End game difficulty and PE

[anubite]

I think balancing an RPG for the end game can be pretty difficult because you have to account a lot of theoretical party complexity. On one hand, you don't want encounters to be too hard, you want the pacing to match the development of the player's party. On the other hand, players can be very good at exploiting mechanics. Although SoA doesn't have this as much, Throne of Baal has many, many moments where my group could just right click entire enemy groups to death without breaking a sweat, seldom having to actually use any of my upper tier abilities to clear most enemies. And really, despite its rough spots, I was able to clear Watcher's Keep before clearing the Underdark in my second playthrough (granted, I chose to skip fighting the final boss, knew I stood no chance) ... and that place is pretty freakin' ridiculous. Even some of the fiercest challenges can be beaten with the right party makeup and tenacity.

 

I know Obsidian has talked about its current stance on difficulty, adding a few modes to enhance difficulty, and to tailor enhanced levels of difficulty by adding diversity and size to monster packs instead of just buffing damage and health.

 

But I was wondering, perhaps, if there should, or could be a rudimentary little check after a certain point, to see how "easy" the last fight was. On higher difficulty levels, I was thinking, what if there were something like L4D's ai director? Towards the end game, this director would kick in after checking to see whether you're right-clicking things to death or not. If you are, it would further add to enemy diversity and perhaps unlock certain abilities for monsters, to further enhance difficulty? And if you start wiping a lot afterwards, it would turn back the difficulty to normal.

 

Just an idea anyway. Not saying it should be all game long, or exist on the easier difficulties. Just that, in spite of a developer's best intentions, a game may still be "too easy" for some players, given the right items and party make-up.

[TMTVL]

So if some poor shmuck who's not really all that good at the combat system manages to get a bunch of lucky rolls, he gets steamrolled in the next encounter?

[Heresiarch]

The whole director thing made a lot of sense for a survival shooter, but I don't think it will work too well for an RPG. For the very reason TMTV mentioned. I remember quite a few fights made easy because of lucky rolls in BG and IWD. Or just one well-placed devastating spell. Being obliterated in the next fight because of this will not make anyone too happy.

 

What I would love though is difficulty settings independent of enemy damage or hit points. Better tactics, better enemy group composition, more abilities even simply adding more enemies makes the game a lot more entertaining than simply making player's party hit like they're armed with wet noodles and go down in a blink if an enemy so much as slaps them.

[AGX-17]

L4D's variable NPC spawning algorithm (let's call it what it really is,) isn't exactly a good fit for an RPG. The "AI Director" is just a fluff "character" meant to draw the ire of frustrated players away from the developers or the game's design.

[anubite]

Well, I admit the "AI director" in L4D probably isn't the best example, because in practice, you're right - all the "AI director" does is spawn zombies of count Y in position Z based on how well a group of players is doing.

 

What I mean perhaps more specifically is:

 

If you are, for let's say 5-15 fights, being scored by some heuristic a "10" or "you're steamrolling the game" - and if you're on a harder difficulty, it adds a few elements of "the next difficulty up" to your game? Perhaps some players wouldn't like it, but, I guess I'm not sure what to suggest. Maybe this is a problem that will solve itself and it's probably too early in development to consider, but I think such a mechanism could be programmed, where it would be able to determine where a good threshold of difficulty is for your experience, by dynamically adjusting it based on consistent fight perfromance (luck shouldn't be a factor after 10 or so battles).

[Game_Exile]

The AI director is a good example because you want to give story reasons for the type of "rank" system you are proposing. The major problem is that if the game has some long term strategic depth, it makes more sense to give the player advantages for doing well in the beginning rather than disadvanatages, not to mention this would likely end up hurting the story anyway. I don't like the idea, myself. The devs should try to make a well balanced game, maybe with some advanced difficulty options that are appropriate for first time playthroughs (if they are going to have difficulty options at all).

[SqueakyCat]

The AI director is a good example because you want to give story reasons for the type of "rank" system you are proposing. The major problem is that if the game has some long term strategic depth, it makes more sense to give the player advantages for doing well in the beginning rather than disadvanatages, not to mention this would likely end up hurting the story anyway. I don't like the idea, myself. The devs should try to make a well balanced game, maybe with some advanced difficulty options that are appropriate for first time playthroughs (if they are going to have difficulty options at all).

