Jarrakul

I'm having what I think is the same problem as the others here. I can load the save (linked below) just fine, but if I try to enter the Vailian Trading Company or the Adra Mill, I get the same "error loading the next map" crash. This occurs even if I revert to an earlier save. I've tried repairing the installation, but it didn't help. https://www.dropbox.com/s/332gfyvtx65sujq/Aelwen%20%28c1b53009-d40f-455e-86e7-fd113c601dc1%29%20quicksave.savegame?dl=0

Yes, how dare the devs cave to the demands of their fans after committing to serve the interests of... their... fans... Waitaminute... At the end of the day, their job is to take backer/fan feedback and do what they think is best to make the best game they can. That's all they did here. I, for one, am glad they made the right decision, because making their backers/fans feel welcomed is important, and I'm sorry to all of you who disagree with it. But this? This isn't what censorship looks like, and calling it censorship is an insult to everyone who actually has to live with censorship.

Oh god chanters sound so cool. They are everything I want a bard to be. :D

The point of narration in a game, I would argue, is to sum up or skim over events. Don't want players to have to spend hours marching around town rounding up basic backstory? Give a couple-second overview in narration. Or in mandatory dialog, it doesn't really make a difference, but narration probably feels less forced in most cases. Want to talk about how bad the weather is during a sea voyage, but don't have anything mechanically interesting for the player to do? Just describe it to them. It's simple, and it works. Not always optimally, granted, but it does the job at least passably. Now, as to the OP's thing about who the narrator should be, I do think having the narrator be a person who really exists in the gameworld can help with immersion, and having it be the player character can be a little too much putting words into the player's mouth. So yeah, I do think having the narrator be an NPC (Belhifet, Oswald's niece, Kelemvor, etc.) is generally a good idea. I don't think it's necessary, and it's not hard to imagine that it wouldn't work for some games, but I think in general that it's a good idea.

I find DA to have a fundamentally better combat system ruined by even-worse-than-BG balance. Playing as a rogue, at least, felt like I had a bunch of abilities that could be useful in any given situation, but I had to figure out which ones would be useful when. But then the mage just dropped an Ice Storm and made all that irrelevant (until the endgame, when I suddenly found myself borderline invincible for reasons which are still somewhat beyond me). At least when that happened in BG, I could go "yeah, but he can't remove a trap." As much as I dislike balancing out-of-combat prowess against in-combat prowess, it was still better than just not balancing them at all. DA1 also suffered from a severe lack of precise information, if I recall correctly, which made it difficult to know what would be effective when until you'd used it multiple times, which made strategizing difficult. DA2 fixed that problem, but introduced... others. As for BG2... it is an oversimplification to claim that mage fights are "counter or die". But they're close, especially as you get into the late game. The difference between having the magic true sight + breach combo (or just an inquisitor) and not having it is huge. It's not quite win vs. die, but it's by far the biggest difficulty swing from knowing a specific spell since BG1 basilisks and Protection from Petrification. While I do like a lot about the BG2 magic system in general, I think mage fights would be a lot more fun if they were more consistently somewhere between the two extremes of countered and non-countered. That doesn't mean there can't be counters, but they should be softer, as should un-countered defenses.

That would be much better, yes, because losing one party member in a party-based game really isn't that bad. But it's still highly random by the standards of damage-dealing effects, and I'm still not totally okay with that. What I'd love to see is something like "Hold targets for 0 seconds on miss, 3 seconds on glancing hit, 10 seconds on full hit, 20 seconds on crit" or something like that. So you could count on it doing something almost all the time, but it would never last so long that the target would be out of the battle altogether. In this way, it remains powerful, but becomes a lot less random. It's now a reliable tool with some random effect, rather than a totally random effect that might do nothing or might insta-win depending on how lucky you get.

