Depth vs. Complexity

[Magnate]

I'm probably still the only one who has played Halls of Mist mystics in the last version. That's an example of board complexity going way over my enjoyment treshold!

Ironband mana increases board complexity a lot, especially when mixed with the tight torch and mist phantasm clocks. The board complexity feels just about right for semi-spellcasters, but with full spellcasters it seems to rise above my pain treshold...

Maybe I just need to get used to it. And the player can easily adjust the board complexity by choosing a different class.

(EDIT: I'm not sure whether I should call this strategic or board complexity. I'm talking about using your resources in an optimal way to solve the various problems on the board. The scale is only one dungeon level, so I hesitate to call it strategic complexity. Maybe it's a borderline case.)

Nope, that's pretty clearly board complexity. Strategic complexity is about *when* stuff happens - this level, next level, 5/10/20/50 levels time.

[Mikko Lehtinen]

I'm probably still the only one who has played Halls of Mist mystics in the last version. That's an example of board complexity going way over my enjoyment treshold!

Ironband mana increases board complexity a lot, especially when mixed with the tight torch and mist phantasm clocks. The board complexity feels just about right for semi-spellcasters, but with full spellcasters it seems to rise above my pain treshold...

Maybe I just need to get used to it. And the player can easily adjust the board complexity by choosing a different class.

(EDIT: I'm not sure whether I should call this strategic or board complexity. I'm talking about using your resources in an optimal way to solve the various problems on the board. The scale is only one dungeon level, so I hesitate to call it strategic complexity. Maybe it's a borderline case.)

[Mikko Lehtinen]

Personally I think Angband is all about comprehension complexity. There is a little board complexity, in those rare situations where you have a number of different enemies in LoS with a range of different abilities they might use - but that almost always leads to you using an escape and looking for a safer battleground.

Yeah, that what I was thinking too!

In Halls of Mist I've been trying to increase board complexity by pushing combat into rooms, letting the player fight two monsters in melee at the same time, having monsters appear in heterogeneus groups, and by introducing tactical terrain features. Having to keep track of time you spend on exploring, fighting and resting increases board complexity too.

That's also the reason why Mist might benefit from non-opposed combat rolls where you don't have to inspect every single enemy in sight to find your hit and parry chances. When there are multiple opponents in most combats, the game needs to be designed so that the situation can be analyzed *fast*. For most other *bands, that is not important. So you may safely forget my comments about the benefits of non-opposed rolls and binary monster stats.

[Magnate]

This is also how Magic the Gathering is designed. Read this.

Three different kinds of complexity are introduced in the article: Comprehension Complexity, Board Complexity, and Strategic Complexity. We've been mostly talking about Comprehension Complexity here.

Thank you - it's a long time since I've read MaRo - I stopped keeping up with MtG about five or six years ago.

Now I'm more convinced than ever that MtG is in fact a counter-example to the Penny Arcade thesis quoted in the OP: it is not only incredibly complex, it *relies* on that complexity for its depth.

That's not to say that I disagree with the thesis, just that MtG is not an example of it in action.

Personally I think Angband is all about comprehension complexity. There is a little board complexity, in those rare situations where you have a number of different enemies in LoS with a range of different abilities they might use - but that almost always leads to you using an escape and looking for a safer battleground. (In fact I think a one-to-many game will always have much less board complexity than a many-to-many game.) There is almost no strategic complexity, beyond inventory and home management. (EDIT: **this** is why I like skill variants so much, because skill choices introduce strategic complexity.)

So that makes our task a lot easier, really: cut down the comprehension complexity, and make sure it ramps up smoothly and slowly throughout the game.

Perhaps we can reach a consensus after all.

[Derakon]

I just watched the episode and it actually goes quite a bit further than Derakon's pithy summary. For example, when counting complexity they don't just count the ruleset, but the amount of rule complexity they player has to deal with, weighted by when and how she/he has to deal with it. So early game complexity is much worse than a slowly revealed approach, and a tutorial helps cut complexity (on this definition) considerably. Complexity in their definition is basically the cognitive load on the player.

Thanks for the expansion of the summary; I guess I was a bit overly brief.

Extra Credits has extra discussions about how to lead a player into a complex game, including on building tutorials (pithy summary: keep interesting things happening even if they don't require much skill on the player's part; don't frontload everything). Since it's generally-agreed that Angband ought to have a tutorial, this would be a good starting point for someone thinking about working on one.

There's also some discussion of "first-order optimal strategies", which are basically the easily-learned, reasonably-effective techniques you use early in a game (e.g. in Angband, Potion of Speed + whack'n'back). The game needs them to get the player hooked -- if there isn't some reasonably-accessible way to do well, then the player just ends up blocked right off the bat and doesn't get to have fun. But at the same time the strategy shouldn't be dominant, because it doesn't require much skill; if it is dominant then the player can just walk through the game applying that strategy to every situation, and they'll be bored. The game needs to guide the player to adapt their strategies, both so that they'll be able to handle the challenges where the FOO strategies break down, and so that they'll have a more varied toolkit to help keep the game interesting.

