Angband Philosophy I: Player choice

[Nick]

I have outlined a few of my ideas about maintaining, but probably not very coherently. This (and maybe even follow-up posts) will be an attempt to be a bit more thorough.

It seems to me that one of the big strengths of Angband is the freedom of choice it gives to its players. There are two fixed objectives, and how you achieve those is very largely up to you. As a consequence, we have

Rule 1: Angband is a game you are playing against yourself.

So if you're "cheating", you're only cheating yourself (unless you're playing in a competition where there are defined rules and an honour system and you go against that).

Now I'm taking the responsibility for maintaining Angband, but that doesn't make me the boss of it. I'd certainly feel a bit silly telling someone like Timo how to play, and if it's like that for Timo, why should it be any different for a new player? Which brings us to

Rule 2: How you play Angband is up to you.

Back in the days of rgra, there used to be a standard pattern when anyone complained about some aspect of the game: a small amount of discussion, and then if they were still complaining, someone would say "You should make a variant".

So one of my big goals is to make the game easier to modify. Don't like hounds, or identify, or the mage spells? Change 'em. And while there will still be a standard (erm, Vanilla) Angband which is the one that's released, I will be encouraging people to mess with stuff and see whether they like it better.

Discuss.

[Therem Harth]

(1) By all means. Competitions should IMO prohibit save-scumming, but I don't see anything wrong with having it as a difficulty option, that is clearly marked in the save file.

- Note "difficulty option", not "cheat option". It makes the game easier. It doesn't make it not-a-even-a-game-anymore, which is what Cheat Death does.

- Also, since I am now working full-time, I can see very well where people would rather save-scum than have to restart from scratch every time a character dies.

tl;dr it's a game. It's supposed to be fun. It has to accommodate real life, not the other way around.

(2) Yes please. I'll admit that V and its quirky codebase have some charm, but that charm quickly wears thin when you get serious about adding new features.

...

So, yeah, looks like I agree completely.

[debo]

What I find interesting now is how the apparent traffic on oook (ladder uploads) is largely divided between Sil and Poschengband, which basically represent the most extreme dipoles that one could take when designing an angband variant. I'm not sure that variants that just change a thing or two will have much traction, but I'm willing to find out. (It also helps that those two variants have fairlyt frequent rewleases , whereas most others have stagnated.)

[Timo Pietilä]

So one of my big goals is to make the game easier to modify. Don't like hounds, or identify, or the mage spells? Change 'em.

Some sort of standardized spell file like monster.txt with easy to understand syntax that game then parses as spell?

N:1:Magic missile
T:Bolt:mana
C:multi
D:<damage code goes here in some fashion>

Something like that?

[Nick]

Some sort of standardized spell file like monster.txt with easy to understand syntax that game then parses as spell?

N:1:Magic missile
T:Bolt:mana
C:multi
D:<damage code goes here in some fashion>

Something like that?

Yes - here's an example:

spell:Stinking Cloud:3:2:27:3
effect:PROJECT
dice:$B
expr:B:PLAYER_LEVEL:/ 2 + 10
ball:POIS:2
desc:Shoots a radius-2 poison ball.

I should point out that much of this is molybdenum's work.

[Derakon]

What I find interesting now is how the apparent traffic on oook (ladder uploads) is largely divided between Sil and Poschengband, which basically represent the most extreme dipoles that one could take when designing an angband variant. I'm not sure that variants that just change a thing or two will have much traction, but I'm willing to find out. (It also helps that those two variants have fairlyt frequent rewleases , whereas most others have stagnated.)

I suspect that "mini-variants" wouldn't get an especially large player base, but they would serve as good test-beds for new changes. "Hey, I did X to the game, and I really like how it plays, you should try it out" is a lot more convincing than "I think doing X would be a good idea".

[fizzix]

I suspect that "mini-variants" wouldn't get an especially large player base, but they would serve as good test-beds for new changes. "Hey, I did X to the game, and I really like how it plays, you should try it out" is a lot more convincing than "I think doing X would be a good idea".

I entirely agree. This is also how I motivated for some game changes like making teleport other a bolt. There was a ton of theorizing, somewhat convincingly, by Eddie and others that if it was a bolt, it would force players to use ASCs and summoning would be too overpowered. Turns out that that wasn't true, and the only way to figure that out was to play a few games with it.

Unfortunately, it takes far more time to play a game of angband, than it does to make a few in game modifications. So this wasn't really feasible to do.

I had asked Nick about pulling a lot of the game constants out into an edit or namelist file. And if I ever get the motivation to start it, it would be great. (Once it's started and I figure out how to do it, or see how someone else did it, it'll be easy to replicate). The idea here would be it would allow the player to really modify some important features of the game without compiling. Stuff like monster spawn rates, OoD levels, item drop curves, vault parameters, level size, monster/loot density, etc etc. Whereas, now the edit files act like a surgical tool, where you can change a single parameter or monster with ease, it's more difficult (you need to recompile) to change some of the larger features of the game. As such we're missing out on a whole range of exploratory space for players to run around in.

[Lionmaruu]

I agree with both points, and that's why I am against any enforced mechanic about level scumming or fully exploring levels and taking your time (like invincible mob spawning or some sort of "hurry" mechanism). That said I would love some birth options that could add those options and more, like no shopping or no level regeneration.