I should note that while I seriously dislike spells like Hold Person, I have no problem whatsoever with spells like Web, Entangle, or Command. The reason for this has to do with the way the random distribution works as the way it makes me play. I dislike Hold Person because if I get unlucky once, half my party is out of the fight. Probably dead if anything happened to be attacking them at the time. I dislike it offensively too, because when I cast it I'm basically saying "I hope the enemies fail their saves so I can insta-win this fight." It's not tactically interesting. Which is not to say there's nothing of tactical interest about a Hold Person spell, but the core effect (save or be rendered useless and defenseless for the rest of the fight) is not tactically interesting. However, I like spells like Web, Entangle, and Command. Command is simple. I like it because it's effective and short-duration. For much of BG1, it isn't a gamble, because most BG1 enemies don't get a save against it, and by the time enemies that do become common a 1st level spell and a casting time of 1 isn't much of a resource to gamble. My fondness for Web and Entangle is somewhat more complicated. At first glance, they seem a lot like Hold Person, but they have two crucial differences (beyond Entangle being a root instead of a hold). First, they trend much more heavily towards the average result. This is because their effect only lasts one round, but it's checked every round. So instead of failing one save and being out of the fight, you fail a save and are out for a few seconds, and then you roll again. Both Web-style effects and Hold Person-style effects may have the same expected value in rounds spent paralyzed, but the Web sums multiple random events to produce a result that trends much more strongly towards the average. The second reason is that Web and Entangle are effects that sit on an area, rather than on creatures. That opens up all sorts of fun tactics, both in how you use them and in how you deal with enemies using them. Their existence adds tactical depth, and should be lauded. But note how they differ from Hold Person. I'm not arguing that status effects should be weak. I'm arguing that they should be interesting. Also, the "you're all noobs for not liking this mechanic" argument needs to die in a fire. I beat Baldur's Gate with SCS and max health enemies. I've only ever run one hardcore run, but I beat it easily (although without SCS, admittedly). I am not somehow unable to deal with the sorts of things BG throws at you. I just prefer challenges that make me respond in interesting ways, rather than challenges that render me unable to respond unless I know about them beforehand and have prepared their specific counter.

So much yes. I think I once brought up something along these lines in times of yore, or possibly I just brought up something tangentially related. Either way, I have... very lengthy opinions on how this might be handled. I'll spare you all the details unless anyone's actually interested, but suffice to say I heartily approve of the concept.

So, I notice some folks blaming the rogues-as-DPS thing on MMOs, but I don't think that's where it comes from. I'm pretty sure it comes from not wanting to have combat specialists and non-combat specialists. That's a dichotomy that crops up a lot in old RPGs (Fallout is probably the best example) and in PnP RPGs, and while it's entirely realistic, it makes for bad gameplay (although if you happen to be Fallout you can make up for it in other ways). This is something that became blindingly obvious to me when I played a d20 Modern game as a Charismatic Hero. I was a socialite to end all socialites. My Bluff checks were the stuff of legend, let me tell you, and it was glorious. Until a fight started. When I fight started, I was suddenly unable to contribute... anything. I mean, I could shoot at things, but hitting was basically out of the question, as was dealing decent damage if I did somehow manage to hit. The greatest contribution I could add was if one of the enemies made the mistake of wasting time shooting at me instead of someone who was actually a threat. As much fun as the character was when he was in his element, a large portion of the time I was utterly useless. Not "not the best". Useless. Nothing I could do would ever swing the direction of a fight. And as such, overall, it was the least fun character I have ever played in 18 years of PnP RPGs. Playing rogues in 2E and 3E is similar (I never played much 1E, but it's close enough to 2E for me to guess). Not quite as bad, because you get one good hit before you're worthless. Not even that good a hit, really, but it's something. 3E is slightly better than 2E in some ways, because if I can flank things I can actually do good damage right up until the moment anything realizes I'm doing damage and I immediately die. Which means I get one good round, instead of one good hit, and then I'm twiddling my thumbs for the next couple of hours. And, of course, if it's 3E and I'm up against undead, I'm just totally useless. Now, I'll be the first to admit that the applicability of that to party-based games is limited, but it still holds true enough to be a serious concern. In games where a given player only controls one character, it's absolutely essential. PS. I'm also amused by the argument that rogues shouldn't be as good at combat as fighters because it makes everyone the same, but paladins and rangers and barbarians are totally fine. So what if the rogue is another warrior class? Why can't we have another warrior class, representing an archetype that is not otherwise well-supported? What's wrong with that?

I can't comment on the musical design process with anything near the same confidence (/arrogance) that I can comment on the rules design, but the sample piece is pretty awesome. This game is going to be beautiful.

Ffordesoon, absolutely that's helpful. Watcher's Keep did a better job of that than Durlag's Tower, in that it was very easy to leave WK and come back to where you left off, but you're right in that you certainly weren't encouraged or expected to do so. If the dungeon is built to be explored in bits and pieces over the course of the game, the issues of its massive size become much less problematic. I'm still worried, though, that it's that big because we made it that big, rather than because the devs thought it should be that big. I don't want forced size. I want whatever size the devs think will make the best mega-dungeon. That said, I do think that if we are exploring it over the course of the game, and if that distributed exploration is forced in some way (say, with locked doors whose keys are elsewhere), then unlocking new levels of Old Nua could be a really fun reward.

