Distinguishing dlevel 60 from dlevel 90

[fizzix]

One piece of advice you hear often on forums here is that players should never hang out around dlevel 60. Dlevel 90 is only marginally more difficult, but the loot is much better. This is good advice, but I think it's indicative of faulty dungeon engineering. So the question is how do you make dlevel 90 more deadly? First I'll give some past answers to this question, and then propose my own idea, to be met with howls of disapproval, as usually is the case.

Prev answer 1: The problem is that there are too many levels. compact the dungeon to 50 levels and redistribute monsters around. I dislike this solution because it takes away from the journey of Angband.

Prev answer 2: Enforce a minimum level for monsters. I.e. don't allow novices on dlevel 99. This doesn't work because there are so few non-unique monsters that occur deep. Furthermore, most of these monsters have DROP_GOOD so they cannot drop consumables. This would require a player to drop to a lower depth to scum for endgame consumables, which is unideal.

Prev answer 3: Increase monster density. While it's true that levels seem to sometimes be sparsely populated, another giant group of trolls isn't going to add any more difficulty. In fact it might make things easier, and certainly more tedious.

Prev answer 4: Add more deep, tough monsters. While this is probably a good idea, it's unlikely on its own to fix the problem. Especially if these monsters can be easily avoided.

My idea is to take the good parts of 2 and 3 and combine them, stealing the good parts of both. I would add additional difficult monsters in addition to what normally is there. The additional monsters would increase in both number and difficulty as the dungeon level gets deeper. The specific algorithm I thought up is as follows.

1. Take dlevel and subtract 50, then multiply by 2. Let this be N

2. Every time you make a room, roll d100 and if the value is less than N, you will add a monster.

3. If the roll passes, add a new monster in the room, constrained so that it has minimum level of dlevel/2.

4. Profit

So at dlevel 60, each room has a 20% chance of having a monster that is level 30 or deeper. At dlevel 90, each room has an 80% chance of having a monster that is level 45 or deeper. The hardest part will be to enforce the minimum level on the monster generation.

[Starhawk]

*howl of disapproval*

[EpicMan]

Answer #6: Replace those deep orc/troll pits on DL90 with new monsters that come in groups, don't have attacks/abilities that make them avoid-or-else, and that can drop consumables. The problem is that players might go back up to DL60 to scum the orc pits again.

Answer #7: Change DROP_GOOD and DROP_GREAT so that they can also drop good consumables (?TL, !Healing, stat potions, etc). By DL90 you could not care less about 99+% of equipment drops. But you will very much want stat potions (for restoring) and high-level healing and escape consumables.

To amend answer #1, perhaps we don't need to shelve 50 levels, maybe just the 30 between d60 and d90.

And really we're just removing 30-50 <b>depths</b>, not 30-50 levels. An ironman game has 100 levels, most games have many more, as players see at least two levels every time they recall, and many levels at depths 90-98 as players scrounge for consumables and/or go unique hunting. How much does the journey to Angband give us with 30+ depths that are best skipped?

[fizzix]

Answer #6: Replace those deep orc/troll pits on DL90 with new monsters that come in groups, don't have attacks/abilities that make them avoid-or-else, and that can drop consumables. The problem is that players might go back up to DL60 to scum the orc pits again.

Not allowing orc pits after a certain depth is reasonable. After all, giant and dragon pits are much better for consumables and gear anyway. No orc pits after dlevel 50, no jelly pits after dlevel 60 and no troll pits after dlevel 70 all seem reasonable. These changes don't really make those levels harder though, just more lucrative.

Answer #7: Change DROP_GOOD and DROP_GREAT so that they can also drop good consumables (?TL, !Healing, stat potions, etc). By DL90 you could not care less about 99+% of equipment drops. But you will very much want stat potions (for restoring) and high-level healing and escape consumables.

This was my very first suggestion on the Angband forums many years ago. People didn't really like it then, but maybe opinions have changed? Honestly, I'd like to remove the whole DROP_GOOD idea and instead replace it instead with higher probabilities for ego items. and DROP_GREAT should get an immediate roll for artifact, since only a few monsters have DROP_GREAT.

To amend answer #1, perhaps we don't need to shelve 50 levels, maybe just the 30 between d60 and d90.

And really we're just removing 30-50 <b>depths</b>, not 30-50 levels. An ironman game has 100 levels, most games have many more, as players see at least two levels every time they recall, and many levels at depths 90-98 as players scrounge for consumables and/or go unique hunting. How much does the journey to Angband give us with 30+ depths that are best skipped?

Yes, this is indeed the question. Which is more desirable from a playing perspective? Dropping to the ultimate depth and then slowly building up power there, gradually killing more and more difficult monsters, OR, making it so that the lower depth isn't survivable until you are reasonably strong, and requiring the adventurer to build up strength to not immediately die in the deep depths?

I'm probably biased in my opinions here. I want dlevels 60-90 to be a journey, I think the game is better if they are levels with their own differences. In my ideal game, the difficulty difference between dlevel 10 and dlevel 30 would be similar to the difference between 70 and 90. But that's me.

[dhegler]

Definitely love the idea of shortening from 100 levels and redistributing monsters. The first 50 levels or so are about 10x more exciting than the last 50. Also, there is a such a lack of new monsters after about level 80 that it just feels like a chore after you reach level 70 or so. Maybe it's just me, but I think 50-70 levels is all you need. Where did 100 levels come from anyway? Is that Tolkienian?

Also, what if you made a game option where you could specify x% "shorten dungeon depth", where you could say 50% and only get 50 levels, with all monster and object levels multiplied by the same factor (ie 50%)? Wouldn't solve a redistribution of monsters, but may be an interesting option.

