Additional gold sink

[the Invisible Stalker]

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.

Or just accept that ladder position has no semantic content. I don't think there's really a way to make it meaningful as long as there are a variety of birth options, play styles, versions, etc. You can fix the version issue, as you said, but the others won't go away.

[mushroom patch]

Or just accept that ladder position has no semantic content. I don't think there's really a way to make it meaningful as long as there are a variety of birth options, play styles, versions, etc. You can fix the version issue, as you said, but the others won't go away.

There is a lot of value added by ladders with a sensible scoring system. They make the game far more replayable for advanced players, because they provide a source of depth that doesn't exist in the play-around "everyone who wins is the winner" mode of play. Competition drives players to explore aspects of the game that are invisible if you just want to win by the time you die IRL.

It's easy to say score is meaningless when you have a game where score really is meaningless (in angband, just a matter of how long you played the character). In games where this isn't true, there's a lot of sport and interest to be had in competing for high scores. See, e.g., dcss.

edit: also, of course, an actually competitive ladder system has to be auditable. In particular, the games must be played online. Different ladders could exist for different choices of birth options, though it would probably make sense to decide on a reasonable set of such options and standardize them as different "formats" with associated ladders for each.

[Zireael]

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.

Good points both on the scoring and the separate ladder for each vanilla release.

[the Invisible Stalker]

There is a lot of value added by ladders with a sensible scoring system. They make the game far more replayable for advanced players, because they provide a source of depth that doesn't exist in the play-around "everyone who wins is the winner" mode of play. Competition drives players to explore aspects of the game that are invisible if you just want to win by the time you die IRL.

It's easy to say score is meaningless when you have a game where score really is meaningless (in angband, just a matter of how long you played the character). In games where this isn't true, there's a lot of sport and interest to be had in competing for high scores. See, e.g., dcss.

edit: also, of course, an actually competitive ladder system has to be auditable. In particular, the games must be played online. Different ladders could exist for different choices of birth options, though it would probably make sense to decide on a reasonable set of such options and standardize them as different "formats" with associated ladders for each.

I suppose it depends on what you want. After I've won I've no particular interest in winning faster. Winning with a different race/class combination, sure. With a more challenging set of birth options or some other challenge, sure. But just faster? No thanks. There's nothing at all wrong with that; it's just not for me. So updating the ladder's scoring is largely irrelevant for me, as long as I can still use it to see how characters of my class and around my level died.

[AnonymousHero]

I suppose it depends on what you want. After I've won I've no particular interest in winning faster. Winning with a different race/class combination, sure. With a more challenging set of birth options or some other challenge, sure. But just faster? No thanks. There's nothing at all wrong with that; it's just not for me. So updating the ladder's scoring is largely irrelevant for me, as long as I can still use it to see how characters of my class and around my level died.

Agreed, I think . I have absolutely no objection to the ladder showing different measures of "winnitude" (or lossitude as the case may be), but I don't the that the concept of "score" makes sense in-game. (E.g. for the list of previous characters you get when you exit.)

EDIT: Just to elaborate a little bit: I have a character who's number #6(ish) on the all-time ladder (which is apparently ranked by EXP), but I actually only played that character to completion because a) it was hilariously overpowered (with 2 or 3 weapons slots, and 50% all-round damage reduction) and I wanted to demonstrate that, and b) I wanted to demonstrate that extending XP to 99M (or whatever it was) and levels to 75 was completely broken in a game that wasn't fundamentally designed for such levels. It doesn't give me any pride to have such a character, but if people want to sort by "max score across all variants" there it is. I don't think it makes even the remotest of sense to compare that character to an Angband 3.0.9 character.

[AnonymousHero]

Oh, yeah, and: Remove score! Down with score![1]

[1] Perhaps even before we remove gold!

[the Invisible Stalker]

Agreed, I think . I have absolutely no objection to the ladder showing different measures of "winnitude" (or lossitude as the case may be), but I don't the that the concept of "score" makes sense in-game. (E.g. for the list of previous characters you get when you exit.)

Okay. I see what you mean. And yes, the scoring in that list is totally broken. Back when multi-user systems were still a thing it might have been worth fixing, but now it would probably be better to sort that list by date rather than "score".

[AnonymousHero]

Okay. I see what you mean. And yes, the scoring in that list is totally broken. Back when multi-user systems were still a thing it might have been worth fixing, but now it would probably be better to sort that list by date rather than "score".

Indeed.

I'd actually be really interested if anyone is actually playing *bands on real multiuser systems any more. I know that there are *hack variants that are still played online with shared score lists, but I think that's kind of a different situation to V.

EDIT: Btw, I do think *bands should perhaps track further interesting stats, that e.g. PosCheng does (inherited from Entro, I think) such as "real time spent with game open". The more of these stats all the games could track, the better. We might even discover new way to devise a score!

