No more GM sucker punches, and the gameplay challenges thereof

[Endrosz]

As I say you can't really make a smart AI, so in order to make him challenging you have to give him an unfair advantage.

 

You CAN make the AI smart, but:

-- It takes a long time, many iterations. You can only start developing the after all of the systems are finalized (if you don't to want to waste man-hours), that means late in the development cycle.

-- That developer time is also costly, and that money is usually spent elsewhere, see below.

-- The result will be appreciated by only a small portion of the players. In the case of an RPG, everyone cries out for more companions, more wilderness areas, more side quests, more quest solutions, more branching choices, more skill and spells, but very few people ask for a good/better AI. See this forum as a direct evidence.

-- As a consequence of all of the above, few people in the game software industry have expertise with adaptive neural nets, fuzzy logic, emergent small entity AI, and so on, because there's little demand for these skills in this industry segment. But at least one of the former is required to write a decent AI FRAMEWORK. It's the framework that counts, what is possible within it; "teaching" the AI the specifics of a game simply takes time, it's not that difficult. I have studied neural nets and fuzzy logic during my university, later I've created AI "brains" for strategy games for several years, and of this is a fascinating subject that few people bother with.

 

If you want a game with good AI, try the Galactic Civilizations series. The AI there is ruthless, efficient, and diplomatic negotiations in GalCiv is almost like talking to an intelligent opponent, who has agendas, knows what's worth what, makes counteroffers (very refreshing!) etc.

Edited April 5, 2014 by Endrosz

[IndiraLightfoot]

I've always wanted to try out Galactic Civilizations, and since a new instalment is coming up soon, you have just given me one more good reason to try it out.

 

I must say, though, as someone used to Fighting Fantasy-books and early official D&D scenarios, where sucker punches are fairly common, that I somehow sadistically enjoy them a lot - of course, as long as replaying is allowed. It could easily get to be very abrupt endings in all those adventures, so I guess "save-scumming" has been there all along, even before CRPGs, and I'm oddly enough used to that meta-gamey feature. Regardless, those almost unfair and leftfield scenarios and scenes are often enjoyable in their own cruel right. Just my two cents.

[Sarex]

I love save-scumming, I have at least 10 save files in every game I play. In DA:O I just spammed new saves, had around 150 of them by the end.

[Chippy]

RTS example: Shogun 2 had the ai create 20 stack armies out of thin air because the developer decided the ai could never beat the player.  They "improved" on that (so I'm told in Rome: Total War, but I experienced it in King Arthur 2) by introducing a capture the flag scenario; so I've just beaten an army outnumbering me 2:1. using great tactics and skill and am just about to kill their general, but because they captured the flags, it's game over.

[Yonjuro]

 

Exactly; this is a completely hypothetical question. So, in my other post, I posed the question, if this encounter is not a sucker punch, what would turn it into one?

 

 For example, putting this area in BG1, say, in Candlekeep in place of Shank or Carbos (the two very lame assassins who try to kill you in the tutorial level) would qualify, but maybe you can think of a less ridiculous example that you would consider a sucker punch.

 

To be honest it would have to be something ridiculously over the top......

 

 

 Well, how about this:

 

 

 

...

 

I don't think that would be possible, but to answer your hypothetical, no it wouldn't I would just need to find another way to kill them. I would probably use spell turning+spell shield+spell immunity and set most of my characters to ranged attack.

 

 Suppose the beholder area had a mutant beholder with an eye of dispelling. Spell shield protects you from beholders because their magic removal eye doesn't work like Dispel Magic (or Remove Magic). Would that modification be ridiculously over the top? Is it a sucker punch?  

 

 Or, simpler, if the game hadn't provided you with a spell shield scroll before meeting the beholders. That's six seconds to no protections at all no matter what buffs you did.

 

 

What people don't seem to understand is that games have to be unfair to be challenging, AI's are dumb and I don't see them getting smarter anytime soon.

 

 Well, hang on. BG2 is carefully designed to be winnable. If you want unfair for real, it would be very easy to mod BG2 to make it unwinnable, and the mod wouldn't have to be ridiculous (where ridiculous would be something like replacing the mephit portals in the intro level with beholder, mindflayer, vampire and lich portals (recall that Imoen has spells up to level 4 at the beginning of the game)). It could do things like the above - give beholders real dispelling and similar creature mods; remove some scrolls and potions etc.

