Game News Torment Kickstarter Update #27: Pre-Holidays Report from the Torment Triumvirate

[Dr Schultz]

Roguey has a good point, even if the wording is troll-bait. Any probability check is gonna cause save-scumming, and that's pretty poor design. I also think adding a mechanic to reduce this chance of failure by spending stat points is making it even worse.

I think chance of failure is okay (but not good) for combat. There's a chance you miss and that makes sense. Combat is long, a lot of things happen, you can easily make up for any mistakes. Chance usually go both ways here as well -- you can get lucky and land a crit, for instance. It can add excitement.

But dialogue?
PC: [Charm - 86%... FAIL] Hey baby, how 'bout GAH FUCK INTESTINES GRAPEJUICE SHIT
NPC: Why I never! [leaves party]
Player: duh, okay, I'll keep on playing.
... or what?

I prefer a system where stats determine options, and those options are pre-determined, even if the options are failures. I mean, I'll see from the text whether or not a lie sounds good or not. If it looks like shit I'll try another option. Or fight. Or whatever.

Franky, the dialogue in your sample is bad design no matter how the ruleset works. Why on earth a non combat situation should be resolved by a single test with a binary outcome?
The problem is the encounter design, not the ruleset. Players should have more options, more possible outcomes, more agency. A flat value wouldn't change anything in this scenario.

 

[Grunker]

Here we go again. Classic Roguey: points out a legitimate issue, coats it in obvious bullshit and troll-paint, and the next 30 pages will be about that.

I like Infinitron 's suggestion, if we can't have the optimum solution (no rolled checks).

 

[hiver]

Its like that thing hiver was talking about... and then it was like that thing Adam was talking about... what, a month ago or more now?

not to mention this update itself. what are you people .. just reading the headlines and titles? or are you all mementoing around or dont bother reading at all?
you too, - journalist.

We chose Numenera for the setting, knowing that we'd be able to adapt the system as needed (because we talked to Monte about it and he was cool with that). That said, we've found the Numenera system does not need as much adaptation as we once thought it would. A lot of the goodness and funness of it is because it puts power in the player's hands, which obviously is totally in line with what we're thinking.

http://www.rpgcodex.net/forums/inde...-successor-thread.79051/page-258#post-2776068

In terms of special abilities unique to your type (Nano, Glaive, Jack), you can't do those things unless you actually earn the ability. So a Glaive can't cast esoteries no matter how much Effort he uses.

The idea in Numenera is that anybody can attempt a task -- like picking a lock, persuading a guard, or jumping across a chasm. The difficulty of the task is determined by the task itself, not by the character attempting it. So jumping across a 6' chasm is (to make up a number) a Difficulty 3 Task. That difficulty determines what target number you have to roll to succeed. If you are trained in jumping, it decreases the difficulty. If you apply Effort, it decreases it even more.

So a trained jumper with low Might might have the same chance of success as a very strong, untrained person who spends their Might on the task. Additionally, a trained jumper with high Might would not only have a higher chance of success, but would be capable of jumps that the other two builds couldn't even try (or they could, but they'd fail). And an untrained character with low Might could maybe tackle an easy jump (low Might doesn't mean no Might), but any serious jump would probably be too difficult even with the Effort he could apply.

So it doesn't eliminate choice at all, but it does suggest your stat pools (Might, Speed, Intellect) are slightly more important than what skills you train in. Monte is fond of saying that Tier 1 characters are actually quite powerful already, unlike level 1 characters in most games. Effort is part of that.

And this is just talking about generic "difficult tasks." As I said in the first paragraph, there are lots of things that will make your character unique beyond stats and skills. Your type, descriptor, and focus (mostly type and focus) give you bonuses and abilities unavailable elsewhere.

I should also add that we're still pre-prototype here. When we finally get something we can play around with, if we discover that this system -- while great for a PnP RPG -- sucks on a computer, we'll change it.

 

[undecaf]

I think chance of failure is okay (but not good) for combat.
...
But dialogue?

Just like there's a chance that you miss in combat, there's a chance that you miss a certain delicate tone while persuading someone, even if you sounded convincing at first.

That said, there should be mechanics for both, stats determining options and - where appropriate - chances (like haggling or persuasion that is pleading for the persons emotions - "[5%]Would you pick up the soap for me, please?" - and not so much for talking sense to someone - "No, you're wrong. E=mc2").

 

[Lord Andre]

When it comes to skill checks the most important thing to me is: are they difficult enough to be meaningful ? If I can pass every thing in the game by pushing the "effort button", then it's pointless.

Save-scumming a problem ? Don't fucking save-scum. Problem solved. <---- non-issue
Challenges are piss-easy across the board ? No solution. <---- issue

Conclusion: Roguey is not some master troll, just another mediocre monkey defending an idea that it decided arbitrarily at some point to be true.