[mushroom patch]

I suppose it depends on what you want. After I've won I've no particular interest in winning faster. Winning with a different race/class combination, sure. With a more challenging set of birth options or some other challenge, sure. But just faster? No thanks. There's nothing at all wrong with that; it's just not for me. So updating the ladder's scoring is largely irrelevant for me, as long as I can still use it to see how characters of my class and around my level died.

I don't understand how this responds to my comment. Do you dispute that having a scoring system along the lines of what's been described in thread and having online multiuser servers w/ a centralized scoring ladder has the effects I describe? If so, I suggest you look at DCSS, which has exactly what I describe and a vibrant online play community.

re: *hacks, indeed, nethack on public telnet has been continuously popular for decades. DCSS is big online too. Angband could be too, it just needs organization and infrastructure, and (as I say) probably better scoring. The huge number of angband variants including Sil, still a hot topic in the roguelike world, could attract some interest.

Also, agreed that many different statistics factor into play on multiuser systems. Real time, winrate, and streaking are all good things to keep track of in an auditable system.

[AnonymousHero]

re: *hacks, indeed, nethack on public telnet has been continuously popular for decades. DCSS is big online too. Angband could be too, it just needs organization and infrastructure, and (as I say) probably better scoring. The huge number of angband variants including Sil, still a hot topic in the roguelike world, could attract some interest.

My impression after watching a number of DCSS LP's recently was that the people playing weren't particularly concerned with score. It was more that "Yeah, I beat the game as a Tengu Enchanter!" (or whatever it was).

I didn't get the impression that the precise score mattered to them -- in fact this particular gamer specifically mentioned that his current character was much better (in terms of achievement) than the current top scorer in his high score list. (Since it was a "challenge" character).

[mushroom patch]

My impression after watching a number of DCSS LP's recently was that the people playing weren't particularly concerned with score. It was more that "Yeah, I beat the game as a Tengu Enchanter!" (or whatever it was).

I didn't get the impression that the precise score mattered to them -- in fact this particular gamer specifically mentioned that his current character was much better (in terms of achievement) than the current top scorer in his high score list. (Since it was a "challenge" character).

There's an active speedrunning/top scoring scene. Speedrunners and megazig characters tend to attract the largest/most consistent audiences, though often enough people just go where other people are speccing. LPs don't often capture the more competitive side of games by their nature. It's true people who think about stats tend to think in terms of winrate and streaking more than speedrunning though. Also, of course, speccing and chatting is good times.

[the Invisible Stalker]

Indeed.

I'd actually be really interested if anyone is actually playing *bands on real multiuser systems any more. I know that there are *hack variants that are still played online with shared score lists, but I think that's kind of a different situation to V.

EDIT: Btw, I do think *bands should perhaps track further interesting stats, that e.g. PosCheng does (inherited from Entro, I think) such as "real time spent with game open". The more of these stats all the games could track, the better. We might even discover new way to devise a score!

How clever is the real time tracking feature? If I wander off from the keyboard for a few hours without saving are those hours counted, or does it figure out that if there haven't been any keystrokes recently then I probably haven't been playing?

[the Invisible Stalker]

I don't understand how this responds to my comment. Do you dispute that having a scoring system along the lines of what's been described in thread and having online multiuser servers w/ a centralized scoring ladder has the effects I describe? If so, I suggest you look at DCSS, which has exactly what I describe and a vibrant online play community.

re: *hacks, indeed, nethack on public telnet has been continuously popular for decades. DCSS is big online too. Angband could be too, it just needs organization and infrastructure, and (as I say) probably better scoring. The huge number of angband variants including Sil, still a hot topic in the roguelike world, could attract some interest.

Also, agreed that many different statistics factor into play on multiuser systems. Real time, winrate, and streaking are all good things to keep track of in an auditable system.

I probably wasn't very clear. My point is that angband characters, even within a version, variant and race/class combination, aren't necessarily directly comparable. I've been playing artifactless recently. I've found it's harder, not surprisingly. Should the scoring attempt to adjust for that? If so, by how much? My worry is that if we try to reduce everything to a single number then there's a big temptation to play to maximise that number. One of the strengths of Angband is that there are a lot of different ways to make it challenging. If you are Timo you can play a bookless, artifactless, hobbit mage equipped with only a permacursed shovel and simply try to get a win. If you are Eddie you can sprint to level 100 and kill Morgoth before he even knows you're there. Any attempt to assign scores to those characters will make one of those choices suboptimal, and thus subtly steer people away from it. Of course we already have scoring systems, in the ladder, in the exit screen, in the comps, etc. But comps are a special case and in the other cases the scoring is so obviously broken that I hope no one takes it seriously. It's the obvious brokenness of those systems which renders them harmless.

[Nick]

Starting score 0, +1 for killing Sauron, +1 for killing Morgoth. That would clarify some things.

[the Invisible Stalker]

Starting score 0, +1 for killing Sauron, +1 for killing Morgoth. That would clarify some things.

I would propose starting score 1 and then -1 for being killed.