Additional gold sink

[Carnivean]

This is not a valid problem statement, AFAICT. What does the accumulation of gold achieve in the game?

Currently, it's a shortcut to a subset of desirable items.

Perhaps it should have been written with "can" or "does" instead of "has to". If a player wants money they have to find it, but it's optional (someone, probably Magic Mushroom, squelches money). The game can be played without money without losing anything.

[Monkey Face]

As far as I can recall (it's been about 30 years) but in the original Rogue, your score for leaderboard purposes was based on your gold accumulation rather than your experience points.

Maybe we should go back to that. Then it becomes a trade off whether to use your gold to buy those BOS in the BM or hoard it some more for a higher score.

[Cold_Heart]

I really like the idea of having an ability to order stuff in shop: pay shopkeeper to try to procure certain item for you using their trade connections, while you are down in the dungeon. The more you give him the better the item will be.

Of course it can end up a mace of slay troll but hey, it could be the one ring

[Derakon]

As far as I can recall (it's been about 30 years) but in the original Rogue, your score for leaderboard purposes was based on your gold accumulation rather than your experience points.

Maybe we should go back to that. Then it becomes a trade off whether to use your gold to buy those BOS in the BM or hoard it some more for a higher score.

This only makes a lick of sense in games with limits on how much content is available. Even in forced-descent in Angband, the fact that monsters can spawn back in mid-level (combined with the total lack of any kind of "clock" like limits on food or light) means that the highest score would go to the player with the most tolerance for tedium.

And if you removed the respawning and stick with forced-descent, then you end up with the highest scorer being the player who mines out the most treasure and gets luckiest with cash drop sizes...let alone getting lucky with, say, hydra pits.

[mushroom patch]

Though in fact, score is determined primarily by XP for winners so the situation you describe remains the case, see e.g.:

http://angband.oook.cz/ladder-show.php?id=10569

Topping the ladder at 64 MILLION turns o.0; In another roguelike, this is past the end of time.

It would be great to reform score so that it is a meaningful measure of something. For this, dividing by turncount seems to work well.

[Derakon]

Though in fact, score is determined primarily by XP for winners so the situation you describe remains the case, see e.g.:

http://angband.oook.cz/ladder-show.php?id=10569

Topping the ladder at 64 MILLION turns o.0; In another roguelike, this is past the end of time.

It would be great to reform score so that it is a meaningful measure of something. For this, dividing by turncount seems to work well.

The ladder's a little off too since all of the topmost entries are ancient.

Dividing by turncount still ends up meaning that the player that spends the most time grinding Greater Draconic Quylthulgs (or similar high-experience-density enemies) will probably have the best score. I mean, yes, you need to get to the point where you can do that grind as efficiently as possible, but there's still a lot of grinding in there. A "proper" score needs to involve non-repeatable actions, IMO. Maybe you could do something like sum up the experience values of each type of enemy killed by the player and then divide that by their turncount. That seems kind of dodgy though since two equivalently-skilled players will score differently depending on the variety of enemies they encounter during their mad dash to the bottom.

This is probably why competitions are graded based on turncount and nothing else.

[Cold_Heart]

A "proper" score needs to involve non-repeatable actions, IMO.

Number of uniques killed modified by how much off-depth that happened, difference between unique level and clvl*2 and divided by turncount

[mushroom patch]

The ladder's a little off too since all of the topmost entries are ancient.

Dividing by turncount still ends up meaning that the player that spends the most time grinding Greater Draconic Quylthulgs (or similar high-experience-density enemies) will probably have the best score. I mean, yes, you need to get to the point where you can do that grind as efficiently as possible, but there's still a lot of grinding in there. A "proper" score needs to involve non-repeatable actions, IMO. Maybe you could do something like sum up the experience values of each type of enemy killed by the player and then divide that by their turncount. That seems kind of dodgy though since two equivalently-skilled players will score differently depending on the variety of enemies they encounter during their mad dash to the bottom.

This is probably why competitions are graded based on turncount and nothing else.

Well, killing Morgoth and Sauron is nonrepeatable. I'd just go with (K_1*max_depth + sauron_bonus + morgoth_bonus)/(turncount_on_win + K_2), where K_2 is not too big, but big enough to avoid weird stuff like getting killed on the first turn by an icky thing next to the stairs resulting in a high score. Something more elaborate could be devised to avoid the potential of decreasing score over the course of the game, I guess. Maybe points separately computed for each dungeon level/bonus, divided by turncount on receiving them, and added up.