 

[tuluse]

The problem is that dialog and skill checks traditionally don't have:

Partial successes
Second attempts with more information
Player input on the success rate of the check
Resources that can be spent to increase success rates, or even get 2nd opportunities


If InXile can add all these things then dice rolling would not be a problem.

 

[Zed]

I think chance of failure is okay (but not good) for combat.
...
But dialogue?

Just like there's a chance that you miss in combat, there's a chance that you miss a certain delicate tone while persuading someone, even if you sounded convincing at first.

That said, there should be mechanics for both, stats determining options and - where appropriate - chances (like haggling or persuasion that is pleading for the persons emotions - "[5%]Would you pick up the soap for me, please?" - and not so much for talking sense to someone - "No, you're wrong. E=mc2").

Haggling in this case would just be a skill check with flavor text whereas something like lying to someone would be a pre-determined string of text with an underlying mechanism of randomly failing.
I have no problems at all with skills determining what pre-defined choices are available (perhaps a better lie, or perhaps opening an option to charm or intimidate instead) as it is in most CRPGs with skill-checks in dialogues, but to have the choices be small individual games of chance is pretty stupid.

Save-scumming a problem ? Don't fucking save-scum. Problem solved. <---- non-issue
Challenges are piss-easy across the board ? No solution. <---- issue

Save-scumming behavior is a result of design. Designing around it does not impede challenging gameplay.

 

[Excidium]

The problem is that dialog and skill checks traditionally don't have:

Partial successes
Second attempts with more information
Player input on the success rate of the check
Resources that can be spent to increase success rates, or even get 2nd opportunities


If InXile can add all these things then dice rolling would not be a problem.

Exactly. Add to that a rolling model that isn't horribly random.

If it works at all like Numenera it's going to be twice as awful in CRPG form.

 

[Lord Andre]

If you want 100% success in something then specialize in that. Build your char so that you succeed even on the lowest roll. C&C.

Also, this obsession against save-scumming is like being a gay homophob. "I hate myself for wanting to suck dick and then I see all these other people sucking dick and enjoying it so I want a law that makes sucking dick illegal for everyone because I'm a mean cocksucker...oh wait."
No one is putting a gun to your head and telling you to punch an icecream truck for an hour and a half.

 

[Infinitron]

People who say they don't savescum are like men who say they don't masturbate.

I also think adding a mechanic to reduce this chance of failure by spending stat points is making it even worse.

I think this isn't necessarily true if the player knows his "stat points" will be regularly refilled - again, much like HP. You don't freak out over every hit because you know sooner or later you're going to rest.

 

[Zed]

People who say they don't savescum are like men who say they don't masturbate.

I also think adding a mechanic to reduce this chance of failure by spending stat points is making it even worse.

I think this isn't necessarily true if the player knows his "stat points" will be regularly refilled - again, much like HP. You don't freak out over every hit because you know sooner or later you're going to rest.

So you'll either end up in a save-scum scenario when you run out of stats, or you wait or get out of your way to get it refilled. Doesn't sound any better to me.

 

[Infinitron]

So you'll either end up in a save-scum scenario when you run out of stats, or you wait or get out of your way to get it refilled. Doesn't sound any better to me.

It's certainly not perfect, but the game is going to have an Effort mechanic one way or the other (it's an integral part of Numenera) so you better get used to it.

As far as refilling goes, remember that the game may have some kind of time pressure mechanic, so you may not be able to do that indefinitely.

 

[Zed]

No one is putting a gun to your head and telling you to punch an icecream truck for an hour and a half.

Nope, but it's something people will want to do because they are looking for a certain outcome that is within their reach.
People want to try until they succeed, if they know they can. It's probably the same mentality behind gamling.

The question is: where in a game do you want this endeavor of probable success to take place? I think most people don't mind it in combat. It's a long process, and you can act after the fact and still reach the same conclusion (you winning the fight).
Dialogue is an entirely different thing. "I know there's a chance I can lie to this guy and get entry to this room... Ooops, failed. Well, I don't want to fight the guy... Since I know I can pass him without a fight, I'll just load again."

 

[Lancehead]

As far as refilling goes, remember that the game may have some kind of time pressure mechanic, so you may not be able to do that indefinitely.

I thought time pressure was part of Crises. Don't think it'll affect Effort replenishing.

 

[Lord Andre]

Does anybody remember persuasion in VTMB ? No dice, you just had to match the difficulty number. 6 was good, 7 ok, 8,9,10 useless. There was nothing that required 10, one 9, two 8s and the reward was shit. You had no way of knowing, and it cost an arm and a leg to get there. Was that good design ?