[Mikko Lehtinen]

I just watched the episode and it actually goes quite a bit further than Derakon's pithy summary. For example, when counting complexity they don't just count the ruleset, but the amount of rule complexity they player has to deal with, weighted by when and how she/he has to deal with it. So early game complexity is much worse than a slowly revealed approach, and a tutorial helps cut complexity (on this definition) considerably. Complexity in their definition is basically the cognitive load on the player.

What is fascinating is how closely this matches Scatha and my view in recent off-line discussion. For example, we are less worried about complexity on rare artefacts or on advanced player abilities that are out of reach in most play-throughs.

This is also how Magic the Gathering is designed. Read this.

Three different kinds of complexity are introduced in the article: Comprehension Complexity, Board Complexity, and Strategic Complexity. We've been mostly talking about Comprehension Complexity here.

[Mikko Lehtinen]

In Sil, it is simpler overall to use opposed rolls because every single skill check works like that so you get it for free. With a single roll, you get a piecewise linear S-curve of increasing chance of success: no success, linearly increasing success, full success. With opposed 1d10s (or whatever) you get a smoother S-curve which I like, though I admit it is slightly more complex.

Ron Edward's Sorcerer, a roleplaying game, uses opposed pools of d10's. You don't add the rolls together. Instead, you player with the best single die roll wins. If there's a tie, count the second die. You then get the degree of success by counting how many of your dice beat the opponent's highest die.

The only problem with any sort of opposed rolls or a changing difficulty level is that you then have to let the player know how good the opposition is. At least if the player is supposed to make decisions based on the difficulty level. If the UI is not very clear on this, the difficulty levels don't actually increase the tactical depth of the game very much. If there's extra keypresses or mouse movements required to find out your chances, people won't bother do keep checking the chances in every fight.

Difficulty levels might actually decrease the tactical depth of the game if the player loses the idea of how good his chances are, or simply stops caring. Compare this to situation where the player knows that his Saving Throw is 53% and uses that information to his advantage.

[ekolis]

Here's a radical proposal: The next maintainer for Angband should be... NOBODY! Everyone can keep all their sacred cows unslaughtered, and if anyone wants to make any changes, well, that's what variants are for! And we could have variants of variants, too! (Hey, it worked for Hengband...)

[Magnate]

I'm honestly pretty surprised that so many of the things on your list are still undone. They are pretty poor examples of game design and were just accreted over time or put in as quick fixes. I think the people who put them in would be embarrassed to see them treated as sacred cows.

That's a fabulous understatement. During the great GPL push I spoke on the telephone to Alex Cutler, and he was astonished that anyone had heard of Angband, let alone was still playing it. I am certain the idea that it had any quintessence that must remain inviolate did not cross his mind.

It's kind of hard to explain why making changes to Angband is such an unpleasant experience - either you live through it, or you read a couple of years-worth of barracking to get the flavour. But if you call it somethingelseband, everybody says how well you've done to fix all those stupid flaws. You really couldn't make it up.

[half]

Seriously, while there are problems with the sil system it's a great starting point. I would seriously consider scrapping the V combat system in pyrel and building off of Sil's system.

You are very welcome to do so. People should feel free to take the ideas introduced in Sil and put them in other games where they think they would work. Perhaps even more obvious than taking Sil's combat system is taking its system for minor skills like saving throw and disarming.

For example, with HP if you have a linear scaling with level and CON, you get a quadratic relation.

Note that in Sil the effect of Con on HP, Gra on Voice, and Str on carrying capacity is exponential rather than linear: each point increases the amount by 20%. This works well for these things. If you want finer grain control, 10% or 5% would also work. The main idea is presumably to make them smooth (C^infinity...) and logical.

I'm personally a fan of striving for symmetry, because it's silly to have a difference between the way a player deals damage and the way a monster deals damage (as is currently the case.) This means removing multiple blows entirely and rethinking how monsters attack.

In Sil it is very symmetric and this was one of the aims. I started off designing player vs player scenarios on paper, making sure that its results for a warrior in armour with a longsword vs a fencer in leather with a shortsword worked out properly...

We moved to single blow combat by default for the player to avoid message spam or extra attack button presses. The main downside was removing a natural aspect of damage reduction having more effect against light weapons users. I'm still a bit sad about that going as it would have been nice simulation/flavour for no extra complexity. As it happens, we do have a few ways of getting extra attacks, but they come through two advanced abilities and through a herb of rage. This does work well.

Right near release, we decided to remove multiple attacks from monsters too and this went very well. We cut down their attacks to 2 at most: a regular attack and a more interesting one. They have a 2/3 chance of doing the regular attack and a 1/3 chance of the interesting one. This gives the variety without the message spam etc.

Even if you go with multiple attacks, I'd really cut it down to 3 at most, with 1 and 2 being common.

* I think saving throws should work as follows, the effect rolls a die and if it's higher than your saving throw, you succumb. if it's lower, you do not. High enough saving throw = immunity from weaker effects. I think this is superior to either competing dice rolls or having the player roll against a static value.