 

 Thinking of a minimal mod that would make BG2 impossible is left as a an exercise for the reader.

[teknoman2]

RTS example: Shogun 2 had the ai create 20 stack armies out of thin air because the developer decided the ai could never beat the player.  They "improved" on that (so I'm told in Rome: Total War, but I experienced it in King Arthur 2) by introducing a capture the flag scenario; so I've just beaten an army outnumbering me 2:1. using great tactics and skill and am just about to kill their general, but because they captured the flags, it's game over.

if you want an RTS example take Empire Earth. in custom maps, since they did not have an AI capable of managing the opponents, they simply gave them cheat codes. every 5 minutes the AI was using the cheat code for +10000 of all resources. it had no need for workers and could spend all population slots for army units, meaning that your base was under constant attack, while you had to have 60% of your population in workers to keep up with the resources

[Labadal]

I'm not really a fan of some fights feeling like a suckerpunch. Fights can be fun and challenging without having to get frustrating. NWN2 and its expansion packs were guilty in some encounters. The same goes for Torment. It doesn't make for fun fights, I didn't feel satisfaction when winning those battles. All I could think was: Finally, this is over.

[Sarex]

 Suppose the beholder area had a mutant beholder with an eye of dispelling. Spell shield protects you from beholders because their magic removal eye doesn't work like Dispel Magic (or Remove Magic). Would that modification be ridiculously over the top? Is it a sucker punch?  

 

 Or, simpler, if the game hadn't provided you with a spell shield scroll before meeting the beholders. That's six seconds to no protections at all no matter what buffs you did.

 

 

That would just be ramping up in difficulty of the fight. Would it be over the top? I don't know as I haven't played that encounter.

 

That would just be bad game design, but the facts are that beholders follow the rules of the game, you are given the tools (spells, summing, the shield of Baldurian), so it is very possible to beat the encounter in many different ways.

 

 Well, hang on. BG2 is carefully designed to be winnable. If you want unfair for real, it would be very easy to mod BG2 to make it unwinnable, and the mod wouldn't have to be ridiculous (where ridiculous would be something like replacing the mephit portals in the intro level with beholder, mindflayer, vampire and lich portals (recall that Imoen has spells up to level 4 at the beginning of the game)). It could do things like the above - give beholders real dispelling and similar creature mods; remove some scrolls and potions etc.

 

 Thinking of a minimal mod that would make BG2 impossible is left as a an exercise for the reader.

 

Unfair doesn't mean unwinnable, unfair means that the rules that apply to you don't apply to the AI. I made some examples in the above posts.

 

My last opinion on sucker punches is this, it's hard to do them when you can save the game where ever you want. In the example of the 2 Beholders, the game auto-saves for you right before that encounter, so even if they surprised you, you will lose nothing from it.

[J.E. Sawyer]

On this general topic, I recently picked up 13th Age and was pleasantly surprised that a lot of the base d20 modifications in their game are similar to the sorts of things we were doing in Pillars of Eternity. The mechanics are not always the same, but some of the goals seem to be.  The way they handle paralysis/petrify/death is pretty well-suited to tabletop turn-based play.  When the character gets hit with one of the "you're out" effects, they get to take one more action before they're effectively barred from performing more actions.  However, on their turns, they continue rolling saves.  It's only after they've failed a series of saves over many rounds that the effect becomes permanent (or at least for the duration of combat in the case of paralysis).

 

The potential danger is similar and the biggest penalty (the character being removed from combat) is still present, but it feels less like a sucker punch because a) the victim can still react to it (once) after being hit b) the players still perform an action on their turns (roll a save, as with ongoing effects in 4E) and c) other characters can also intervene before the victim passes the point of no return.

 

Surprise! Death saves I think usually feel like Jedi Outcast snipers.  They don't test your gaming abilities as much as they test your patience, since a lot of your success depends on trial and error followed by metagaming hard counters to learned threats after a reload.

[Hormalakh]

wow...i can go for that Josh. That's a good way of doing it without having to get rid of "OP" spells like flesh to stone, disintigrate, PW:Kill, etc.

