion that ESP simply detect the existence of creatures instead of what type they are is also a good one. We could bring back the old & symbol to represent unknown creatures that aren't in LOS.

As for revealing items in vaults on entering the level: I'm opposed to this largely because it makes the risk/reward calculation for cracking the vault too easy. This is also why I think the abundance of object detection in modern Vanilla is not such a great thing. Being able to easily see what's on the ground makes it too easy to know that you don't need to go anywhere near that big pile of monsters.

I think it would be fun if creature types that have been "probed" (eew) become fully identifiable with esp - but don't have this carry forward from character to character. This would make probing more useful than it currently is, and would add another strategic element to the game (the value of probing things vs the risk inherent of wasting a turn while in LOS).

[Magnate]

Where?

Run 3.1.1 or 3.1.2 with the -r command line switch and it will deposit a new monster.txt file in your lib/user dir. The monsters stay the same, but the depths and xp rewards are changed for better balance. It makes for an interesting game - but hounds appear too early for me!

[Magnate]

Just a suggestion: While there are lots of good ideas here, I am wondering if the maintainers have come up with a goal of what they want Angband to be, and then filter through the suggesions that accomplish that goal. Should Angband be longer or shorter, easier or harder, better ui, more race/class variety, simpler or more complicated? I know there is an active maintainer group, but have they agreed on 2-3 goals of what all the changes in Angband are working towards?

There have been alot of great changes in the last couple years, BTW, both in gameplay and in the code organization, which is why I will be spenging practically every minute of free time I have in the next couple months getting NPP caught up with the current vanilla source. The only thing I don't see is an overall theme of where it is all going. Gameplay seems to be headed in a couple different directions at once.

That's precisely because there hasn't been a conversation about overall goals since Takk wrote http://trac.rephial.org/roadmap, which is at least two years ago (before I had commit access). There was more recently a discussion on angband-dev about 3.1.3 which I summarised in a previous thread about where the game was going, but that was more about specific task goals rather than overall game direction. You'd have to ask Takk what his overall goal(s) are but I believe they were focused on tidying up the code, making collaborative development easier, making the game more accessible to new players and available on a wider variety of (non-obsolete) platforms. I don't know his views on easy vs. hard, though he has acknowledged that there are now too many ego/artifact wieldables and too few consumables dropped - both of which will be further adjusted for 3.1.3

I think he also holds the view that those who currently happen to be exercising their commit access don't have any greater say over where the game should go than the rest of the community - so he's just as likely to be persuaded by debates here as on angband-dev or anywhere else. That said, changes are more likely to be coded by those who want them than by those who don't - that much influence is unavoidable!

Personally I don't think the game is too easy at all - I've been playing for ten years and have only once got beyond 4000', and I don't think I'm exceptionally inept.

I am thinking about Timo's suggestions of making less info available - Sangband has a nice implementation of partial ESP, where a monster out of LOS is represented by a flickering asterisk, which may or may not be quite in the right place ...

[Jungle_Boy]

I remember in Rogue there was a monster, Q that would immediately run to and sit on any gold near it so you'd have to kill it to get the money. Having some monsters do something like this with objects could be interesting. Although you'd have to do something to prevent @ from standing off and peppering them with arrows.

[PowerDiver]

I am thinking about Timo's suggestions of making less info available

I have very mixed feelings about it. On the one hand, fog of war is exciting. But I really hate the idea of taking notes on how many blows hit a wyrm so I can keep track of its health. I also hate the idea that I fight something, teleport away, teleport back and the situation is treated as fog of war again even though I know what it is.

[ChodTheWacko]

Perhaps the easiest compromise to code would be:
1) Detect object only shows you objects when the spell is cast (like detect monster)
2) Detect object only shows that am item exists there.
It doesn't tell you if it's a potion/weapon/whatever.
Maybe all items are displayed as a * or something.

It would also be interesting to add some 'fake objects' that detect object incorrectly detects. Detect object doesn't need 100% accuracy.

So at least you don't know there is a potion of strength sitting there undefended.

Nerfing detect monster will making diving signifigantly more dangerous.
Perhaps that's the point, but I don't know if we want to increase the chance of instakill.