That is a pretty good system, and very simple. In Sil, it is simpler overall to use opposed rolls because every single skill check works like that so you get it for free. With a single roll, you get a piecewise linear S-curve of increasing chance of success: no success, linearly increasing success, full success. With opposed 1d10s (or whatever) you get a smoother S-curve which I like, though I admit it is slightly more complex.

[half]

What a great thread! I've found this the most interesting thread I've read for months.

I just watched this episode of Extra Credits, which talks about the relationship between depth and complexity in videogames.

I just watched the episode and it actually goes quite a bit further than Derakon's pithy summary. For example, when counting complexity they don't just count the ruleset, but the amount of rule complexity they player has to deal with, weighted by when and how she/he has to deal with it. So early game complexity is much worse than a slowly revealed approach, and a tutorial helps cut complexity (on this definition) considerably. Complexity in their definition is basically the cognitive load on the player.

What is fascinating is how closely this matches Scatha and my view in recent off-line discussion. For example, we are less worried about complexity on rare artefacts or on advanced player abilities that are out of reach in most play-throughs.

They even talk of complexity as a currency to spend in order to get as much depth as possible, which is a major part of the Scatha/Half game design ethos (in Sil and other games). It is not the only part (e.g. there is a lot of weight on flavour and consistent theme in Sil) but it is a major part.

I believe Sil has already done many of these.

Yes. Indeed when starting to design Sil, I just basically started with Angband and then either cut things or replaced systems which gained very little depth for their complexity with tighter and more interesting systems. It was quite some time before I actually started adding unique new things, and by that stage there was a lot of saved complexity to spend so we could have a lot of fun with some splashy new features.

I'm honestly pretty surprised that so many of the things on your list are still undone. They are pretty poor examples of game design and were just accreted over time or put in as quick fixes. I think the people who put them in would be embarrassed to see them treated as sacred cows.

[fizzix]

I like simple rules because they allow me to put in more tactically interesting features without confusing the player. As an example, a skill roll in Halls of Mist is always 1d100 under your score. That simplicity allowed me to add lots of content that utilizes that rule: the best example is Mist's terrain features that ask for various skill rolls all the time (Jumping, Spell Save, Perception, Alchemy).

Also, when your core systems are easy to learn, you can add in some new, interesting subsystems to your game! (Mist examples: goddesses, magic circles.)

Perhaps the aim should not be to remove complexity, but to move it to places where it results in fun gameplay. Derakon mentioned Magic the Gathering. This is exactly what has been done in that game.

I do realize that I'm mostly saying what Derakon already said in my own words.

I agree with this outlook. It's where I've gone in my mind's progression also. However, as you note, first you need to simplify and then you can add. I also like added things to start appearing either after the game progresses more or as additional challenges/campaigns/variants.

[Mikko Lehtinen]

I'd also wager that simplicity for the sake of simplicity is also bad.

I like simple rules because they allow me to put in more tactically interesting features without confusing the player. As an example, a skill roll in Halls of Mist is always 1d100 under your score. That simplicity allowed me to add lots of content that utilizes that rule: the best example is Mist's terrain features that ask for various skill rolls all the time (Jumping, Spell Save, Perception, Alchemy).

Also, when your core systems are easy to learn, you can add in some new, interesting subsystems to your game! (Mist examples: goddesses, magic circles.)

Perhaps the aim should not be to remove complexity, but to move it to places where it results in fun gameplay. Derakon mentioned Magic the Gathering. This is exactly what has been done in that game.

I do realize that I'm mostly saying what Derakon already said in my own words.

[Magnate]

Simplicity is good. That is to say that complexity simply for the sake of complexity is bad. I'd also wager that simplicity for the sake of simplicity is also bad.

+1. Damn but you're good at that pithy soundbite thing. Ever considered politics?

My suggestion would be to, rather than trying to simplify things arbitrarily or by consensus, to simplify the things that would bring Vanilla back in line with what it was before. Re-simplify the elements that were at one time simple. Remove complexities that were (nearly) needlessly added.

Examples would require extra work, and others more familiar with hard-core old-school Angband could make the case better than I, should anyone care to.

This is very interesting, as your last point immediately made me think of Timo, and for all Timo's criticisms of modern V I don't think he's ever mentioned needless complexity. I'll be the first to admit that we made some mistakes in the 3.x versions, but it's hard to think of stuff we made more complicated for the player. It would be a fair criticism to say that we didn't do as much simplifying as we could have done, but that doesn't seem to be what you're saying.

[buzzkill]

Buzzkill: you know what also is good? Specific examples. Polemics against the current state of the game are all well and good as long as they properly support their thesis.

Examples would require extra work, and others more familiar with hard-core old-school Angband could make the case better than I, should anyone care to.

The cause of simplicity could, if left to it's own devices, only result in said simplicity rather than actual simplicity. However, if this cause has a definite predetermined bearing, then such tangents are more easily avoided.

Just sayin', feel free to ignore me. I almost deleted that first post just before I posted it.