Question about the design philosophy.

[Pourush]

I read the design philosophy, and I have a question I want to ask about it:
Why is the design philosophy concerned with specific parts of a good videogame, rather than on the bigger issues of a good videogame, like it being fun, not a waste of time, and for some games, increasing real-world skills?

Also, I'm not sure if I should post in here, or in Crazy Yiuf's Corner. I posted here because it was technically game design discussion, even though it didn't necessarily fit the subtitle to the thread. Did I guess the right board?

[cerebovssquire]

One can't really work with the design principle "it should be fun", not only because this is very generalised (it can mean so many things), but also because "fun" is rather subjective. So it is better to have specific design goals, which will result in the game being fun for a specific target audience. Obviously the design goals are intended to make the game "fun" (with a roguelike definition of the word) and interesting.
Of course these design goals are partly geared against what, according to expectations, are unfun for most people (namely, tedium and monotony).

Likewise no moral philosopher (this is the worst example, but it is the first that comes to mind) will write a book saying "be good" but rather a book about "this is how you can be good in my opinion".

[mageykun]

Welcome to the forums!

Crazy Yiuf's is kind of our general chat. Things can get silly over there. Not that you couldn't discus philosophy over there (in fact, philosophers tend to pop up with surprising frequency), but you're better off here if you want a serious discussion about design principles.

To answer your question, I'd say our philosophy concerns itself with specifics because of the manner in which it's meant to function. It's meant to be a kind of mission statement, and a guide for future improvements and changes. Crawl is somewhat of a long term project, with an impermanent set of contributes and developers. A shared, clearly stated philosophy is necessary if you want to unite the efforts of these disparate people to work in the same direction over time.

Goals like "Being fun", "not a waste of time" etc by themselves are problematic because they're rather vague, non-specific and subjective. In almost all cases you want a game to be fun- but there's no universal metric on what constitutes fun.

Edit: hah, only ninja-ed once. :p

[Pourush]

Fair enough. But I have another question.

The game has a lot of different skills it requires. The basic skill it seems to train is risk management, but due to a newbie not understanding the way the game system works, they play a game that trains that skill only partially, due to the fundamental uncertainty as to where risk is and isn't. Here are a few notable examples of what a newbie is unaware of:

What are the notable playstyles of the game, and what do they require?
Why do I die so early on in the game?
What are the differences between what I can expect from the different races?

These things may be clear to a person who plays the game a lot or gets help. But to a newbie, it seems to make the game less fun to know the most basic consequences of the rules. So if the clarity acquired from either help or experience is so important to this game, why is clarity only a minor design goal?

[bobross419]

Some folks have an easier time with some playstyles than others... some playstyles folks find immensely boring while others might find them fun.

With that said, the in game help is pretty clear on what can be expected from the different races:

High Elves
This is a tall and powerful Elven species who advance in levels very slowly, requiring half again as much experience as Humans. They have good intelligence and dexterity, but suffer in strength. Compared with Humans, they have fewer HP but more magic. Among all Elves, they are best with blades and bows.

Magicians
===========================
A magician is not an available character background by itself, but a type of background, encompassing Wizards, Conjurers, Enchanters, Summoners, Necromancers, Transmuters, various Elementalists and Venom Mages. Magicians are the best at using magic. Among other things, they start with a robe and a book of spells which should see them through the first several levels.

Wizards
A Wizard is a magician who does not specialise in any area of magic. Wizards start with a variety of magical skills and with Magic Dart memorised. Their book allows them to progress in many different branches of the arcane arts.

With that said, I think it can be assumed that anyone thinking of playing DCSS will be at least familiar with the classical classes (Fighter, Ranger(Hunter), Wizard, Summoner, Necromance) and these classes aren't really all that difficult to grasp. Considering that I've ascended (I'm not all that great) with all of them but Hunter (only because I got bored and quit) it isn't too far of a stretch that most of the new players aren't getting hampered by an initial choice that they are familiar with in some way.

Don't forget the Tutorial (which breaks down the basics pretty Barney style) and Hints Mode (which points out some of the things that a lot of us already take for granted).

You can't really expect someone to be able to play a game without some kind of instruction... How would you know that Mario can get coins and stuff from certain blocks? It is completely counter intuitive to slam your head into something metal.

[BlackSheep]

I think, too, that the tutorial is still being refined, so as to make it easier for new players to get acquainted with the game.

[njvack]

Mario even had its little demo video that showed the basics of platformer mechanics.

But Crawl isn't designed to be a game you can pick up and master right away -- it involves learning the mechanics of quite a lot of complex systems and their interactions, in addition to memorizing a bunch of stuff (eg, how hard a griffon is compared to a hippogriff). A lot of the joy in a game like this comes from learning and mastering those systems. Realizing that you've just survived a centaur on D:3 is as fun, IMO, as grabbing the orb. Or getting up and running with a new, unfamiliar build.

In contrast, a platformer like Mario or something like Angry Birds is designed to be simple and intuitive, and provides joy in rather fundamentally different ways.