It's worth noting that there are subsections of the Baldur's Gate community that have made good use of the rolling mechanic. Challenges like no-reload playthroughs with your first stat roll can be a lot of fun to some folks. But for the vast majority, the random-sum nature of the stat system meant that you just kept rolling until you got the stats you liked, which really just wasted the player's time. What I'm saying here is that I think Walsingham has the right idea. It's a great option to have for the Ironman folks, but it kind of stinks for everyone else. Although I'd still prefer a system that combined point-buy with random variance, as I mentioned before.

I am not a fan of "dead and hasn't noticed yet" status effects, like long-duration stuns and holds. Hold Person might as well be save-or-die for all the difference it makes, except that it's slightly more merciful to the player. If you're gonna have long-duration hold spells, you might as well scrap them and just replace them with save-or-die effects, because it'll accomplish the same thing. Not that I'm advocating that, mind. I hate save-or-die effects with a passion in RPGs. It's just worth realizing that some status effects have, in the past, been functionally the same thing as save-or-die.

I would love to see weather have small in-game effects. Lephys's wind stuff would be cool, and a simpler thing could be done with weapon range and fog. Rain could slow movement slightly, and maybe thunderstorms have a chance to briefly stun everyone every so often as a particularly large boom sends people reeling for a second. I have no idea how feasible any of that is, but in infinite-funds-land it'd be awesome to see.

I actually find the Skyrim map (my main frame of reference for the 3d style) to be somewhat difficult to use. It's not sufficiently exaggerated to clearly depict much of anything but snow and mountains. If we get a 3d map, I'd prefer something that looked more like the last couple Civilization games, with everything huge and exaggerated and very visually clear. Of course that ends up looking kind of cartoony in 3d, so I think the best result would be achieved by having something 2d that looks like a traditional paper fantasy map. That's not to say I don't understand the desire for a Skyrim-style 3d map, those are definitely cool, but it's not my preference. Also, 2d is probably cheaper, and honestly there are a lot of things in the game I'd rather the time/money be spent on.

Honestly, I'm just kind of concerned about the size of the thing. It's absolutely massive, and frankly I'm not sure that's a good thing. Durlag's Tower was a wonderful lore experience, and Watcher's Keep had some really cool and varied gameplay, but both of them felt really long, imo, and by the end I was definitely glad to have them done. Old Nua is set to be, what, three times the depth? If that translates to three times the size, that's a ludicrously huge dungeon. From a pacing standpoint, that just feels like a tad much to me. Now, obviously I trust the devs. They're smart folks, they'll figure something out. But I can't help but feel that Old Nua is as big as it is because they were trying to get more Kickstarter money (a worthy goal, make no mistake) rather than because that's the size they felt would be the most fun.