[Yonjuro]

 

 Suppose the beholder area had a mutant beholder with an eye of dispelling. Spell shield protects you from beholders because their magic removal eye doesn't work like Dispel Magic (or Remove Magic). Would that modification be ridiculously over the top? Is it a sucker punch?  

 

....

 

 

That would just be ramping up in difficulty of the fight. Would it be over the top? I don't know as I haven't played that encounter.

 

That would just be bad game design, but the facts are that beholders follow the rules of the game, you are given the tools (spells, summing, the shield of Baldurian), so it is very possible to beat the encounter in many different ways.

 

 Agreed, but I had a point with these specific examples. The point of the Spell Shield example is that Spell Shield might as well be called Beholder Begone. The game has provided you with a  specific important tool.

 

 As you mentioned, there are other tools too, but at some point an encounter becomes what Mr. Sawyer referred to as a sucker punch in a recent interview. Reasonable people can certainly disagree on when that is (I'm not convinced that the Beholder encounter we've been talking about qualifies though I can see why somebody would think so), but more on that at the end of this post.

 

 

 Well, hang on. BG2 is carefully designed to be winnable.....

 

Unfair doesn't mean unwinnable, unfair means that the rules that apply to you don't apply to the AI. I made some examples in the above posts.

 

 I see. I thought you were saying something else.

 

 

...

My last opinion on sucker punches is this, it's hard to do them when you can save the game where ever you want. In the example of the 2 Beholders, the game auto-saves for you right before that encounter, so even if they surprised you, you will lose nothing from it.

 

 That's certainly a reasonable view. With auto-save, the cost of losing that encounter is minimal. My view is different; I always try for a no reload game because reloading makes the game less fun for me.

 

 I don't always succeed. E.g., I certainly reloaded that time when I stumbled into the twisted rune with a low level party in my first playthrough.

 

 If an encounter requires you to fail once to get needed info and then reload and destroy that makes it unsatisfying (for me and for the other no reload types). I would prefer an encounter that is challenging but possible to beat (maybe with some good luck but without the hindsight of reloading) the first time and that continues to be challenging afterwards. 

[Lephys]

 

You should know enough to think "okay, this thing's probably going to have some really damaging breath attack that's going to easily hit a lot of people, and I'm going to need to be ready to figure out how that works and avoid it or mitigate it somehow."

IT's A DRAGON!

 

That's like saying 'You should totally warn people that necromancer has undead followers!, 'you should totally warn people a giant will do massive damage' or 'you should totally warn people a banana is yellow!'

 

Okay, clearly a misunderstanding here. Totally valid, because, in just text, the meaning could go either way.

 

All I meant was "you, the player, should know that, because it's a dragon." Not "You should have that information because the game forcibly crammed it into your head somewhere."

 

I'm 99.9% sure you read it as the 2nd one. And yeah, if I meant that, I'd be crazy.

 

Granted, even if a dragon wasn't such a common creature in games and lore, something like that should still come up in the game world, because it should be common enough knowledge. Some town got attacked by a dragon, and it burned to the ground. Clearly fire is involved when dragons are. Someone doesn't have to explain that to you or anything, and you don't really have to go out of your way to learn it.

 

Anywho, I also just wanted to point out that Josh kind of hit another nail on the head regarding suckerpunches: the ability to react to it in a significant way. If it's just "Oh, you weren't prepared enough to prevent this all together, so it's happened, and that's that... you're all dead," that just sucks. Because, what are you going to do? Try again. Or never play the game again. One or the other. But, if it's "Okay, this is really, really bad at the moment, but you can probably slowly climb your way out of this metaphorical pit and salvage the situation if you're clever and adaptive, even if you didn't know about this before," then it's STILL not easy. The creature's still a menace, and there's still a chance you'll become horribly dead-ed. It's just not "rock beats your scissors, RESTART!"

[Volourn]

Adventuring, by its nature, is dangerous. The instant deaths and other nasty effects exist to illustrate this fact.  Fighitng a beholder or other deadlyc reatures who have such powers isn't supposed to be easy, welcomed, or cool. It is supposed to put fear into the player.

 

Sawyer's example seems like a neat way to do things but  it does make things less dangerous hence less challenging. When you are exploring the CAVE OF DOOM you gotta expect some doom. Having  multiple  second chances doesn't make it more fun or rewarding to survive a beholder. Heck, in the example, it turns  flesh to stone into  a simple parlysis/stop spell that may  turn more 'permenant' a half dozen rounds later.