 

In case you haven't read this update:

 

http://forums.obsidian.net/topic/60707-update-9-the-power-of-modes-and-godlike-races-our-23m-stretch-goal/

[Game_Exile]

In case you haven't read this update:

 

http://forums.obsidi...m-stretch-goal/

I have read it, and I don't like the description of any of the "modes" that they mentioned. None of them look appropriate for a first time playthrough, and they even hint at lazy design.

 

1) Expert Mode- The main feature here is apparently the elimination of "helper" information for things like skill thresholds and reputation modifiers. What exactly is the point of doing this? Are there going to be enough signs elsewhere so you can 1) figure out how the skill and repuation systems work, and 2) estimate the effect your skill and reputation will effect things without having this helper information? If not, then you are just guessing the entire game, and If you already know what will happen from previous playthroughs, then this mode adds zero challenge. It just sounds like some slapped on BS the devs invented so they don't have to design a game with any strategic depth. And the weighty gold idea sounds like it could be interesting, but only if there is some strategic depth connected to carrying weight.

 

2) Ironman Mode- I can't help looking at the jump from unlimited saves/reloads to ironman mode without suspecting that the devs are designing the ironman mode for people who know what every choice does, meaning choices in the game will not be very complex (which is, once again, lack of strategic depth). I mean, this game will likely take 20+ hours to complete. At this point, my hopes for it would be higher if, instead, I saw plans for a well designed checkpoint system that is integrated with the story. Of course, for all I know, this game could turn out to be complex, challenging, and well balanced for Ironman Mode, so I don't outright object to this option.

 

3) Path of the Damned- Throwing in every enemy from every difficulty level is a stupid gimmick. And I hate the attitude behind this mode. It's like they think challenge in a CRPG is just a gimmick, and they don't care about how complex and well designed challenges support the story and setting (and vice versa).

 

I wouldn't be upset if they scrapped all these extra modes and just designed a challenging and complex game with a good checkpoint system.

Edited December 1, 2012 by Game_Exile

[BasaltineBadger]

Just make the main boss reasonably difficult and make a few hidden bosses, possibly extra boss granting a new ending for more experienced players. The main antagonist doesn't have to be most powerful man in the game.

[SqueakyCat]

I wouldn't be upset if they scrapped all these extra modes and just designed a challenging and complex game with a good checkpoint system.

 

You'd better brace yourself for disappointment then:

 

Question: PE will have a "save anytime, save anywhere" system like the IE games?

 

Feargus: Yes, save anywhere you want.

 

You can find the quote in the following link (under the comments posted on Oct. 4). These comments are from the Kickstarter page in which Feargus regularly posted.

 

http://forums.obsidian.net/topic/60874-updated-summary-of-the-qa-in-the-kickstarter-comments-page/

 

If you are adverse to using the save system, you can always just use the Autosaves to simulate (more or less) a checkpoint system.

 

I personally like the different modes they are offering and the ability to turn on/off options to suit individual preferences. Some people like it easy and others like it bonecrushing hard.

[Game_Exile]

I know they will have a free save and reload system as the default option. I ignore it when I talk about how a game will be designed and balanced, because obviously you can't balance any sort of challenge in parts of the game where you can save and reload any time, unless you assume the player won't save or reload any time. In which case, you are balancing for a checkpoint system which you won't even implement (which is retarded, but retardations like this are all too common). Again, getting rid of the other extra modes and having a challenging game with one serious optional checkpoint mode would be OK with me.

 

Some people like it easy and others like it bonecrushing hard.

It's not just whether it is easy or bonecrushing hard. It's also a matter of 1) making everything fit the story and setting, and 2) having long term strategic challenges that make more of the setting and story outside of combat scenarios more interesting.

 

If you are adverse to using the save system, you can always just use the Autosaves to simulate (more or less) a checkpoint system.

This is a bad idea because, first or all, autosaves will not be well balanced, or it is anyway unlikely that they will be.

 

But more importantly, when a player has a challenge in front of him he is searching for all available ways to overcome it, and save/reload will almost always be the most obvious, most ("mentally") energy efficient way to do so. Pretty much every time the player encounters any hint of difficulty or frustration, he has to use a lot of blunt discipline to keep from saving and reloading, or else he will simply turn parts of the game into short trial and error segments. This is important, because, in the context of the game (and its setting/story), the player is always looking for the most efficient ways to get good results, not looking for a "challenge" per se. So if you want meaningful challenge that supports your setting/story, and not contrived challenge that hurts it, you will not force the players to do what amounts to picking a difficulty/challenge option every time they are doing something in the game.