Not all Roguelikes are Crawl-like (though the prototypical games in the genre are) -- Diablo, The Binding of Isaac, and, say, Porta Lucis are all arguably roguelikes with more accessible mechanics. But they're all less complex systems, at least in some respects. (Disclaimer: I haven't played Diablo 3.)

Whew! That was a long way of saying: Crawl is a well-designed example of a rather niche type of game.

[njvack]

I think, too, that the tutorial is still being refined, so as to make it easier for new players to get acquainted with the game.

Yeah -- someone even did usability testing of the Tutorial. It's neat.

[bobross419]

Actually playing through the tutorial... it pretty clearly lays out the basics.

Edit: Ya Tutorial pretty well covers the basics and Hints Mode is pretty awesome in the amount of game mechanics that it covers without getting stuck in the nitty gritty of formulas and stuff.

[Pourush]

What I'm concerned about is how well does somebody like me, who kind of knows about video games, and picked up the basics of this game pretty quick, but isn't necessarily certain about how to play the game in such a way as to not die. My experience with it is that, in the early game, there's not a lot to do, and because of training from Dungeons of Dredmor, I always explore every dungeon level completely and get bored in the process, thus leading to me quitting the game without saving--because I'm too fickle with my build choices to possibly stick to one build for multiple sessions. It seems I might have raised invalid points in the previous post I made.

[dpeg]

Pourush: Welcome to the game (although I have a hunch it's not the perfect target for your gaming needs).

I have initiated the philosophy section and written large chunks of it, so I feel responsible to reply. (Note 1: I am not an official developer anymore, but Crawl is still very dear to me; I follow what's going on, and I have some ideas I'd like to see get into the game. Note 2: The replies by cerebovssquire (blue) and mageykun (green) were on the point; I'll try to add something new.)

Roguelikes are developed in a larger context -- the design space is huge, considering that Rogue, Nethack, ToME, Letterhunt, Diablo, DoomRL are all roguelikes. Crawl happens to be a roguelike with a long history and a rather large crew (between four and ten active developers at any time, I would say). When I wrote the philosophy section, I was mostly concerned with trying to distill what makes Crawl work so well (compared to Nethack, where I came from). These are the major and minor goals you can see there. Later on, the philosophy section was used, first among developers but then also by players, to say that something is Crawly or un-Crawly. So it gave Crawl a more specific place in the huge design space for roguelikes (although at times it is interpreted too strictly, for my tastes at least -- the recent posting that "doing Lair is a no-brainer", for example).

A completely different way to interpret the philosophy section, which addresses Pourush's question, is that it says to some degree what's unfun to the *devteam*. (And of course, there can be tensions among developers and there is no clear procedure how to resolve those.) Fun is subjective. We have no chance to make the game universally fun for everyone. And honestly speaking, we'd be in an awkward position if we tried, what with our baroque interface, the aged graphics (don't even mention the ASCII), the turn-based and sudden death approach to game.
But even among more hardcore players, fun is diverse. There probably are players who really enjoy grinding in games. (I'd bet that many who grind don't really enjoy it but who am I to judge.) However, everyone in the devteam agrees that grinding is no fun but rather a symptom of bad design. This does not mean that Crawl is grind free but it does mean that we'll think about how to remove a source of grinding if one shows up. In other words, if you happen to like grinding, then Crawl is probably not a game for you (speaking in general, not to Pourush).

So the Crawl philosophy does not try to say what a roguelike is (it assumes that the players know that) and it does not try to say how a game can be fun (because that's impossible). It just lays down a number of basic principles that are accepted among developers (and also most players). More or less sticking to those principles makes sure that Crawl is more or less fun to those who actually play it.

Regarding Pourush's last set of questions: Crawl is not a game to teach you its inner workings in any detail (for example, because they're very formulaic -- by the way, I think I was the developer most adamant about not displaying numbers, this might change over time). However, the idea is that if you learn from your deaths, you can improve. It is true that you won't play perfectly without access to long lists of numbers, formulas, maps (i.e. spoilers) or even the source, but it should be possible to improve your game -- i.e. to get deeper and deeper, obtain a rune, and then win. The game has been won unspoiled, and that's important to us.
I read that you're bored by Crawl's large levels. That's absolutely fine but an indication that Crawl might not be for you. Crawl is one of the (rather few) huge roguelikes (I guess the most known others are Angband and derivatives, Nethack, ToME). It has a large number of those large levels. Winning game of first time victories often take 20 hours. Some players like this epic grandezza, others cannot stand it. Obviously, that's a property which won't change although steps are taken to make the game a bit slimmer.

There is a solution to all your problems, though: Give Brogue a try: https://sites.google.com/site/broguegame/
That game is much smaller in scope. I can vouch for its excellent design -- it does many things better than Crawl can do, and it explains its mechanics in a very clear way. It is also hard but in a good way (a lot to learn before you win).