[Derakon]

Orc pits already practically never show up past, oh, 3000' at the most? And troll pits start vanishing shortly thereafter. All that remains are giant/dragon/demon/undead pits and zoos.

DROP_GREAT already automatically gets an artifact roll. Any item that makes it to the ego level gets a chance to be an artifact, and DROP_GREAT ensures that the item is at least an ego item.

I had a change awhile back to contract the dungeon to only 10 dlvls. It was pretty gnarly; I didn't make it very far. It wouldn't be too hard to contract the dungeon by an arbitrary factor if you wanted to, though it'd be more difficult if you wanted the difference in difficulty between each level to be more or less constant.

[CunningGabe]

Not allowing orc pits after a certain depth is reasonable. After all, giant and dragon pits are much better for consumables and gear anyway. No orc pits after dlevel 50, no jelly pits after dlevel 60 and no troll pits after dlevel 70 all seem reasonable. These changes don't really make those levels harder though, just more lucrative.

This should more-or-less be the case already in 3.3. When I refactored the monster nest/pit stuff a while back, I changed the way that the type of pit is chosen. It used to be something crazy like roll 1d(DL), and if you get less than 20, it's an orc pit, etc. So 25% of pits at DL80 were orc pits. Now, orc pits almost never show up after DL55. (But they can happen very infrequently.)

Getting slightly more into the nitty-gritty: the commonness of any type of pit is normally distributed, with mean at the average level indicated in pit.txt, and with standard deviation 10DLs.

[Angelus]

Try playing without that silly preserve on. You'll stop a few times between 60 and 90, I guarantee it.

[qwerty]

Could we add flags to food and consumables to indicate level (like good great) then when a monster drops an item of that level you roll a dice to see if a consumable of the same level gets dropped as well. This might reduce the killing of helpless novices to get stat potions as you will be more likely to get them from higher level monsters?
Just a suggestion.
Cue howls of derision.

[d_m]

Could we add flags to food and consumables to indicate level (like good great) then when a monster drops an item of that level you roll a dice to see if a consumable of the same level gets dropped as well. This might reduce the killing of helpless novices to get stat potions as you will be more likely to get them from higher level monsters?
Just a suggestion.
Cue howls of derision.

This might work, but I think it will maybe cause other problems.

I know Gabe and I (and others) have talked about monster drop profiles. I think many variants have these and they seem like a nice way to solve this problem.

[Derakon]

I seem to recall seeing one of the devs saying they wanted to add OF_GOOD to the game, which if I read it correctly is "Object flag: good". I.e. you'd be able to modify object.txt, add GOOD to the flags list, and have that item be eligible for drops from monsters that only drop good items.

[jens]

This is indeed an important issue. It took me a couple of years to realise that the diving speed after the grind at stat gain, and a bit of grind for some good eq, should suddenly turn to a free fall...

On the other end of the scale some people want to skip the first 20 levels of the dungeon. Just as I feel on that issue, I feel here as well. I want Angband to be a long game, I want to use all the levels, I want all levels to be interesting, need different tactics, be relevant for @ development.

This is a big issue. Many issues that might at first seem unconnected to this has a big impact.

For me it starts with redesigning combat. The most important combat statistic is IMO BLOWS. At the moment we jump start beyond the halfway point on the possible scale of blows. This means that the dlvls needed for @ development become a lot more compressed than they need be.

The next issue I'd look at is the problem of avoiding the tough monsters. That is stealth and detection need to be nerfed/redesigned. I've been thinking about adding a whole lot of new tougher monsters for the empty levels, but knowing that people can just avoid them limits the possibilities (only ever vigilant and undetectable monsters apply).

[Derakon]

So you don't think that monsters should be avoidable? Some of the most fun I've had in Angband is trying to dodge around the 80% of monsters that could kill me trivially so I can get to the 20% that I can actually handle.

As for combat, I'm generally fine with the idea of redesigning it, so long as we don't introduce more grinding. Thus I tend to disfavor approaches that create level-based breakpoints or sharp cliffs in the dungeon difficulty curve. The player should not feel like they have to stop and mindlessly kill monsters for awhile before they can get on with progressing through the dungeon.

[Pete Mack]

So you don't think that monsters should be avoidable? Some of the most fun I've had in Angband is trying to dodge around the 80% of monsters that could kill me trivially so I can get to the 20% that I can actually handle.

+1.
Angband is just a grind if there aren't lots of monsters you can't reasonably handle. (Pit Fiend in GV at dl 60 comes to mind...)

[jens]

So you don't think that monsters should be avoidable? Some of the most fun I've had in Angband is trying to dodge around the 80% of monsters that could kill me trivially so I can get to the 20% that I can actually handle.

This is one of the interesting things about designing an open source game, every time you have an idea, someone will feel that the opposite is better ;-) But after some talk some improvement is usually possible for both :-)

OK, so people need to be able to sneak around. But they do not need to do it on every level all the time, i.e. there is some room for design that would accomodate both of our views (and btw, I am not opposed to sneaking, I just feel that it should be a choice, not the default as it is today).

As for combat, I'm generally fine with the idea of redesigning it, so long as we don't introduce more grinding. Thus I tend to disfavor approaches that create level-based breakpoints or sharp cliffs in the dungeon difficulty curve. The player should not feel like they have to stop and mindlessly kill monsters for awhile before they can get on with progressing through the dungeon.

Of course I mean a redesign that takes everthing into account and produces a result that everyone will always be happy with ;-)