Any randomly generated large box in the dungeon may become named “Pandora’s Box”. It works like Schroedinger’s Box in that it’s a normal non-artifact box which is never trapped, and the special behavior when it’s opened is to spawn a bunch of nasty monsters around the player and subtract 2 Luck (and interrupt the normal box-opening process). However, there is something good for the player inside.

Eating a mimic corpse typically makes you mimic some object besides a pile of gold, giving you an opportunity to identify it if you don’t already know it. You get the message “You can’t resist the temptation to mimic a [object]” with that object’s randomized description being appended if it was previously unknown.

The odds of getting an object you haven’t discovered are controlled by your Luck.

• If Luck is 0, you mimic the first random object type the game picks.
• If Luck is positive, the game rerolls up to Luck times to try and find an unidentified object type.
• If Luck is negative, the game rerolls up to Luck times to try and find an identified object type.

Container traps’ effects scale based on dungeon level (ideally on the level a container generated, not on the current level, so a player can’t bring a chest up to level 1 to make it use the weakest possible trap.)

In addition to adding some more effects, also add some effects which cosmetically are the same as other ones, but are more or less damaging. This would, however, make untrapping containers even more useless than it is now; perhaps a trap should stay armed even after you set it off. Another idea is to give certain traps an arming item, like a potion of oil for the fire effect and a land mine for the explosion effect. Successfully untrapping it will deposit the arming item into the container.

Also, make container traps independent of Luck, because currently they encourage the player to ignore containers until Luck is maximized.

Killing a mindless pet does not incur any penalties to alignment record or Luck.

Sokoban puzzles that have no spare boulders by default can have a little attached room containing several easily retrieved spare boulders, but it incurs a Luck penalty to open the door to this room.

Luckblade’s enchantment always equals your current Luck. (Enchanting it or engraving with it would then presumably do nothing since these do not affect Luck.)

Leprechauns very rarely drop a four-leaf clover on death. This is a vegan comestible that provides 1d3 luck (or just acts as a luckstone) when carried and can be eaten to gain 1 Luck. (The four-leaf clover appearing from sacrifices is not this item and cannot be picked up, though that could become confusing.)

An artifact luckstone which increases or decreases Luck by 5 rather than the ordinary 3 for a luckstone when carried. (Luck might still be capped at -13 and +13 for technical reasons, rather than being allowed to go to +15 and -15.)

A cursed scroll of fire should create a pool of lava underneath your feet. Possibly only in certain circumstances like with negative Luck.

Carrying a substantial number of cursed items in your inventory has some denigrating effect on you. One suggestion is that enough cursed items counts as a negative luckitem similar to a cursed luckstone.

The quest leader either sometimes or always asks for the quest artifact back when you return with it, to replace the weird flavor in most quests that the leader is desperate to get the artifact back but is then completely fine with you adventuring off with it after retrieving it.

If you refuse to give them the artifact back, they turn hostile. If it’s a chance-based system, the chance might be dependent on your alignment record, so a character that has misbehaved is less likely to be found worthy of keeping the artifact. However, since you must have +20 alignment to enter the quest and get a massive amount of alignment from killing the nemesis, it’s questionable whether anyone’s alignment will ever be in a bad enough state to increase the chances. Luck has also been suggested as a variable that influences the chance, but most players will already have maxed their luck by the end of the quest.

In order to keep the player from just dodging the quest leader on their way out of the quest so that they can keep the artifact, various enforcement mechanisms have been proposed:

• Incentive-based; returning the artifact gives you +5 alignment record (and set to 0 if negative) and +3 to luck. Rewards should be better than the incentive of killing the quest leader to get the artifact back, or else penalties applied if you kill them (e.g. the artifact blasts you even though it’s your own quest artifact).
• Strict: you must show the artifact to the quest leader and have the quest marked completed in order for the invocation to work. Alternatively, the quest leader has the Bell of Opening and will only surrender it when you present the artifact, whether or not they demand it back. Or the quest nemesis has cursed the Bell and the leader will lift the curse after you show them the artifact.
• Delayed: you don’t have to go without the artifact forever, but the leader demands it back for some time so that they can fix whatever problem the artifact needed to solve. This could manifest as either a variable number of turns, or a requirement to go explore a certain number of new levels/area, before they’re done with it.

A new game ending where you can take an fake Amulet of Yendor into the Planes. This is only achievable by having your fake Amulet be unidentified, and have some combination of high Luck, being crowned, and high alignment record. The idea is that your god is so blinded by your excellence that they allow you into the Planes without checking that the Amulet is real.

In the Planes under this scenario, the Wizard’s harassment stops (if indeed he was harassing you at all; you don’t need to kill him or do the invocation to get into the Planes under these conditions).

When you get to Astral and offer the fake Amulet, you trigger a different ending in which your mistake is realized and you are condemned to become one of the player monsters roaming around on Astral until the hero with the real Amulet shows up.