I'm sorry, I don't really know what to say, but this is simply untrue. The numbers aren't the same. When I'm looking at characters with THAC0s and saves in the ballpark of 0, and armor classes in the ballpark of -6 and spells with maybe a -3 to the save, it's pretty obvious that attacks are going to hit a lot more often than people are going to fail saves. And that is exactly the sort of thing you get around late-BG1/early-BG2. A number of the spells you list aren't damaging spells, so direct comparison is difficult (and I will also not argue that Protection from Magic Weapons was anything near balanced). As for the others, you're kind of right. There are a number of factors that can be altered. Of course, there were a few high-level, single-target damaging spells in BG, and there will likely be a few in PE, so that one doesn't really work. As for situational-ness... of course different spells are more useful in different situations, but if save-or-die spells were anywhere near as good as you seem to think they are/should be (and in the IE games, they simply weren't), I guarantee that these spells will be the most useful in a great many situations, because a 50+% chance of killing any given enemy with a single spell is amazing. As for casting time, yes that is a factor that can be played with, and actually I think you could do something interesting with very long casting time (1 minute or more) spells that just killed a target, no save allowed. As it is, I'm just gonna say that in order for this to be sufficient for about a ~50% chance of killing the target outright, the casting time difference is gonna have to be big (on the order of several rounds or equivalent), and that's just bad game design when you spend a ton of rounds doing something that then just fails because you got unlucky on one roll (fun fact: that's what this boils down to anyway). Negative caster effects tend to produce the sort of NPC-PC asymmetry that, to my mind, typically destroys immersion (you won't use Disintegrate because you can just reload if you die, but the enemy can use it all he wants because what does he care about loot?). But hey, if what you want is a win-button against any boss that makes you not get loot for the fight... wait, no, that's still dumb. Well, no. Because I'm not arguing for mindless simplicity. I'm arguing for you not being able to win a battle on round 1 because you got a single lucky roll. I'm arguing for well-tailored difficulty. I'm arguing that not all complexity is good, and you don't seem to get that. Which would worry me, if it weren't crystal clear to me that the devs disagree with you. And this is where I decided not to bother anymore. Seriously, you accuse me of not dredging through 17 pages of thread to find your views on what made save-or-die effects good, while you can't be bothered to spend two seconds Googling one of the most popular board games in the world? Fine. Have a wikipedia link. http://en.wikipedia.org/wiki/Go_(game) Importantly, the rules of Go can be taught in about five minutes. The depth of strategy makes Chess look like Parcheesi. Now, to Hiro Protagonist: Go is relevant because it's the disproof of the "simple is bad" argument in games. Obviously I'm not arguing that PE should be Go: The Video Game, but Go proves that simple games do not necessarily lack strategic depth. Go is, in many ways, the pinnacle of elegance in game rules, and although PE is not aspiring to be Go, that elegance is something that all games should aspire to, even if they are building something more complex. In other words, if your game is gonna be complex, you'd better get a lot out of that complexity, or else you end up with the universal TV remote of video games. What you miss, crucially, is that, with the exception of the rogue's Evasion ability, none of those changes are random. There's no "roll to see if your opponent is fire immune." Even when the rogue increases the randomness through Evasion (which, notably, is not itself random), it doesn't reach save-or-die levels because if the rogue fails his save, he just takes some damage. Possibly enough to kill him if he's squishy, but that's because of largely-deterministic or low-variance factors like hit points. He doesn't just die because he got unlucky once. He dies because he has a low Con and then got unlucky once. Which, frankly, is something that just sort of happens to characters with low Con (seriously don't dump Con). Maybe that's its own problem, but it's not the same problem. As I've said repeatedly (you'll note that I'm actually willing to repeat things for you), the problem with extreme randomness is that it leads to difficulty curves that make no sense at all. A random fight with goblins could be very difficult while a dragon in the next room could be a cakewalk, just because of how lucky you've gotten. Sense good difficulty curves tend to be important for immersion, sense of progression, and just general fun, this is a big problem. Let me illustrate with a story. A couple months ago, I was playing BG2 for the umpteenth time. It was late in the game, and I had just finished trashing Firkraag's minions. Now, I love Firkraag. He's literally my favorite enemy in the game. I was a bit overleveled, I'll admit, but even so I was looking forward to a good fight. And so I cast my buffs, walked in, and started fighting. About 10-seconds in, he just dies all of a sudden. Turns out my Silver Sword's vorpal effect proc'ed and Firkraag failed his save. Seeing him just die to a random effect like that, not because I was skilled but because I got lucky, was almost physically painful. Nothing that game has ever hit me with has hurt more than seeing Firkraag just die because he happened to roll a 1 at the wrong time. So I reloaded, unequipped the Silver Sword, and fought him again, because stupid random save-or-die effects had trivialized a fight that should've been at least challenging. Now, of course you could argue that I could've just not used the Silver Sword, but from a designer's standpoint, that situation shouldn't even be allowed to happen. It'd be like dropping a +10 sword into the beginning of BG1 and going "it's okay, some people just won't use it." Extreme randomness makes the game less fun, and therefore should not be allowed. And then this. This would be fine if you'd been willing to make a passing effort to find out what Go was. If you hadn't just asked me to reiterate an argument that I think I've made with every post in this thread. If you were willing to put in a fraction of the same effort you're asking me to put in. But now? No. I'm sorry. I'm done. You're clearly not interested in discussing this in a reasonable fashion, and I'm not willing to indulge you with an argument conducted in an unreasonable fashion. I've made my points for all to see. There is nothing further to be gained. Goodbye.

Oh, I'm not arguing that I felt Soldier's Peak was good. It wasn't. I just felt like the main game was perfectly complete without it.