Having a random based variable in the equation gives a tangible reward for maxing a skill, even if it is a meta reward e.g. less time spent reloading like a faggot. And if you can't get something in 2-3 tries you move on. What's the big deal ?

 

[Monad]

Save-scumming a problem ? Don't fucking save-scum. Problem solved. <---- non-issue
Challenges are piss-easy across the board ? No solution. <---- issue

Save-scumming behavior is a result of design. Designing around it does not impede challenging gameplay.

No it's not. It's a result of player behavior and choice. I don't think designing a game around trying to prevent players from reloading is a good thing.

 

[Excidium]

I think this isn't necessarily true if the player knows his "stat points" will be regularly refilled - again, much like HP. You don't freak out over every hit because you know sooner or later you're going to rest.

Oh yeah. Stats are replenished on rest, where you roll 1d6+level and then distribute the results as you see fit. Also first "rest" is instant (can even be taken in combat) and then increases each time I don't remember how many steps until the last one takes a whole day and it resets. More randomness and easily metagamed stuff for Roguey delight if they keep it same

 

[Lady_Error]

Eschalon I believe got around this pretty easily by generating a seed whenever you entered any map and saving it.

That is indeed a good solution, though personally I don't think save-scumming is such a big problem. Those who want to do it, should do it. Sometimes it can be fun.

 

[Azrael the cat]

There's a basic mechanical reason why %chances in dialogue is bad, in a way that doesn't apply to combat.

In combat, you're rarely in a situation where a non-trivial combat requires only 1-2 rolls. Abilities that create that possibility, like save-or-die spells, are doled out carefully, with severe restrictions on how often you can use them. Even if power word: Kill had only a 5% chance of working, letting a mage spam it like magic missile would be poor design.

Generally in combat you're facing a series of rolls - and with multiple characters on both sides of the combat, that series of rolls can get very large very quickly. So what you have is a situation like professional poker: sure, each event is chance-based, but the odds are knowable, fair and you can each manipulate them based on skill and knowledge. So over the course of a decent-length game the odds even out, and it becomes as chance-based as basketball or football - sure, sometimes you might just get lucky with a big play and cause an upset, but the best few players will make the final most of the time.

Rpg combat doesn't go quite that far, as that would require ultra-long combats, but it goes far enough that in a well-designed system you can't play the dice to beat an encounter that should be far too strong for you. There's a small window of difficulty where reloading might help, and a very very small chance that you might get lucky on a big play, but really in a party-based combat, save-scumming will only net you the opportunity to take back a bad move (which is exactly what it SHOULD allow you to do...and you should be allowed to do that without breaking the game, which is why a combat where win/lose is decided by a single roll, like 1st level characters in 2nd ed D&D, is crap).

But dialogue skills simply don't have that many rolls to make the chance modifier even out. Even in a major dialogue with really tough checks, you might have to pass 3-4 (at utter most) skill checks. More commonly you have one check, one diceroll, and it would look silly if you didn't (nothing to make tough dialogues stand out if every shopkeeper gets 4 shots at resisting your haggle skill). That makes it a completely different ball-game to dice-rolls in combat. It isn't poker anymore, it's walking into a casino and taking a single spin on the roullette wheel. Now obviously skill systems will modify the chance, so I don't mean that it's too random. I mean that the roll is too influential - you only need to savescum ONE roll and you've completely changed the outcome (where, the rare 'big play' aside, changing one roll in a party v party combat will only be decisive if things were already pretty close).

That brings in a very different savescumming problem. Instead of letting you take back dumb turns, it transforms the chance system into busywork like the minigames that now plague crpgs. Why do I have to play frogger to fucking unlock a door that I know I have the skill for - can't I just unlock the damn thing? That very easily chances to 'I know I've got the skill to pass this, why the fuck should I have to go through the chance system.'

 

[janjetina]

One must take a look at the task itself, i.e. its possible success and failure states. The best option would be having multiple 'partial success' states, as well as more nuanced outcomes that lead from the failure states, as well as other features mentioned by tuluse. Effort is already in.

Some tasks, like picking a lock, don't really have much to offer in terms of immediate outcome, but the designers should ideally take care that the immediate outcome does not lead into a dead end (taking an example of a lock, we can have 1 - picked it, 2 - failed to pick it, may retry, 3 - jammed it, need to find an alternative way in (e.g. use explosives, which will alert the guards, or search the level for a key and see if you can swipe it) , 4 - jammed it, alerted a guard by making noise, leading to an encounter which could be solved in a few possible ways, not only through combat).

In such a case, the effect of randomness would be ameliorated, because the failure would not lead to a dead end (of course, one must take care not to award player's stupidity). The obstacle is in finite resources, so the designers need to make a compromise between quality and quantity. Unfortunately, in cRPGs quantity prevails.