You can chat to the Minetown Watch to pacify them if they’re angry, but it requires high Charisma and Luck to pull off (and is never guaranteed). There probably also needs to be some limiting factor so that you can’t retry this.

Non-chaotic monks should suffer a Luck penalty for eating meat, not just an alignment penalty.

Revamp luck so that it’s mostly permanent and tied to game milestones such as Oracle, Sokoban completion, clearing Medusa, passing the Castle, finishing the Quest, and getting crowned. All of these increase base luck by 1 (like the full moon benefit, only more so.) Carried luckitems adjust by -2 if they’re “net cursed”, +2 if net uncursed, and +4 if net blessed. All other things that affect luck still do, with no way to prevent them timing out back to the new base value.

Remove the air currents in Sokoban that prevent you from floating or flying over the pit and hole traps. Instead, apply the normal luck penalty. Though it’s been pointed out that this would allow a player to bypass the entire branch and then negate the whole luck penalty if they can find or create a coaligned unicorn and throw a few gems at it.

Fuzz the turn on which Luck times out by hashing ubirthday together with (turns/300), and taking the result modulo 300. This means that Luck will still time out once every 300 turns, but the player can’t predict the exact turn on which it happens, and figuring out the exact turn it happens in one iteration is unhelpful for any future iterations.

If Luck is timing out once per 600 turns, still do the above, but do nothing on every other iteration. Determining which set of “every other” iterations to use could also be done by hashing ubirthday and seeing if the result is odd or even.

A monster whose attacks scale in direct proportion to your Luck - the luckier you are, the more damage it deals.

Tourists begin the game with 1 or 2 Luck.

Non-excessive luck system: start by disentangling luck and religion, and convert rnl checks into other checks. Luck can only be -1, 0, or 1 and can never time out unless a full moon or Friday the 13th ends. Some things can reduce Luck to a base of -1, but nothing can increase it to a base of +1 other than carrying positive luckstones.

Worshipers of The Lady (neutral Tourists or some neutral Priests) have a permanent full-moon effect with Luck timing out to +1 due to the Lady’s influence.

Your god considers both your luck and your alignment record when deciding whether to grant a prayer, and weighs alignment record more heavily. That is, if you have negative luck but good alignment, your god will often still grant the prayer, and positive luck might save you from bad alignment, but not often.

You can wish for luck, which increases your Luck by d5 points.

When you are turning to slime, higher luck increases the odds of a container trap or magic trap being a tower of flame, and increases rather than decreases the odds of any tower of flame trap (container, magic and fire) hitting you.

Some, most, or all roles begin the game with high or maxed Luck. Of course, without a luckstone, this times out over a few thousand turns.

Dipping or dropping gold pieces in a fountain very rarely increases your Luck.

Rooms are generated dark much more often when your luck is negative.

One of the playable races, possibly halfling, should get a permanent +1 to Luck similar to the full moon effects.

Have a value that tracks the next turn Luck will check its timeout. Once the timeout ticks, it will re-randomize this value to some other turn.

Give a Luck penalty for destroying a Sokoban boulder by polymorph shudder (or polymorphing it in any way, really, since turning it into anything else removes it from the map).

Your success chance of writing an unknown scroll or spellbook should be based on Intelligence, not Luck.

Determining whether you escape or dodge trap effects should be based on stats like Dexterity and maybe role and XL, rather than Luck.

Luck ideas:

• Luckstones don’t remove luck timeout, they merely extend it.
• Luck is much less controllable, impossible to get any clues about except through enlightenment, and various random events (like what?) can give or remove luck. A skilled player would know how to bias the odds towards luck-increasing events. Controllable sources of gaining luck like sacrifice and throwing gems at unicorns are removed. It no longer times out.
• The luck timeout rate is less predictable, and a luckstone will help but it might still time out quickly if the RNG hates you.
• The base number that luck times out to is equal to the total number of luckstones in your inventory (so carrying 2 noncursed luckstones means luck times out to +2). Luckstones’ weight is increased.
• Remove luckstones in favor of an amulet of luck. Luck timeout can no longer be slotless.
• Luck times out at a rate of 1 point per (600*(number of luckitems + 1)) turns.
• Do away with luck timeout entirely in order to make it impossible to micromanage luck. Luck is now conferred entirely by luckitems. The problem with this approach is that NetHack has about two ways to gain timing-out luck and many, many ways to lose it.
• Luck can be gained by enchanting luckstones.
• Different types of luckitems which stack with other luckitems but not with each other, like a four-leaf clover you can get from leprechauns.

The Oracle stands on a neutral altar, but killing her sets your luck to -10.

For Luck balance, there should be some Luck-based effects that are really good at low Luck but get kind of uninteresting at high Luck.

Luck factors into the chance of successfully getting the enchantment you wished for on an object.

Luck’s effect on whether a projectile breaks is nerfed. Instead, the enchantment is the primary factor, and is more involved than just “+2 or above”.