 

Like everything having to do with improving videogames, it is a matter of making your game more immersive, i.e. having things more interesting for the player to "do" and wrapping them in an illusion. Though the "illusion" and "the things that the player does" are actually one and the same, and they appear to be more separate or less separate from each other depending on how well your game's mechanics match it's aesthetics.

[SqueakyCat]

Game_Exile - Thanks for your well thought out reply.

 

I agree, in theory, to much of what you've said. I simply don't like (personally, of course) a checkpoint system in a RPG. Naturally, it's somewhat for selfish reasons, but I don't always have a long stretch of time to play and most good RPG's require many, many hours to complete.

 

Even when I do get a block of time, it seems there's always interruptions -- friends or family dropping in or one of life's many minor emergencies that needs handling.

 

I am also a player who likes to try different strategies in combat and will sometimes reload (even after winning) just to try a different strategy or group of spells.

 

We all have our own individual playstyles. I certainly respect your view and playstyle as well.

 

If memory serves, I was under the impression that PE will be more like BG2 with smaller maps to explore (sans Big City). If that's the case, wouldn't an autosave happen each time you travel to a new area? Just trying to think of some way that would appeal to your playstyle.

 

In the end, though, I realize your issue with the design is the balancing. Hopefully, OE will get it right. Mr. Sawyer also seems very much into balancing everything so hopefully it will turn out to be a game you'll enjoy.

 

Edit: It's also interesting that Feargus only said "anywhere" and not "anywhere and anytime". Hopefully, you won't be able to save during combat or other times that they deem appropriate.

Edited December 2, 2012 by SqueakyCat

[Dragoonlordz]

The monsters you face I hope do not (all) auto-level increase to keep matching your own. Limit the ones that can auto-level to those with high enough possible intelligence to learn new tactics and find and use new equipment. If I can squish a rodent at lvl 5 I do not want that rodent to magically be able to squish me at 50 since they do not wear better armour, do not equip better weapons, doubtful that are smart enough to organise a different tactical approach to your party in combat. A mutated version might be stronger but the original type would not pose a greater threat outside of greater numbers..

 

However I may be able to beat and Orc at say lvl 10 who has poor gear and low intelligence, lacking decent tactical knowledge of our party and methods but when we as characters have equally poor armour and weapons plus lack spells and have not yet learned the best tactical ways to beat that unit type then it should be hard to beat. Later on in the game if that same Orc type is using same weapons against my better armour, same armour against my better weapons, same tactics due to low intellgence against our improved knowledge of their methods based on past experience then that unit should remain easy aka not all of sudden be stronger then us.

 

There is logic to changing monster type as progress and also having a auto-level range but that range should not be from lvl 1 upto max hero level for both those aforementioned Orc or Bat individual unit types. An auto-level idiology in the game should depend on the type of monster (imho) and rely partly on use of ranges for types. If the monster is not one capable of higher function learning or adapting then it should not auto-level past a certain point (imho) for example if cannot auto-level past the 1-10 range or 11-20, if it is a type that is capable of such adaption then should auto-level between ranges and bringing in new unit types of that species right upto maximum level. Model variations should be used for those adapted and improved units to suggest or show this improvement from past versions fought to the player. Both from visual, equipment and tactical change. The non-adapted ones however should still exist in their weaker form meaning they will never evolve or appear in the 41-50 auto-level range, they can be used to back up the stronger new units.

 

From a character perspective as opposed to player, he or she is not actually getting much stronger through the game outside of better armour, weapon and spells, he or she may learn new spells and may have better armour or weapons but some monsters should be able to do that too for a lot of those more intelligent species. They in theory should be able to find new weapons, wear better armour, learn new skills and spells even if not witnessed on screen by the player this should be happening in the world itself as time progresses from a logical standpoint which gives validity to new unit types of same species. Some will do this while others will remain poorly equiped, those poorly equiped versions should not be able to take me (who is by this stage better equiped and knowledgable about that species weaknesses and tactics) down at least not without greater numbers on their side or a newer higher inteligent, better equiped variation of that monster overseeing the weaker types actions and tactics.