First, I would never, ever, under any remote circumstances point to the IE games as examples of good game balance. So what they did is kind of moot. However, if you want to go down that road, I would point out that, although the die rolled was the same for both attack rolls and saving throws, it was virtually never true that each had equivalent chances of success. At low levels, the odds tended to be comparable, but THAC0 and save numbers improved much faster than armor class and save penalties. Hence, by mid levels, your chance of failing a saving throw was vastly lower than your chance of being struck by an attack. You are once again assuming my argument only compares spells to physical attacks. That's not what I'm doing. It's not a dichotomy that has any bearing here. Both spells and non-spell abilities can potentially have save-or-die components, or they can have damaging components, or even both. My point is that a save-or-die effect has to be balanced against similar effects that deal damage. Save-or-die effects will, by necessity, have to involve a lower probability of success, or else they will simply be better. Ah, the "it's simple so it sucks" argument. Tell that to a Go player and see what kind of reaction you get. But to address the issue in more depth, of course the same does not necessarily become better as it becomes simpler. Nor does it become better if it becomes more complex. Simplicity is one goal, feature-completeness and depth are others. Complexity is not a goal, but rather the cost of feature-completeness and depth. Elegance is when the complexity cost is low compared to the payoff. My argument is not that adding complexity is bad, it's that the potential benefits (greater difficulty variance due to higher randomness) are outweighed by the cost in additional complexity (difficulty in tailoring difficulty to produce an optimal player experience). Fireball is not extreme unless its damage looks, for example, like a uniform distribution from 1 to 200. This makes it highly extreme, because the variance in its results is going to be huge. Sometimes it'll do basically nothing, and sometimes it'll kill everything, just based on pure random chance. The mere ability to kill a lot of things does not make something an extreme random effect. For that to happen, it's effects have to vary highly based on random chance. In other words, for something to be an extreme random effect, it has to be extremely random. As for Hold spells, and all other long-duration stuns, yes you're absolutely right. They are exactly as extreme as save-or-die effects, and should not exist for exactly the same reasons. "Dead but hasn't realized it yet", as someone on the Baldur's Gate forums once put it, is not functionally different from dead (except that it's not as mean when used on the player, I suppose, but that creates an asymmetry that I'm not fond of). Such spells are extreme and should be removed. Short-duration stuns are fine. I'm confused by your use of Josh's quote here. He's referring to a very powerful creature instantly killing a much weaker creature because of large power discrepancies. I'm not arguing against that. I think that should be able to happen. You're referring to a very powerful creature instantly killing anything if it gets a bit lucky (and apparently you're okay with it being just as easy as hitting an attack, judging from your earlier arguments). That is not an analogous situation. At all. If I may ask, what is it you think save-or-die spells add to the game? I've given my thoughts on the subject, and you've argued that I'm wrong about them. So I want to know what you think they add.

It's true. Coran in particular is the god of the Firewine. The place is still a pain, though.

The idea that it's all random, and therefore all has the exact same sort of probability spread, is kind of nonsense. Damage typically has a higher chance of being successful but a lower payoff. Save-or-die necessarily has a higher payoff, and therefore needs a lower chance of success to be remotely balanced (assuming equal number of targets, etc). Hence, the probability spread is quite different. If you don't understand that, I'm afraid I don't know what to say. This is basic probability here. Yes, obviously the system can handle different sorts of durability. But the more the system relies on additional systems for fundamental issues of durability, the more complex it becomes and the more difficult it is to balance. Sense the presence of save-or-die effects has a massive effect on the durability-contribution of any characteristic that can be used to resist such effects, the presence of save-or-die effects makes balance much more complicated to maintain. Not impossible, mind, but far more difficult. Actually, I'm not taking Lephys's anything. As I said, I haven't read most of the thread. This is purely from me, with my 18 years of D&D and crpg experience. I'm not comparing death spells to sword blows. I'm comparing them to everything that deals damage. That includes sword blows, but it also includes fireballs. And, I'm sorry, but the mere idea that save-or-die spells are anything but extreme in their effects is simply ludicrous. Beating the target is, by definition, a highly extreme effect against that target. Now, as to one-hit-kills, I'm pretty sure Josh was referring to the fact that stamina totals will be such that sometimes things will die in one hit, just from damage. That will naturally apply to some enemies and not to others (for example, a guard might be one-shotted, but a dragon won't be). This is fundamentally different from having abilities (and frankly, I don't give a **** if they're spells or sword blows) with the potential to one-shot literally anything in the game (or anything in the game that isn't arbitrarily immune).

Since I don't have my books at the moment, I just did some quick Google research. Tolkien dwarf women almost certainly had beards, as inferred from a stated similarity in appearance to the men and the fact that it's quite clear that beards are highly characteristic of dwarves (to the point that it would be impossible to mistake someone for a male dwarf without a beard). D&D, naturally, varies according to campaign settings, but both Greyhawk and Forgotten Realms have bearded dwarf women. Interestingly enough, they apparently often shave in both settings, so I guess I have to eat my words regarding that never showing up. EDIT: Edited for clarity.