[AnonymousHero]

Well, killing Morgoth and Sauron is nonrepeatable. I'd just go with (K_1*max_depth + sauron_bonus + morgoth_bonus)/(turncount_on_win + K_2), where K_2 is not too big, but big enough to avoid weird stuff like getting killed on the first turn by an icky thing next to the stairs resulting in a high score. Something more elaborate could be devised to avoid the potential of decreasing score over the course of the game, I guess. Maybe points separately computed for each dungeon level/bonus, divided by turncount on receiving them, and added up.

Perhaps sprinkle the character level in there as well. (Would have to be the maximum level attained to avoid silly things like intentionally level-draining before the Morgoth fight and such.)

Personally, I think I'm leaning towards removing score entirely from ToME 2.x. It's rather a lot of code for very little benefit, if any.

[the Invisible Stalker]

This is precisely what I did in FAangband, although it happened more as a side effect of another store existing than as an attempt to make endgame gold relevant.

My general position on this is that having more gold than you can use in the late game is simply a consequence of the way the game unfolds, and I'm really reluctant to use it for anything meaningful. Maybe the player should be allowed to build a solid gold house.

I agree. This is one of those rare cases where Angband teaches you something valuable in real life: Money is very important when you don't have any, but then becomes less important as you acquire more and after a certain point it becomes largely irrelevant. Of course I've only ever actually reached that point in Angband.

I like FA's travelling merchant, by the way, but I'm sure I've said that before.

[mushroom patch]

Perhaps sprinkle the character level in there as well. (Would have to be the maximum level attained to avoid silly things like intentionally level-draining before the Morgoth fight and such.)

Personally, I think I'm leaning towards removing score entirely from ToME 2.x. It's rather a lot of code for very little benefit, if any.

Yeah, maybe divide by character level as well, but I don't know. On the other hand, if you can achieve a high character level quickly, I don't think that should be penalized. Obviously, you shouldn't get a higher score for reaching a higher character level. This is the problem with the current system: It rewards behavior that has nothing to do with winning.

I think there are two critical components to having "sport" in a roguelike game: A solid scoring mechanism and an online player community with spectating, game recording, and scoreboard/ladder for online games only. As it stands, there's little sport in merely winning at angband (or most other roguelike games). The competitions on the forums are good, but they're small. They operate on an honor system, which is fine for friendly competition, but doesn't scale well.

[MattB]

Maybe you could do something like sum up the experience values of each type of enemy killed

FAangband already has this.

[MattB]

Am I missing something here, or would anyone have a problem with a ranking that works like this?...

Winners: ranked by lowest turncount
Everyone else: ranked by deepest level, then by lowest turncount.

Does anyone really care, outside of competitions, about their non-winners' place on the ladder? Maybe you want to compare one of your own loser's position compared to another, and this would do that quite adequately.

This method will ensure that no top spot on the ladder is ever impossible to beat.

Further, and I've said this before, I firmly believe that there should be a separate ladder for each release of Vanilla. When 4.0 is officially released, it will be great fun, like an alternative compo, as people try to beat it in the fewest turns possible. There are already separate ladders for (no doubt excellent and worthy) variants with, like, four entries. All we need is a distinct ladder created for Vanilla every couple of years or so. Please.

[MattB]

Remove gold! Down with gold!

"Some men aren't looking for anything logical, like money - they can't be bought, bullied, reasoned or negotiated with - some men just want to watch the world burn." - Alfred Pennyworth, Batman's butler.

[mushroom patch]

Am I missing something here, or would anyone have a problem with a ranking that works like this?...

Winners: ranked by lowest turncount
Everyone else: ranked by deepest level, then by lowest turncount.

Does anyone really care, outside of competitions, about their non-winners' place on the ladder? Maybe you want to compare one of your own loser's position compared to another, and this would do that quite adequately.

This method will ensure that no top spot on the ladder is ever impossible to beat.

Further, and I've said this before, I firmly believe that there should be a separate ladder for each release of Vanilla. When 4.0 is officially released, it will be great fun, like an alternative compo, as people try to beat it in the fewest turns possible. There are already separate ladders for (no doubt excellent and worthy) variants with, like, four entries. All we need is a distinct ladder created for Vanilla every couple of years or so. Please.

Yeah, I think that's basically the essence of a good scoring system. No score but turncount would make sense (for winners), given there's a single objective in the game. Seems like if you lose, it hardly matters what your score is anyway, but yes, something involving max depth and turncount would make sense (although, I'd say you want turncount at which you achieved your max_depth or sth more complicated to avoid one's score going down over time, see my comment up thread).