[Pourush]

Yes, I know Crawl might not be for me. But not necessarily because of the large levels. It's more that in the beginning, can own any monster that spawns in and then just rest up to full. Even if you're a spellcaster without offensive spells, if you're a troll or spriggan you can wreck everything with your fists. It might be different for different races, but for the ones I play most (spriggan, troll) it happens so I don't know about the other races. If it's a group of monsters, you can usually do the same thing, but you'll have to flee eventually. And then go right back in, because they still aren't that threatening. It's a pretty long beginning, so you can wreck everything without a thought for quite a while. I'm pretty unlikely to play it again unless that gets fixed, and I'm not necessarily going to be around for that long, which is okay.

But not having to make any major combat decisions in the early game sounds like it violates the bit in the game's philosophy concerning "no-brainers". Furthermore, I hear that in earlier builds, monsters would wreck characters pretty quick. I read "Several good players maintain that it [pillar dancing] is necessary to beat D:1 consistently.", and it seems like that's been fixed. As for all that talk about game design philosophy, I was testing it to make sure it held up to scrutiny. And the second post I made can be safely ignored. This seems to be near the time where I take my leave, I think I've made all the points that are of interest. Unless there's arguments I haven't anticipated arising.

[mageykun]

I rather think pillar dancing was over-touted. You really don't have to resort to it nearly as often as some of those guides make out.

As to your other point... I find most builds in the early game have what I like to call a gimmick. A thing, that you can do well, by virtue of race, background and/or god, that acts as your initial response to monsters. What this is can vary- but all the recommended combinations have something (what often makes the unrecommended classes hard is they don't have reliable gimmicks). I don't feel it's a no-brainer through to have an initial gimmick to build your combat strategy around. You still have to make tactical decisions in it's application, and strategic decisions in how you will grow and adapt your build based on what resources you find. Plus, there's the initial race / background choice where you decide on a gimmick.

[cerebovssquire]

Well, in the late game your tactical mistakes often won't even screw you over, but will instead force consumable usage at worst. In the early game, one minor tactical mistake will often result in suddenly unavoidable death. I don't see what your problem is. I guess minmay might have a point. Or you are playing too many spriggans and trolls, which changes a lot. Picking two of the four strongest early-game races obviously makes it seem easier than it is to most combinations.

Yes, you have a lot more tactical options later on. However, tactical diversity does not equal tactical depth. Tactical depth is directly linked to weight of choice, meaning that you need to feel consequences for a bad choice. So while you might have 5ish ways to go about a fight early on, and more than 20 later on, the choice between 5 ways of combat early on is much more meaningful than the choice between 20+ means of combat later on.

This is not necessarily a bad thing, to a degree. If the late game was as hard as the early game, permadeath would absolutely suck because death would occur frequently after hours of playtime.

[kefir645]

The only problem I have with randomness in Crawl is dice rolls. In board games, they add to the excitement and player involvement. Not so in computer games where RNG rolls all dices and in Crawl you won't even see the numbers unless you're playing a wizard mode...

For instance, 1d8 (and dices like XdY where X is many times less then Y are all over the place in Crawl...) just means that the spell (magic dart at full power in this case, but it's even worse with IMB which is 2d22...) can deal 1-8 damage and it's just too random. Why not change it to something saner like 3-6 (3d2)? I heard it's going to be fixed, I hope it's true.

[sardonica]

The only problem I have with randomness in Crawl is dice rolls. In board games, they add to the excitement and player involvement. Not so in computer games where RNG rolls all dices and in Crawl you won't even see the numbers unless you're playing a wizard mode...

For instance, 1d8 (and dices like XdY where X is many times less then Y are all over the place in Crawl...) just means that the spell (magic dart at full power in this case, but it's even worse with IMB which is 2d22...) can deal 1-8 damage and it's just too random. Why not change it to something saner like 3-6 (3d2)? I heard it's going to be fixed, I hope it's true.

If THIS freaks you out, remember that any action has a flat chance for failure or success.

How about instead of making crawl deterministic, you adjust your strategies to compensate for the chance of failure?

[kefir645]

The only problem I have with randomness in Crawl is dice rolls. In board games, they add to the excitement and player involvement. Not so in computer games where RNG rolls all dices and in Crawl you won't even see the numbers unless you're playing a wizard mode...

For instance, 1d8 (and dices like XdY where X is many times less then Y are all over the place in Crawl...) just means that the spell (magic dart at full power in this case, but it's even worse with IMB which is 2d22...) can deal 1-8 damage and it's just too random. Why not change it to something saner like 3-6 (3d2)? I heard it's going to be fixed, I hope it's true.

If THIS freaks you out, remember that any action has a flat chance for failure or success.

How about instead of making crawl deterministic, you adjust your strategies to compensate for the chance of failure?

It doesn't freak me out, I just think Crawl would be a better game if it was slightly more deterministic. I'm not talking Desktop Dungeons deterministic, just less dices that range from, say, a scratch to an instakill. As for a flat chance of failure for every action, that one makes perfect sense.

[njvack]

Well, in the late game your tactical mistakes often won't even screw you over, but will instead force consumable usage at worst.

You underestimate my capacity for tactical mistakes.