Chess is not an RPG: The Illusion of Game Balance

[Bedrockgames]

That's not really true though. Looking at 3e, mostly because that's the system I'm most familiar with, a 35 point buy character is operating in all respects at one level higher than a 25 point buy character - his HP, AC, saves, bonus spells, skill bonuses - are all one level higher.

Whether you consider that to be broken or not is a matter of taste, but, the math certainly says its a pretty big advantage.

I mean, if you were to go to your DM and say, "Ok, I'll make a 25 point buy character instead of 4d6 drop lowest, roll twice. But, I start at 2nd level instead of 1st like everyone else," do you think any DM would go for it?

Heck, can I play an 18 point buy character in your game, but start two levels ahead of everyone else? Would that be perfectly fine?

At the end of the day, die rolled characters are almost always higher value than point buy characters. And that's a balance issue. It might not be a huge one, but, it does make a difference. I know that my earlier D&D experiences vary wildly from, say, [MENTION=4937]Celebrim[/MENTION]'s because we were very generous with character generation. The idea of a fighter that didn't have a percentile strength was a foreign one to my groups. Why would you play a fighter if you didn't have percentile strength?

I certainly won't deny that point buy characters produce greater parity and so yes, they are probably going to give you tighter balance and be further on the balance spectrum. I would contend though that does depend on how you define balance and what you want from it. Personally I find random results produce balance by having a fair starting point (every one has the same odds) and by eliminating the likelihood of power builds (because you can't customize as easily when your stats are random). But yes, characters will have greater parity using the point buy method. That is a whole different argument from what folks are talking about here. What is being debated is why people like 4d6 drop the lowest and I think it is a little bit nuts that posters are essentially calling those who say they like 4d6 because its random, liars. This is what I am objecting to. If you find point buy more balanced, I am fine with that. We can disagree on some of the distinctions surrounding the word and still respect each other's opinions.

 

[Umbran]

What is being debated is why people like 4d6 drop the lowest and I think it is a little bit nuts that posters are essentially calling those who say they like 4d6 because its random, liars. This is what I am objecting to.

As well you should. It is a kind of jerkish thing to do - some folks seem to feel that since the folks so described are largely unnamed masses, it is okay to tell them that they are lying, or somehow wrong about their preferences - doing so to win an argument is one of those behaviors that give the internet a bad name.

 

[Hussar]

I dunno. I gotta go with Celebrim on this one. If randomness was truly being sought, then why is randomness only mitigated in one direction - ever upward? And, considerably upward.

Like I said, if you tried to ask your point buy DM if you could go with a lower point buy but add levels, you'd get laughed at. But, with random die rolled characters that almost always default to a higher and often much higher baseline, you get to add levels to your character right off the bat.

I think there is something here besides, "Oh, i just like randomly determining stats". If that's all there was too it, then why have so many extras that mitigate that. Heck, even back in Basic DnD, you got to swap points from stats (at 2:1) so you could get those exceptional stats on pretty much every character.

 

[Celebrim]

The array you give has an average value of 12.
4d6 drop lowest has, if I recall correctly, an average of 12.24

If and only if you don't reroll "hopeless" characters. If you have a table rule that says, "Reroll "hopeless" characters", then depending on what the bottom percentile of characters that are being excluded, the average is much higher than that. And if you don't have a table rule that says you reroll hopeless characters, you often have players working around that in some fashion - suiciding characters, cheating, creating multiple characters and then choosing which to play, etc.

So, I'm not sure how that's a "stomp". Taken straight, 4d6 drop lowest is, of course, more likely to generate high numbers - but it is also more likely to generate *low* numbers.

Again, mostly that matters (particularly for 1e) if in fact no rerolls are allowed. Once you start throwing out the garbage results, it's just more likely to generate high numbers, plus perhaps also some low numbers at the same time. Low numbers plus no high numbers typically means the character is discarded through some sort of social agreement or metagame activity.

My thesis since challenging this notion that random chargen was balanced is that random chargen is not balanced, creates large imbalances in play, and as a result pretty much everyone that used it (certainly everyone I've encountered and played with) over time developed some sort of metagame methodology to cope with the possible imbalances that actual randomness would necessarily produce. The result was random chargen with strong conditions that made it effectively unrandom or else while still random was irrelevant to the game the player would play. I haven't really argued that this secondary development was bad. It is in fact a functional response to the problems created by random chargen. Indeed, per Gygax's own discussion, the alternate methods in the DMG were an evolved response to the problems created by earlier more highly random chargen. They were however but one step in that evolution, because the popular Method I itself (for instance) still didn't control for the randomness enough.