 

 

"Clearly fire is involved when dragons are."

 

Of course, not all dragons breath fire.

[Sabotin]

You are going to have multiple second chances however you turn it, the question is how to best implement that to keep players concerned with gaming and not metagaming.

[Labadal]

I think having second chances in Pillars of Eternity for death spells is a good move. Why? Because as I understand it, there are no resurrection spells or items in the universe (yet).

 

In the IE games, if you got hit by an instant death spell, it was always possible to resurrect fallen heroes.

 

I'm not sure second chances should be applied for every status ailment. Maybe some ailments go away after a set amount of time? I don't mind if they are permanent as long as the player gets some means of cure, either using spells or items.

 

Something like The Flesh to Stone status should be permanent until the player manually (or an AI controlled party member) cures that status with a spell or item.

[J.E. Sawyer]

Adventuring, by its nature, is dangerous. The instant deaths and other nasty effects exist to illustrate this fact.  Fighitng a beholder or other deadlyc reatures who have such powers isn't supposed to be easy, welcomed, or cool. It is supposed to put fear into the player.

 

They do in tabletop.  In a game where you can reload, such as ours (excluding Trial of Iron), they don't create fear as much as annoyance.  For many players, getting sacked by a Disintegrate effect simply means a reload and a metagame revision of their prepped defenses.  In the second attempt, what would in tabletop induce fear isn't even an annoyance; it's simply obviated by a hard counter.  BG's Protection from Petrification is a 2nd level wizard spell and grants blanket immunity.  If you have a wizard and have that spell prepped, BG's fights against basilisks are trivially easy.  If you don't, they become as swingy as most low-level BG encounters.  Challenging in the sense that there's a good chance you won't get through them, but not particularly interesting since the individual die rolls have such a huge effect on the outcome.

[Sarex]

They do in tabletop.  In a game where you can reload, such as ours (excluding Trial of Iron), they don't create fear as much as annoyance.  For many players, getting sacked by a Disintegrate effect simply means a reload and a metagame revision of their prepped defenses.  In the second attempt, what would in tabletop induce fear isn't even an annoyance; it's simply obviated by a hard counter.  BG's Protection from Petrification is a 2nd level wizard spell and grants blanket immunity.  If you have a wizard and have that spell prepped, BG's fights against basilisks are trivially easy.  If you don't, they become as swingy as most low-level BG encounters.  Challenging in the sense that there's a good chance you won't get through them, but not particularly interesting since the individual die rolls have such a huge effect on the outcome.

 

But then isn't it like that for anything else in the game, you can reload if you fail or are caught unprepared. Where do you draw the line at what is annoying and what isn't, as this is a very subjective thing. If you step in to a deadly trap you simply reload and avoid it, if the enemy does a certain combo and gets you, you reload and plan for it. I am not sure I follow how that is different from anything else, is the problem that you can't outplay it or what?

[Hassat Hunter]

Likewise... a non-punch combat encounter that lasts 10 minutes but gets me in the end is sure to annoy me more than any suckerpunch ever can.

Having to revert 30 minutes due to autosaves will annoy me more than any suckerpunch (fortunately, there's no such thing here).

Having to watch an unskippable animation after dying before I can get in will annoy me more than any suckerpunch.

 

Like Sarex, what's the line?

[J.E. Sawyer]

But then isn't it like that for anything else in the game, you can reload if you fail or are caught unprepared. Where do you draw the line at what is annoying and what isn't, as this is a very subjective thing. If you step in to a deadly trap you simply reload and avoid it, if the enemy does a certain combo and gets you, you reload and plan for it. I am not sure I follow how that is different from anything else, is the problem that you can't outplay it or what?