 

We as the player as opposed to character though however see statistical information and levels even if the character does not. The intelligent types of monsters in theory are not frozen in time, time passes for them just like it does for the characters and within that time they learn just like you do. So while may face same species at any point in the game on top of new enemy variations of that species (unit types), those new variations should be improved (stronger) via equipment, spells and skills plus tactics (intelligence) aka higher level range but the weaker less adaptive ones should not be at that same higher level range. They should also apply new methods and tactics to beat you. In reality there is no level bar above our heads as people, the difference between player and character perspective. We may learn new things or find better equipment as the character but the same should be applicable to the opposing side if they are humanoid or intelligent species.

 

Non-intelligent species or adaptive ones like bats as an example and such should not auto-level past certainrange no matter what point in the game your at (imho) and those must remain as weak as the first time you beat them, they can be increased in difficulty by increase of numbers but not those specific individual bats getting stronger.

 

The variety in types should apply tactical idiology, you may encounter weaker Orcs in my example early on and also again later one and by weaker I mean poorly geared and low intelligence. If encounter later on they (that specific variation) should be as easy to kill as was prior. To increase the difficulty would be increase the frequency of the better equiped and higher inteligent variation of that species later in the game, bring in new forms and use the weaker version as cannon fodder to back up those stronger visual, equiped and methodological variations. Reinforcement systems applied for larger conflicts, tactical variations in battle strategies that improve over the course of the game from the enemy/opposing forces learned from past mistakes and battles.

 

The TL:DR version...

 

Intelligent, adaptive monsters and species should auto-level but only between a set level range (along lines or 1-10, 11-20, 21-30 etc) between those ranges the method using increased variable unit types should be applied as way to bridge the gaps. A monster from the set range 1-10 should not be able to kick my arse when I am in the level 41-50 range. A new unit type of the same species should be used or brought into play which is in the range more equal to my equipment and skillset, tactics and statistics. The lower 1-10 variation could and should be used merely (imho) as cannon fodder to increase the numerical tactical advantage of the 41-50 range opponent I am facing. Meaning both continue to exist in the game, only the purpose has changed for that weaker unit type. In hindsight I thought getting what I mean across would be alot easier than in reality turned out to be...

Edited December 3, 2012 by Dragoonlordz

[Ieo]

I'm not in favor of the OP's idea. The RNG may make that sort of implementation patently unfair, and there was no definition of "difficulty" either. In fact, only the developers have a very good idea as to what variables go into "difficulty" and what mechanisms can and should be tuned. The reason why they have all those extra hard modes is simple--the so-called hardcore have different ideas as to what constitutes "difficult." Ultimately, I'm not concerned about any of that because:

 

There will be special 'hard' modes that aren't mutually exclusive.

There will be tuned difficulty levels covering the gamut.

 

...

IMO, the best RPGs use story elements to reinforce the game play. That game play may be combat, choice mechanics, skill use, etc. This does not mean the game play should be unplayable for "casual" players. At Obsidian we value narrative and game play equally.

 

Whatever the "easy" mode ends up being, it will be very manageable for that type of player. We've already announced manageable stretch goals to include some WICKED options for you hardcore types (believe me, you are gonna get punished ). The folks that want an easier time of it, will have that option as well. As long as making the game accessible does not influence or "dumb down" the difficult settings (which it won't), everyone wins. After all, we are making this game together. We want you ALL to enjoy it.

... The default mode should, and will be, a challenge for the average RPG player. I've not spoken to Josh about specifics for an "easy mode". This will either be a mode in and of itself, or we'll have a few options to lower specific gameplay features to more casual levels. Again, the goal will never be to trivialize the default difficulty settings.

 

 

The monsters you face I hope do not (all) auto-level increase to keep matching your own.

 

You're talking about level scaling. Project Eternity will not have level scaling except the main campaign (think F:NV, apparently). It sounds like there will be encounter scaling depending on what region the player decides to visit first (makes sense--BG had this, I believe, and non-railroad exploration games should have some flexibility for that scaling difficulty).

 

http://www.pcgamer.c...dian-live-chat/

http://forums.obsidi...96#entry1218696

[Game_Exile]

Squeaky, thanks for your sympathetic remarks.

 

Even when I do get a block of time, it seems there's always interruptions -- friends or family dropping in or one of life's many minor emergencies that needs handling.