[Timo Pietilä]

I agree with both points, and that's why I am against any enforced mechanic about level scumming or fully exploring levels and taking your time

I'm guessing "about" -> "against".

I agree. This is why I hated Sil. It forces you to dive which is too big violation against my preferences that I can't play it. No matter how good it could otherwise be. Forced diving is also against basic angband philosophy: infinite dungeon, infinite time to complete game.

[taptap]

I'm guessing "about" -> "against".

I agree. This is why I hated Sil. It forces you to dive which is too big violation against my preferences that I can't play it. No matter how good it could otherwise be. Forced diving is also against basic angband philosophy: infinite dungeon, infinite time to complete game.

Angband has in fact heavy handed mechanisms about level scumming in place, it simply inflates most early finds out of utility. Just ask yourself how much of your early game equipment is still with you in the endgame, how often you use potions collected early, later on etc. And the choice about inflation is as much a "player choice" as is forced descent in other games.

[Timo Pietilä]

Angband has in fact heavy handed mechanisms about level scumming in place, it simply inflates most early finds out of utility. Just ask yourself how much of your early game equipment is still with you in the endgame, how often you use potions collected early, later on etc. And the choice about inflation is as much a "player choice" as is forced descent in other games.

Forced descend by definition is not player choice. Inflation is inflation, but it doesn't force you to do anything.

[fizzix]

Forced descend by definition is not player choice. Inflation is inflation, but it doesn't force you to do anything.

This is correct. There's a big difference between a game that allows you infinitely scum for most inferior gear and one that forces you down whether you are ready or not.

From a game design standpoint, you can actually make a spectrum between "tight" and "loose." A maximally "tight" game will give you exactly what you need in order to progress. All characters (of the same class/race) that get to the same stage will be roughly equally powerful. A "loose" game will allow for a tremendous amount of variation in gear and character progression at given points. Linear RPGs are tight, Roguelikes, in general, are loose.

Limiting to roguelikes, out of ones I know, Brogue is probably the "tightest" game. Sil is on the tighter side but still not too tight. Somewhere in the middle are games like DCSS and ToME. Angband (and bands in general) have traditionally monopolized the "loose" side of things.

One of the reasons to want a "tight" game is it gives you something to balance against. If you know the player only has 20 levels of monsters and items, you need to make sure that they'll get enough gear in those 20 levels to win the game some targeted percentage of the team. The tighter the game, the more precise the allocation values are. With a loose game, you don't need to get allocation levels as exact, since any slop can be compensated for by the player exploring side-quest areas, or repeating levels. With infinite levels, you can be really sloppy with allocation, and still be fine. And previous versions of Angband, before Myshkin, Magnate and I started really looking at statistics, were really really sloppy.

I think we've actually managed to do something positive in that we've tightened up the allocation levels without actually tightening up the gameplay (unless you choose a tight option like forced descent or ironman).

[taptap]

Not sure loose/tight is really the question here. You have a choice of balancing mechanism. If you choose inflation (instead of some kind of limitation), you can allow infinite repetition. This "player choice" is paid for with inflation, which is in my opinion the worst imaginable mechanism you can have in a game, especially one relying on a combat system that notoriously scales badly. Now, I don't expect this to change and agree that another balancing mechanism isn't necessary with inflation all around, but I find it a bit odd that this is advertised as "player choice" without even mentioning the price paid to make it viable.

[Derakon]

The "player choice" in question is a natural way for players to set their own desired difficulty level. If you're having trouble, then you can slow down and grind for a bit until you're better-prepared. If you feel that you're doing well, then you can effectively skip a portion of the game and go straight to your optimal level of challenge.

Games that operate on a timer (implemented by whatever means) tend to balance things such that the player needs to do as much as possible within the constraints of the timer if they want to do well. Which means that there's no high-level choice involved at all: you do as much grinding as you can get away with before you're forced to move on. I'm oversimplifying here to an extent, but even in Sil where you're free to dive "early", most players (as I understand it) will hang around at 950' until they're forced by the game clock to move on to the endgame.

I'm not trying to make a value judgement here. Neither system is inherently better than the other. Angband is prone to having players play so cautiously that they stop having fun, simply because they can; meanwhile, I know many players that find "timered" games stressful (just as many players find that form of stress to be fun!). It's all about selecting the playstyle that suits you best.

Stepping back into the mainstream, consider a game like, say, Castlevania: Symphony of the Night. If you play "normally" (killing enemies that cross your path and finding most of the new equipment) you'll likely find it to be a fairly easy game. If you're having trouble, then you can go grind for a bit and your improved stats will significantly improve your survivability. Or, if you're highly skilled, then you can play with self-imposed challenges -- I've played the game under a "no equipment" constraint where you have to punch everything to death, for example.

Timered games are capable of supporting that kind of flexibility of gameplay, but not to the extent that untimered ones can. That's where the "player choice" that Timo was ranting about comes from.

[krazyhades]

I find Nethack's endgame to be somewhere between the two, but a step to the side. The game can last as long as you want, but in the end you need to be prepared for a race (time pressure) to make your escape.