 

It is not like many things in the game.  E.g. hit point loss is typically arrived at via the comparatively slow aggregation of rolled hits that normalize over the course of many attacks.  A vulnerable character standing in a bad position is a tactical problem for which there are many potential solutions that can be arrived at over a longer timeline.  This is why many A/D&D players often say that 4th-12th level A/D&D is the "sweet spot" for challenge in encounters.  From 1st to 3rd level, combat is extremely swingy.  The player's tactical decisions have less influence on life and death than the die roll.  For practical purposes, a wide variety of standard attacks at low levels are death attacks (or at least incapacitating attacks) for many characters.  In the mid- to upper-teens, A/D&D combat is dominated by attacking an enemy's weakest defense with an incapacitating status effect or overwhelming damage hit (e.g. in 3.X, a fighter's Will, which is notoriously fragile).  The best defense?  Hard counters vs. specific status effects or damage types, which are extremely difficult to plan for.  Combat scenarios that build around hard counters turn into puzzles for which there are not good solutions if you don't have the hard counters prepped (assuming you have access to that hard counter at all).

 

Where I draw the line are single-check long-duration incapacitation effects against which hard counters are the only practical defense.  Their effects on combat and the immediate aftermath promote reloading and metagaming more than in-the-moment adaptation to temporary or progressive incapacitation.  Yeah, you can lose a fight via progressive aggregation of damage over 30 seconds to 2 minutes.  This is a different type of problem with a more diverse variety of solutions than a fight you lose (or win a Pyrrhic victory) in 12 seconds because a basilisk waxed two party members as soon as you came around the corner.

[Sarex]

It is not like many things in the game.  E.g. hit point loss is typically arrived at via the comparatively slow aggregation of rolled hits that normalize over the course of many attacks.  A vulnerable character standing in a bad position is a tactical problem for which there are many potential solutions that can be arrived at over a longer timeline.  This is why many A/D&D players often say that 4th-12th level A/D&D is the "sweet spot" for challenge in encounters.  From 1st to 3rd level, combat is extremely swingy.  The player's tactical decisions have less influence on life and death than the die roll.  For practical purposes, a wide variety of standard attacks at low levels are death attacks (or at least incapacitating attacks) for many characters.  In the mid- to upper-teens, A/D&D combat is dominated by attacking an enemy's weakest defense with an incapacitating status effect or overwhelming damage hit (e.g. in 3.X, a fighter's Will, which is notoriously fragile).  The best defense?  Hard counters vs. specific status effects or damage types, which are extremely difficult to plan for.  Combat scenarios that build around hard counters turn into puzzles for which there are not good solutions if you don't have the hard counters prepped (assuming you have access to that hard counter at all).

 

Where I draw the line are single-check long-duration incapacitation effects against which hard counters are the only practical defense.  Their effects on combat and the immediate aftermath promote reloading and metagaming more than in-the-moment adaptation to temporary or progressive incapacitation.  Yeah, you can lose a fight via progressive aggregation of damage over 30 seconds to 2 minutes.  This is a different type of problem with a more diverse variety of solutions than a fight you lose (or win a Pyrrhic victory) in 12 seconds because a basilisk waxed two party members as soon as you came around the corner.

 

Then why not simply dial it back using difficulty settings. If you introduce permanent death in the upper difficulties of the game, then why not also scale those hard-counter spells with difficulty.

 

I get that you don't like metagaming, but to be honest some(a lot?) of us do. As you may have read in this and other threads, some of those hard-counter/metagame fights were the most memorable and exiting for us. They were a challenge to beat and prove to yourself that you could beat the hardest thing the game throws at you. In fact the best fights were completely optional.

 

I really think that there is room for both and that there is no reason to make metagaming a bad word.

[Hormalakh]

I think another aspect of this is also the risk/reward analysis that occurs in these games. The party members are extremely "expensive" aspects of the game: if a member dies a good chunk of gameplay is lost completely if they are not "renewable" - you can't bring back a teammate who has leveled up to where he is. Basically, at that point, there is no reason to continue the game because it's impossible to bring another team member to that level quick enough. Contrast this with games where your "team" is renewable. The old-school XCOM game for example, even if you lose one guy, it doesn't take that long to bring in another one to that level (unless of course he's a high-level offier, at which point people will reload). Or RTS games, where individual units are expendable.

 

I think there are two ways to sort of "beat the player's system" in these sort of risk/reward analyses that cause players to quickload/quicksave.

 

One is to make consequences occur later than the player has time to react to. For example, imagine a "witch's curse" spell that doesn't work immediately and has some sort of randomization for whether it works or not (so that the player isn't sure if s/he's really cursed), but it starts working during the next battle. The player then has to make a decision: should I quickload and try to completely avoid all curses or do I risk the quicksave and hope that the curse didn't affect me. This sort of late-consequence can help make "big-hit" abilities more difficult to avoid because players can't see immediate results and have to decide based on incomplete information - a much more scary thought for "save-scummers."