It would also be fun to add more damaging potions that you can get burned on by id-by-use. Not so bad that you will stop using id-by-use (hello curse weapon), but something like say, 'drop a stat by a point'. Something painful to set you back, but not game breaking.

- Frank

[Derakon]

We used to have stat-draining potions. They meant that nobody bothered using ID-by-use on potions once they passed 200'. Now nobody bothers using ID-by-use on potions after, oh, 1000' instead. This is a distinct improvement. We don't want to make ID any more painful than it is already.

I'm also not a fan of unreliable detection (monsters with WEIRD_MIND annoy me; I'd rather they just not be detectable by ESP at all). Spell unreliability is encoded into the failure rate of casting the spell (or activating the rod, etc.). Once the spell goes off, it should have a reliable result.

In general it sounds like there's a fair amount of support for making exploration of the dungeon into actual exploration as opposed to simply traversing known areas that haven't been cleared yet. The more knowledge the player has, the more prepared they are, and the less likely they are to get into a situation they're not comfortable with. But getting into those situations is much of the fun -- not knowing if you're going to survive (but having a chance to, if you don't make mistakes) is good design.

[nullfame]

1) Detect object only shows you objects when the spell is cast (like detect monster)

I am in the opposite camp. I agree with PowerDiver's sentiment which, I believe, is if I can take notes to gain an advantage the game should just do it for me. So if objects are displayed for only 1 turn I can take a screenshot (since they don't move) and refer back to that. That's broken.

How about you can only remember areas you explored for 500 turns, after that you forget those portions of the map? That would be harder but no fun.

If anything I think there should be a "recall last detect monster placement" feature. I hate wasting another charge on my rod of detection because I don't remember if Q is in the moat or in the room. Of course that belongs in the "make the game easier" thread

2) Detect object only shows that am item exists there.
It doesn't tell you if it's a potion/weapon/whatever.
Maybe all items are displayed as a * or something.

At first I liked this idea and even thought of proposing it but a recent game experience has lead me to be against it. I came across a vault with Time hounds, Pazuzu, and the Mouth when I had no chance against any of them and poor terrain to boot (long hallway leading to it, sleeping ogres between me and baddies, them with reverse LOS letting them cast at will once I opened it). If I saw a bunch of * for objects I probably would have destructed it and looted artifacts (which is why I think destruct should remove, not destroy, unidentified artifacts).

Instead I saw a ring of speed which was what I was looking for. So I cracked it and spent a long, tense time getting in there... killing the hounds, receiving lightning balls, summons, time breaths, phasing away, being chased... you know, having fun.

Choosing your battles, the chances you take, and where you explore is about risk/reward ratios. Transparent values of both sides of the quotient helps the player make smarter decisions and have more fun. If you hide reward I substitute a constant value and take fewer risks, and almost never chance it when risk is high.

[Pete Mack]

There have been alot of great changes in the last couple years, BTW, both in gameplay and in the code organization, which is why I will be spenging practically every minute of free time I have in the next couple months getting NPP caught up with the current vanilla source. The only thing I don't see is an overall theme of where it is all going. Gameplay seems to be headed in a couple different directions at once.

This is exactly right, though I would argue that the correct order is
1. Fix the code.
2. Add (or repair) features

It is my fondest hope that 1 is complete and we can focus on 2.
The extremely rough targets I would hope for here are
1. Reduce tedium
2. Maintain a high-risk environment.

Some conclusions frome thes
* consumables (potions, ammo, and scrolls) should be sufficiently available to win comfortably in well under 1M turns. (This was definitely the case in 3.0.9 and earlier.)
* ID-by-use should not be significantly penalized compared to ID-by-spell-or-scroll.
* Some (many) monsters should be hard to kill when you first meet them.
This is the too many ego-and-artifact problem
* There should be more and more dangerous out-of-depth monsters.
If we are going to reduce total average turncount (see reduce tedium), we need to maintain the net risk per game. IMO the best way to do this is a limited nightmare mode, where no level can be assumed to be truly safe.

