Sil on Roguelike Radio

[Darren Grey]

Originally Posted by half

I'm very puzzled by this bit. I don't understand how these sentences fit together. It seems like you are in support of a system where every monster of a given type dies in about the same number of attacks, but then you say that adding variance to this just leads to 'the game of averages'. Isn't the no-variance case even more average all the time?

Nope, because the variance comes in how groups of enemies behave, or the enemies themselves behaving turn by turn. You already have this sort of variance in Sil, with monsters moving about and using certain abilities, and I'm sure a huge amount that I've yet to see. By having variance in the turn by turn damage you end up obscuring these longer term variations. Or at least making them less relevant, and harder to plan around.

Too much randomness obscures tactics, in my opinion. It can have advantages, I do not deny, but it's negative impact can outweigh that. To me it feels like a heavy glove around the hand, buffering all contact with the cold metal it grasps.

Also, you point to a creature that was no threat taking longer than average to kill, but couldn't it also take less time to kill, making it less boring? Doesn't it do that just as often?

Not exactly, because of how the mathematical spread works. You're just as likely to get an overpowering roll on your seventh hit as on your first. Simple example: consider an enemy with 4 HP, and you attack with a 1d3 weapon. Some quick fiddling in Excel shows that there is no chance of it being killed in 1 turn, 40% chance of it dying in 2 turns, 58% of 3 turns and 2% of 4 turns. Contrast with a flat 2 damage - 100% death in 2 turns. This is a very simplistic example, and larger spreads work differently, but I wanted to show that random damage system weigh towards longer time to kill an enemy with a set amount of HP, due to last blow overkill. In particular contrasted with a deterministic system where the player can optimise against overkill. And don't forget that the 2% figure shows up a fair amount across many battles, and gets overinflated in our human brains.

Of course there's all sorts of balance things tied into this as well, and with higher numbers it becomes less relevant. It's still an issue though, in particular with weak enemies now and then taking a long time to kill. What may be a rare chance of the rolls can be very frustrating when it happens. Sil is particularly vulnerable to this as it has a very low floor to damage, and the random armour makes even that floor uncertain.

I think that can great, but I didn't want a system like that in Sil, and indeed very few medium or large roguelikes play that way. Most have pen and paper RPG style combat systems like Sil does. Most don't have loads of interesting tactical positioning abilities like Sil though. I don't see what Sil is doing worse than average here.

Does not a flaw in a gem stand out more ugly than a flaw in glass?

Of course I understand if it's not something you want. This comes down heavily to personal opinion. I have blogged before about my great distaste towards such random mechanics. Sil's combat is perhaps the worst I've seen for this pet peeve of mine in how the modifiers effects the dice rather than the end result, and the armour is random each turn rather than flat. Most roguelikes have modifiers that give +1 to overall damage instead of 50% chance of +1, which has the effect of providing more guarantees in combat. In Sil it seems the whole combat damage system is set up to maximise the thing I hate!

Especially if compared with a game that I know you love: ADOM, which has a more complex, more opaque, and probably less well designed combat system than Sil.

Heh, I know ADOM isn't great. It is a lot more reliable in combat than Sil, with many of those modifiers getting simplified immensely during play - in particular the damage dice on weapons get heavily outweighed by flat modifiers. Combat is mostly very predictable. But it still has a lot of bad design in combat, and in particular a lot of overly detailed systems that end up having very minor effects on play. If I praise ADOM it's because I've been playing it for many years and it's hard to forget your first love. You might feel the same about Angband. In contrast I've been scathing towards ADOM II for carrying on with this convoluted system (expanding it to more complex areas, even). I think a modern roguelike should be held to higher standards. We shouldn't just copy the imperfect designs of the past. Sil is excellent in this regards in many areas.

A better example for a large game with a complex but deterministic combat system is ToME4. It may not be too obvious from an outside look since it's heavy on stats, but when playing you can very reliably know your damage range against any enemy. There is very little numerical randomness most of the time. You can plan moves ahead with reliability, which is important with its cooldown system. There are still surprises in enemy AI, and risks you can take, but in general you feel much more in control of what's going on. Importantly this means that when you die you don't have an RNG to blame, just your own choice of actions.

[bisonbison]

A simplified combat results window might start with something like:

Turn 1034:
- @ x T{1}: hit, 0 dmg
- T{1} x @: missed
- T{2} x @: hit, 7 dmg

Turn 1033:
- @ x T{3}: hit, 12 dmg, T{3} was killed!
- T{1} x @: hit, 3 dmg
- T{2} x @: missed

Which could be helpfully supplemented by a monster window that actually listed individual monsters with health and morale levels:

C{1} | ** | 2 squares away | fleeing
C{2} | ******* | adjacent | confident
w{1} | ******* | 10 squares away | sleeping


You could then turn on a verbose combat mode that listed the die rolls for to-hit and dmg.