Sadly, I am not currently in my home and therefore do not have any of my D&D books around to quote from. Honestly, my use of "a lot" was deliberately vague. Normally I'd just wait to reply to you until I had sources, but the second bit needs to be addressed. I assume you mean that I'm saying they aren't conventionally sexy, because that's what I said. Typos happen, that's cool, but I just want to be clear that that's how I read your post. Having said that, it is absolutely a matter of taste. Sexiness in general is a matter of taste. I like masculine women and feminine men. Not everyone does, but I do. That doesn't tend to extend as far as beards, but you know what? A beard on a woman wouldn't be a deal breaker for me (except as far as a beard on a man would be a deal breaker; some beards are just icky). So, I guess there's your counter-example right there. Where do you get off telling other people what they're attracted to?

Off-topic, but I just want to say how awesome it is that the freaking project lead pops onto these forums to answer our questions and concerns on a semi-regular basis. It really feels like we're being respected and listened to, and that's just awesome.

On the subject of save-or-die spells (and I haven't read the entire thread, so forgive me if this has already been said), the point isn't that luck==bad. The point is that, when one is designing probabilistic systems, it is very important to understand how different probability distributions affect gameplay. Save-or-die effects give really weird value curves, for two reasons. The first reason is obvious. They either work or they don't (sometimes with some minor secondary effect), and if they work they're amazing. So even if they have the same expected utility as another attack, they'll have much more extreme results in both directions. It's the difference between rolling 1d20 and 1d6+7. They have the same average, sure, but they have very different ranges, and lead to different sorts of gameplay. 1d6+7, for example, is far more affected by static modifiers, while 1d20 is more affected by the opportunity for repeated attempts. To go back to the subject of this discussion, save-or-die gameplay is all-or-nothing, and therefore only synergizes with effects that increase its chance of success. Normal damage, on the other hand, synergizes with all other sources of normal damage, because their effects are additive. This leads to very different sorts of gameplay (although I won't argue here that one or the other is "better"). Equally important, save-or-die effects are far more affected by random chance than damage effects are (because damage effects are are small, additive randoms that, together, become very strongly inclined towards their total mean, whereas save-or-die effects have a much weaker tendency towards their overall mean), and therefore they create a less stable difficulty curve. In fact, the same battle may have very different difficulties between reload. This is not totally without advantages, as it creates a certain unpredictability that can be fun, especially on repeated playthroughs. However, it also makes battles virtually impossible to correctly balance, and therefore severely damages the developer's ability to create any sort of pacing or growth effects in difficulty (do not take this to mean the developers lose all ability to do this; I only mean that they lose a large chunk of that ability). The second key difference is not actually been damage and save-or-die, but between straight damage and %-damage. Save-or-die can be construed as effects that deal 100% of the target's health in damage. Like all effects that deal %-based damage, these effects scale very strangely depending on the opponent. At their best, these effects give players a useful way to take down high-durability targets. Some spells should indeed work better against some foes. That's cool. At their worst, however, %-damage effects produce some very strange difficulty curves, especially at the high percentages and when percentages are the only effect applied (10 damage +2% is very different from 15% damage in how they each scale against different opponents). Certain characters now find high-durability targets to be just as squishy as low-durability targets, which is very difficult to balance for. Since save-or-die effects deal 100% damage, they are the logical extreme of this principle. If a group has a save-or-die effect, hit points and other conventional defenses become irrelevant to any party using save-or-die tactics (which, as noted above, synergize poorly with damage tactics, so a mixed-form party is at least somewhat unlikely). This once again makes it difficult to balance the difficulty between players, particularly between players employing save-or-die tactics and players employing conventional damage tactics. So, my argument is this: Randomness is not bad. Some degree of chance can be useful to add variety to encounters, particularly across multiple playthroughs. However, as with many things, extremes can become problematic, in this case because they can significantly alter the game's fundamental difficulty curve. Save-or-die effects are necessarily high-variance random effects, and therefore add a very large element of randomness to the game. Even when save-or-die effects are purely optional, their mere presence adds another layer to the balancing problem and takes developer time, and players who use these effects will still often experience a very strange difficulty curve that undermines the developers' original intent for the game. While I can understand the desire to add randomness to the game, adding the high level of randomness created by save-or-die effects seems to create more problems than it solves.