Having a deterministic step function check is a lazy hack job solution that feels artificial. However, people who are really adept at a certain task amost never fail at it, so purely probabilistic approach is not appropriate. It is appropriate for those characters who are partially qualified to perform the task at hand, so they may or may not succeed. Given that, my personal preference would be having the 0% success threshold skill level (always resulting in a failure, which may or may not be catastrophic, depending on the task), the 100% success threshold skill level, while with intermediate skill levels probabilistic approach would be used.

Of course, the character attempting the task (and by proxy the player) should have information on difficulty of the task (which need not be perfect - an adept would know that he would succeed at the task, but other characters should get the amount of information based on their skill level and attributes, like perception and wisdom (a low wisdom character "doesn't know that he doesn't know", but a player should suspect something if his character with 0 skill in disarming traps and low wisdom thinks he can easily disarm what looks like a complicated explosive device)).

 

[hexer]

If reloading is such a problem just add some sort of price for it ... gold, bottle caps of whatever the currency is. Then you'll think twice about reloading all the time.

 

[ZagorTeNej]

Roguey has a good point, even if the wording is troll-bait. Any probability check is gonna cause save-scumming, and that's pretty poor design. I also think adding a mechanic to reduce this chance of failure by spending stat points is making it even worse.

I think chance of failure is okay (but not good) for combat. There's a chance you miss and that makes sense. Combat is long, a lot of things happen, you can easily make up for any mistakes. Chance usually go both ways here as well -- you can get lucky and land a crit, for instance. It can add excitement.

But dialogue?
PC: [Charm - 86%... FAIL] Hey baby, how 'bout GAH FUCK INTESTINES GRAPEJUICE SHIT
NPC: Why I never! [leaves party]
Player: duh, okay, I'll keep on playing.
... or what?

I prefer a system where stats determine options, and those options are pre-determined, even if the options are failures. I mean, I'll see from the text whether or not a lie sounds good or not. If it looks like shit I'll try another option. Or fight. Or whatever.

Mostly agree but why you do feel chance of failure is not good (but merely OK) for combat? I think a certain amount of randomness in combat is always needed, it makes it more unpredictable and exciting (as you youself said) and besides, (almost) no one's gonna save and reload because you character missed once (or even several times in a row as long as you're hanging in there) and/or opponent hit you unless it's fatal for your char (which is very rarely the case in any CRPG) so it's quite different to say dialogue checks in that regard (which I agree have just one required value to pass it, you either meet the requirements or you don't).

 

[Dr Schultz]

The problem is that dialog and skill checks traditionally don't have:

Partial successes
Second attempts with more information
Player input on the success rate of the check
Resources that can be spent to increase success rates, or even get 2nd opportunities


If InXile can add all these things then dice rolling would not be a problem.

That's exactly what I wrote in a post that still await mod approval :D. A single skill test with a binary outcome is bad design no matter how the ruleset works. Non-combat gameplay shouldn't be a tax for your non-combat abilities (at least not in Torment). It should be real gameplay, like combat.

 

[Dr Schultz]

People who say they don't savescum are like men who say they don't masturbate.

I also think adding a mechanic to reduce this chance of failure by spending stat points is making it even worse.

I think this isn't necessarily true if the player knows his "stat points" will be regularly refilled - again, much like HP. You don't freak out over every hit because you know sooner or later you're going to rest.

Actually stat points ARE the "hit points" of Numenera. They track your health status in the game. You are dead when all your pools are depleted. So using effort is like trading HP for better % off success.

I think this isn't necessarily true if the player knows his "stat points" will be regularly refilled - again, much like HP. You don't freak out over every hit because you know sooner or later you're going to rest.

Oh yeah. Stats are replenished on rest, where you roll 1d6+level and then distribute the results as you see fit. Also first "rest" is instant (can even be taken in combat) and then increases each time I don't remember how many steps until the last one takes a whole day and it resets. More randomness and easily metagamed stuff for Roguey delight if they keep it same

The "rest" part of the effort system should be changed in the videogame, if you ask me.

 

[Dr Schultz]

A single skill test with a binary outcome is bad design no matter how the ruleset works.

But the Effort mechanic doesn't necessarily change anything about that - you either fail or you use Effort and succeed. Still binary. There are simple ways to achieve non-binary outcomes that don't require convoluted dice roll mechanics, like the low-intelligence dialogue options in the earlier Fallout games.

True. Effort doesn't change anything in this area. It just aims to give players more agency (and flexibility) when they attempt a test, and it works fine. The rest is just encounter design: 1 test with 2 outcomes is bad design, more tests with various degrees of success is good design.