[jdh]

Originally Posted by bisonbison

A simplified combat results window might start with something like:

Turn 1034:
- @ x T{1}: hit, 0 dmg
- T{1} x @: missed
- T{2} x @: hit, 7 dmg

Turn 1033:
- @ x T{3}: hit, 12 dmg, T{3} was killed!
- T{1} x @: hit, 3 dmg
- T{2} x @: missed

Isn't this what the interface outside the combat rolls window already shows you (though the damage done only flashes up temporarily)? I only find myself looking at the rolls when something is going wrong, mostly to find out whether I've just been unlucky or if I have little or no chance to hit and do damage.

[Psi]

I will throw out my views, but I suspect it is fairly futile as we have a fundamental difference of opinion!

Please excuse my heavy snipping...

Originally Posted by Darren Grey

By having variance in the turn by turn damage you end up obscuring these longer term variations... Too much randomness obscures tactics, in my opinion... Contrast with a flat 2 damage - 100% death in 2 turns... In particular contrasted with a deterministic system where the player can optimise against overkill. And don't forget that the 2% figure shows up a fair amount across many battles, and gets overinflated in our human brains... It's still an issue though, in particular with weak enemies now and then taking a long time to kill... What may be a rare chance of the rolls can be very frustrating when it happens...

I am not sure I could disagree more. One of the really refreshing things about Sil is the fact that it is unpredictable and dangerous and you have to react to things not going your way. I would hate knowing that a monster was guaranteed dead in however many turns. The very mantra of roguelikes is randomness and with that comes risk and reward and this is what makes playing exciting. Sure it hurts when you the rolls go against you, but was the reward actually worth the risk you took? Given that Sil is a short game, it does not actually take long to get back to where you were anyway.

Originally Posted by Darren Grey

This comes down heavily to personal opinion. I have blogged before about my great distaste towards such random mechanics. Sil's combat is perhaps the worst I've seen for this pet peeve of mine in how the modifiers effects the dice rather than the end result, and the armour is random each turn rather than flat. Most roguelikes have modifiers that give +1 to overall damage instead of 50% chance of +1, which has the effect of providing more guarantees in combat. In Sil it seems the whole combat damage system is set up to maximise the thing I hate!

As you say, this is personal opinion - and again I would go the other way and say that Sil has the best combat mechanics I have seen. I have always favoured O-style combat for the very reason that +1 modifiers to damage are just too powerful in V. As I am sure you are aware, in O-style combat, the modifiers add a percentage to the base damage dice of the weapon which is in effect similar to what happens in Sil.

[Mikko Lehtinen]

Darren,

you seem to want two separate things that overlap a little:

• quick combat analysis
• deterministic mechanics

Quick combat analysis

I can very much symphasize with your desire to analyze the battlefield with a glance. That's the reason why I'm planning to start using only non-opposed percentage skill rolls in a future version of Halls of Mist. (Non-combat skills already work like that.) You will instantly know what your chances to hit are, and can concentrate on smart tactics instead of calculating probabilities. Monsters have a binary armour stat. If they're armored you use Piercing skill to hit with a sword, otherwise you use Slash. Of course I'm sacrificing lots of nuance to be able to do this, and it's clearly not for every game.

Reducing the number of random rolls does not necessarily lead to deterministic combat mechanics. You would probably be happy if either hit chance or damage was randomized, but not both? This I can get behind with. The combat would still have almost as much randomness in it, but the situation would be quicker to analyze. Having randomness in too many places muddles up the thought processes.

I feel that if your game is mostly about fighting against diverse groups, quick combat analysis is key. If your game is mostly about fighting against single opponents or homogenic groups (like Sil?), making each and every swing of the sword interesting is more important.


Determinism

By a coincidence, my favorite turn-based strategy game at the moment, Domination (sequel to Massive Attack [edit: should be Massive Assault! I always mix the game with the trip-hop band]) has completely deterministic combat mechanics. A tank always deals two points of damage, and five points is enough to destroy it.

This is not very popular in turn-based strategy games, maybe because it might lead to analysis paralysis. If you can calculate the result of a battle beforehand, many players feel that they must do exactly that. Complicated calculations with multiple units take time and effort, and are only fun for a minority of players.

Domination manages to work very well because it has an Undo command! You can easily try many possible tactics before you commit to one. That's faster and more fun than calculating by hand.

OK, so I wouldn't enjoy deterministic wargames unless they are as slick as Domination. But roguelikes don't usually have much board complexity -- they're one against many -- so completely deterministic mechanics would probably work much better than in wargames.

BTW, do you feel that critical hits are as boring as random damage rolls? Let's say you had a 15% chance to deal triple damage.

[LostTemplar]

You would probably be happy if either hit chance or damage was randomized, but not both? This I can get behind with.