[Timo Pietilä]

If we are going to reduce total average turncount (see reduce tedium), we need to maintain the net risk per game. IMO the best way to do this is a limited nightmare mode, where no level can be assumed to be truly safe.

Are you talking about creating "difficulty levels"? I vote against that.

I don't think "reduce tedium" and "reduce turncount" are even related. You could have fun with char that does multi-million turn wins if the main dungeon is interesting. Exploring can be fun. Cracking vaults can be challenging. Knowing everything the level can offer after one turn (enter level, cast clairvoyance/activate Palantir, "[") is boring.

Problem with "tedium" is that there is too much to do before final fight that is just matter of patience: Consumables to collect, uniques with nasty escorts to kill, gear to find. We should have some faster, but more risky, way of doing all that. Especially that consumables collection.

There should be less "blend into general mass" items. We should have items that are "unbalanced" but very rare in comparison to others. Items that make last fight laughably easy *if* you find them. Artifact consumables. Potions of invulnerability. Making every single item usable at any situation makes them all boring. Availability of many items in shops should be reduced for many things. Like arrows, bolts and shots. Versatility, not equality. Make items and things good at one thing and and crappy at others. Make it difficult to find right combination. Force them to make hard choices.

[ewert]

There should be less "blend into general mass" items. We should have items that are "unbalanced" but very rare in comparison to others. Items that make last fight laughably easy *if* you find them.

I disagree with this sentiment. Things that are superrare and superpowerful are tempting, but for general gameplay they are much less rewarding than rare and powerful items, IMHO.

Bladeturner. Haven't found one, ever. Heck, goes for PDSM too. =P Next game I will try dlvl 100 farming (just destruct M away every time you want to clear the level), is it easier to find +10 OoD items instead of +12? =P And I really really play a lot of time at dlvl98 just killing pits of U and D and clearing vaults.

So, we already have The One and Bladeturner as uberrare items, I definitely don't think we need more.

If you meant there should be less "Yet Another Westernesse Weapon" and more artifacts or TopThree Gondolin/Fury weapons (MoD, SoS, BoC), but less weapons found in general, then I'll agree.

[Tiburon Silverflame]

If we are going to reduce total average turncount (see reduce tedium), we need to maintain the net risk per game. IMO the best way to do this is a limited nightmare mode, where no level can be assumed to be truly safe.

How much of an impact will this really have on turncounts *per level*? If the net savings per level is only, say, 50 or so time ticks, the risk reduction probably can't be measured.

[nppangband]

This is exactly right, though I would argue that the correct order is
1. Fix the code.
2. Add (or repair) features

It is my fondest hope that 1 is complete and we can focus on 2.
The extremely rough targets I would hope for here are
1. Reduce tedium
2. Maintain a high-risk environment.

I like those first two goals.

I think your targets are great! That is what we have been trying to do all along in NPP with things like quests, more pits/nests, themed levels, varies terrain and in general more opportunities and options to take risks with a corresponding reward if you succeed. People who want to play conservative can do so, but people who want want more danger have plenty of ways to find some. I know some of the ideas still need a little balancing, but most of them work pretty well.

I also thing a hard decision needs to be made about the number of levels in Angband. There are barely any new monsters after 3500'. I know 100 is a nice, round number, but the monster and object list seems to be set up for about 80 levels. Either that, or we need around 100 or so higher level monsters for levels 3500'-3800'

[fizzix]

I like those first two goals.

I also thing a hard decision needs to be made about the number of levels in Angband. There are barely any new monsters after 3500'. I know 100 is a nice, round number, but the monster and object list seems to be set up for about 80 levels. Either that, or we need around 100 or so higher level monsters for levels 3500'-3800'

While there aren't really any new monsters after dlevel 80, there are quite a few uniques. I fear that adding new higher level monsters would just make more monsters to avoid, and make summons that much more annoying and impossible to deal with.

Instead of adding new monsters, maybe we should change the monster density (or enforce minimum levels) in these last levels. Disable deep descent, create stairs down and TL down (unless you can recall to deeper), and only have one down stairs per level. That would make descending the last 20 levels harder without adding new monsters. Each level would be like a 'quest' level, except the quest is 'find the stairs'.