 

The other is to reduce the "risk" aspect of the equation. If your character dies, that's fine. You can always hire another one at the adventurer's hall with similar levels, stats. Just pick up the equipment, and travel back, then hire one (with a cost that's high, but doesn't pass the threshold of "too expensive"). It again doesn't bring back the "scariness" of table-top, but at least it makes the battles less likely to be "annoying" because the "annoying" factor of reloading is equivalent to the party getting another adventurer. Obviously, this isn't true for NPCs like Aloth, Eder, etc.

 

Then why not simply dial it back using difficulty settings. If you introduce permanent death in the upper difficulties of the game, then why not also scale those hard-counter spells with difficulty.

 

I get that you don't like metagaming, but to be honest some(a lot?) of us do. As you may have read in this and other threads, some of those hard-counter/metagame fights were the most memorable and exiting for us. They were a challenge to beat and prove to yourself that you could beat the hardest thing the game throws at you. In fact the best fights were completely optional.

 

I really think that there is room for both and that there is no reason to make metagaming a bad word.

 

 

I think that the issue I have with meta-gaming, is that it takes you out of the "mood" of the game. In a lot of other game genres, that "excitement and fear" of playing occurs during the actual game-play. From minecraft to MOBA's to other games that don't allow me to "meta-game", that aspect of "losing what I've worked for" is what makes it scary and ultimately enjoyable. For people here who think "replaying the last 30 minutes of a game" is "annoying" why isn't Dark Souls or Demon Souls "annoying" when you die? What makes that game more interesting to play and "scary" as compared to games like BG/BG2 where I can quicksave/quickload/rest after each battle?

Edited April 8, 2014 by Hormalakh

[Hormalakh]

I also want to say that I don't want to take away what Sarex says about "enjoyability" from those difficult encounters - those do exist. But I don't think those encounters are enjoyable due to their metagaming potential. They're enjoyable despite their meta-gaming potential. In other words, what players see as enjoyable in those encounters actually has nothing to do with being aware of what is coming up or being able to quicksave/quickload; the reason they are enjoyable is because they provide an interesting monster, a different way of playing your party, or a very nice reward after defeating the tough boss (carsomyr +5 after beating the Red Dragon in Windspear)

[Cubiq]

I get that you don't like metagaming, but to be honest some(a lot?) of us do. As you may have read in this and other threads, some of those hard-counter/metagame fights were the most memorable and exiting for us. They were a challenge to beat and prove to yourself that you could beat the hardest thing the game throws at you. In fact the best fights were completely optional.

I don't see how you could be proud of the fact that you got your ass handed to you the first time, and then you came back prebuffed with direct counters to that, and feel victorious.

The exact example is the beholder cave. You go in, you see there are beholders, you die, reload, and come back with cloak of spell turning, and ease-walk it.

In fact it will be much harder if there aren't hard counter prebuffs in the game, but instead use the mechanic Sawyer proposed where you will require to react correctly when you get hit by a killer spell mid combat.

Edited April 8, 2014 by Cubiq

[Hormalakh]

As for traps: I think again it deals with the "immediate data/incomplete data" conundrum. Imagine if each dungeon had traps set with a RNG for each that "turned on/off" or another RNG which changed the location of each trap on every reload. Sure, you might know where the traps could be, but in each reload, you aren't sure where they are. You have incomplete data. The best way through this sort of game would be to play it slowly and hope your trap detection is high enough or use a tank to explode each trap becasue the meta-gaming knowledge doesn't help you: you have some data (there are traps here, and maybe a few of the locations where the traps might be) but you don't have complete data (where the traps are and whether they are activated with each reload).

 

Minesweeper is fun because it's random. You have some data (10 mines in an easy map) but not all of the data. Playing the same game of minesweeper over and over again (complete data) isn't fun.

Edited April 8, 2014 by Hormalakh

[Fatback]

I agree with with cubiq how can it be the hardest thing in the game if 1 buff makes it obsolete. Wouldn't it be more fun if it was 5-6 things you had to do to overcome the challenge. Those things might also be different based off of your team composition.

Edited April 8, 2014 by Fatback