Using random once is enough IMHO. Just hit / miss or a simple damage random. Probably what I would like is a damge roll as a normal distribution with some mean and deviation, determined by stats (both attacker and target in some simple realistic way), where negative damage values are replaced with zero and called 'miss'.

[Darren Grey]

Originally Posted by Psi

The very mantra of roguelikes is randomness and with that comes risk and reward and this is what makes playing exciting. Sure it hurts when you the rolls go against you, but was the reward actually worth the risk you took? Given that Sil is a short game, it does not actually take long to get back to where you were anyway.

It's long enough! And I disagree that the mantra of roguelikes is randomness. This isn't Snakes and Ladders. I've always been drawn to the chess-like feel of roguelikes, planning actions turn by turn. Having to deal with things not going to plan is of course part of the fun, but too much randomness removes the ability to plan at all. It leaves you in a position of trying to have your character set up to overwhelm the randomness. Which can be fun on its own, but it can sacrifice the tactical depth.

Have you played MicRogue or Zaga-33? Great examples of roguelikes with purely deterministic mechanics. They still have random enemy placement, and enough complexity of monster movements that you can never plan far ahead too perfectly. They're short too, so you can get in and learn the systems quickly. Have a try and see how they feel to you.

I'm not saying Sil should be like that, mind. A bit of randomness I'm perfectly fine with. But it should be there to provide variation and unpredictability on top of the base tactical play, not something that overshadows the whole system.

Originally Posted by Mikko Lehtinen

you seem to want two separate things that overlap a little:

• quick combat analysis
• deterministic mechanics

Yup, the two have some very serious areas of overlap Random mechanics reduce the value of combat analysis. And that sort of analysis is precisely what I enjoy! Sil allows a lot of analysis in many areas thanks to its open mechanics, but this gameplay element is smothered in battle by the dice rolls.

Monsters have a binary armour stat. If they're armored you use Piercing skill to hit with a sword, otherwise you use Slash.

Nice idea! I'm not sure I've seen such a binary system for armour before, but it makes a lot of sense from a game design perspective.

Reducing the number of random rolls does not necessarily lead to deterministic combat mechanics. You would probably be happy if either hit chance or damage was randomized, but not both?

Yes, it's mostly the damage rolls I'm moaning about. The to hit mechanics are much steadier from battle to battle, and it's possible to have modifiers that weigh them heavily in your favour - or, indeed, in the enemy's favour, making some enemies simply unwise to engage without good investment in accuracy. I think the to hit system could do with a bit of clarity added (a bit like with my Stealth suggestions), but the fundamental dice mechanics involved I'm not hugely opposed to.

BTW, do you feel that critical hits are as boring as random damage rolls? Let's say you had a 15% chance to deal triple damage.

Well, when the enemy has that it can be rather terrifying But as a player power it's not something I mind. Though I think it's a little pointless if you can't game it properly, whilst being potentially overpowered if you can

In general a small chance of a random bonus to a player is not a bad mechanic - it's not something they can rely on, but it's not something they have to constantly plan against either. When it happens it feels like a nice reward that gives them a warm fuzzy feeling inside. But if an enemy can do crits then you need to always face them with at least triple their normal attack damage left in health, otherwise the risk is too high of a killer blow. Of course there can be balance to this - if a crit only does +50% damage then it doesn't have as big an impact.

In some of my games I've had the more powerful player abilities rely on randomness much more. On average they do much more damage than the regular attacks, but the inherent unreliability involved reduces their value somewhat. Giving the player the choice between a random system and a deterministic system is quite fun Usually these chaotic systems come with some special player risk though - part of my philosophy that power comes at a price and he who plays with chance leaves himself at the mercy of the dice.

[Mikko Lehtinen]

Originally Posted by Darren Grey

Random mechanics reduce the value of combat analysis. And that sort of analysis is precisely what I enjoy!

We both clearly enjoy having all the relevant information at hand, without needing to calculate anything. But I like to make risk management decisions based on that information; you seem to like puzzle-like combats where it's possible to find a "correct" answer.

[Mikko Lehtinen]

Perhaps one thing that makes risk management more boring in roguelikes than in wargames is if you have only one thing that you can risk: your life.

If losing a combat had less fatal but still permanent consequences, risk management would be more fun.

[Nick]

Originally Posted by Scatha

At least we really enjoy the risk-management game: is it worth taking a small chance of disaster to kill the wounded giant/steal the sword from the dragon's hoard?

Originally Posted by Darren Grey

Nope, never! This is a permadeath game that takes several hours to complete. You can't take such risks. Get to a position of overwhelming power or minimised risk and then proceed.

This exchange is the nub of the thread for me.

To me, Sil is a *short* game. I have frequently played the same game of an Angband variant for a couple of hours a day for days or weeks - here's an example. I died from taking a risk, and I don't regret it.

If that means I'm crazy, then I don't want to be sane.