 

I am also a player who likes to try different strategies in combat and will sometimes reload (even after winning) just to try a different strategy or group of spells.

A checkpoint system where you can quit+autosave any time but only reload from checkpoints will solve your interruption problem. And if you want to experiment with abilties and strategies, I don't see why you would have to be able to reload right before each fight in order to do it, unless you want every fight to turn into a trial and error run.

 

Combat and enemy encounters should be designed so that the player would have to figure out how certain abilities or combinations of abilities work in order to deal with them. The demands of enemy encounters would be simple and varied at first, then go on to be more and more complex as you learn, so that you will have to properly use more and more of your characters' combat abilities as they increase. This should go for most of the game's strategic components, even outside of combat. Later challenges will demand that the player bring your knowledge and experience from earlier in the game, where there was maybe some more room for trial and error, but still some demand to learn the ins and outs of the resources available to you. This is the kind of "difficulty mode" people want (even if they don't know it yet).

 

BTW, the above demand is one of the major reasons why the game should be balanced for the predesigned party NPCs, rather than for multiple player generated party members which will all be availble at the beginning of the game (if this will even be an option). It will be pretty much impossible to balance for so much variability in party composition, and still have good progression in the complexity of the game's challenges, at least in combat. Honestly, the more I hear about "options" and "customIzation", the more it sounds like an excuse for the devs not to properly design and balance the game.

 

In the end, though, I realize your issue with the design is the balancing.

To be more clear, my issue with free saving and reloading is that it is a **** mechanic that is impossible to balance for. Both design and balance ultimately refer to the same thing, only with balance you are approaching the game more from the player's perspective, with an emphasis on "challenge".

[Amberion]

So if some poor shmuck who's not really all that good at the combat system manages to get a bunch of lucky rolls, he gets steamrolled in the next encounter?

This. I remember playing Mega Man X... 7? It had a difficulty system whereby if you blew through a level with high performance, the next few levels got harder. I took the game back. The encounter difficulty should be a result of player choice, not automated based on player skill.

[JFSOCC]

So if some poor shmuck who's not really all that good at the combat system manages to get a bunch of lucky rolls, he gets steamrolled in the next encounter?

You could average it over several encounters, so that it's more gradual. that should compensate for statistical anomaly.

[anubite]

There is no concern of luck. If you steam roll 10+ fights in a row, luck simply cannot be a factor. That's not the issue at hand, the issue is whether such a system should be done, how the monsters should be scaled (and I do not mean like Skyrim, I know a flat level scaling system for monsters is dumb, I specifically mention giving monsters new abilities, specific resistances/traits, or AI to enhance their strength). Creating a heuristic to track player progress accurately could easily be done with low error and it would be trivial to "reverse" the changes done by an AI director if you are suddenly losing.

 

I can see an argument of, "If a player has found a winning strategy. Let them win." But I am specifically worried more about the case where my Minsc at the beginning of ToB could demolish entire groups of enemies on his own. Detecting that kind of thing would be a little harder and I guess if people think, "If you can demolish the game, you should be able to." Is a fine way of thinking, I can't really disagree, though making the game slightly harder when you start demolishing it, I don't think would prevent anyone from having fun. Also, this suggestion of mine was specifically geared towards the end game.

[Amberion]

There is no concern of luck. If you steam roll 10+ fights in a row, luck simply cannot be a factor. That's not the issue at hand, the issue is whether such a system should be done, how the monsters should be scaled (and I do not mean like Skyrim, I know a flat level scaling system for monsters is dumb, I specifically mention giving monsters new abilities, specific resistances/traits, or AI to enhance their strength). Creating a heuristic to track player progress accurately could easily be done with low error and it would be trivial to "reverse" the changes done by an AI director if you are suddenly losing.

 

I can see an argument of, "If a player has found a winning strategy. Let them win." But I am specifically worried more about the case where my Minsc at the beginning of ToB could demolish entire groups of enemies on his own. Detecting that kind of thing would be a little harder and I guess if people think, "If you can demolish the game, you should be able to." Is a fine way of thinking, I can't really disagree, though making the game slightly harder when you start demolishing it, I don't think would prevent anyone from having fun. Also, this suggestion of mine was specifically geared towards the end game.

I still think such difficulty options should be left entirely in the hands of the player, whether they want to enable them or not.