The step I would like to take is to recognize that those sometimes unspoken metagame agreements actually existed and were in fact ways of evading the actual results of randomness. So for example, stating, "I like random chargen, but hopeless characters should be rerolled", suggests at minimum an insertion of choice and fudging into the random dynamic that shows actual randomness was being highly mitigated by a desire for greater balance. It also suggests, as I asserted earlier, that mathematical analysis of Method I is flawed compared to how it was actually used in play. In actual play, as I said, it often stood for the excuse to set stats in a more balanced manner, with more playable character, because the actual method wasn't method I, but "Method I until everyone gets (at least approximately) what they want." And that is not in fact random.

 

[Jhaelen]

Well, I'm with Celebrim regarding rolling for stats. We'd been playing for many years with randomly rolled stats, and what I've found was that our houserules to mitigate the randomness got more elaborate and complicated all the time. When 3e introduced the point-buy method, I used it and never looked back.

(Warning: anecdote coming in I had exactly one player who complained and wanted to roll dice. That player was known to be rather lucky when rolling his d6. Still, I told him: "Well, you get to roll up one character using 6x 4d6 and you have to take it, no matter what you roll." And this time he wasn't lucky. When he then asked if he could use point-buy instead, I said "Sure!". And that was the last time a player wanted to roll for stats. (Okay, anecdote over )

In 4e I even took it one step further and only allowed a single default stat array.

I'm not doubting Bedrockgames when he states, he enjoys the randomness, but in my experience what most players enjoy is the gamble, the hope to get something better than 'average'.

Personally, I've always been happy to get rid of randomness when creating characters given the chance. I'd even take point-buy options if they resulted in sub-par characters compared to the chosen dice roll method. And, yes, part of the reason is that I often suck at rolling stats!

Having said that, there's something I consider even more important than using point-buy instead of rolling dice for stats in D&D, and that's gaining a fixed amount of hit points every level instead of rolling dice! It's something that was standard in 4e, but is used in no other edition before or after. No matter how good or bad your stats are, rolling badly for your hit points three times in a row can cripple your character, especially the fighter-types who use large dice. I've been really disappointed that 5e doesn't have at least an option for this.

 

[Jhaelen]

And since Celebrim mentioned troupe-style play:

I'm a big fan of Ars Magica - actually it's my favorite rpg. And one big reason (of many) is the troupe-style play. In our games, creating the covenant, where all player characters reside, was a group effort. We'd brainstorm about what kind of characters we wanted to populate it and then distributed them, so everyone got to create a bunch of them in addition to a player's magus. And since our typical stories rarely involved more than one or two magi, every player seemed to be interested to carefully design the companions, experts, and grogs, as well. It certainly makes sense to make sure you're going to enjoy playing any type of character if you only get to play your magus every once in a while. For our group this resulted in some of the best roleplaying I've ever experienced.
It's definitely a great thing to have such a rich cast of characters available. Normally, only the GM has the luxury to play someone else each session. In our Ars Magica campaigns we would sometimes even play different characters in one session. I suppose it's not for every player, but for our group of experienced players who have all at some point been GMs, it was awesome.

Luckily, in Ars Magica, there's no rolling dice during character creation. You can freely choose your stats and buy advantages (and disadvantages) to get better (or worse) stats. And although I'm not generally a fan of boon/penalty systems since they can always be abused by min-maxers, we've not had any problems with it in Ars Magica so far, which I attribute to the troupe play.

 

[Celebrim]

I dunno. I gotta go with Celebrim on this one.

Thanks for the vote, but that doesn't give me a lot of confidence.

It's sincerely weird to have you agreeing with me.

If randomness was truly being sought, then why is randomness only mitigated in one direction - ever upward?

I think that there is a lot more to my argument than that, and it begins with, "Randomness isn't the goal in and of itself, but a means to some end."

In the 'old days', I think randomness was used because no one necessarily had better ideals about how to get to those ends. Randomness was the method we had, it wasn't seriously questioned, we didn't have the technical language to talk about it, it was a sacred cow, and so that's what you went with. But to make it actually functional, in my experience of play and in discussing it with other people, it was honored in the breach more than the use. The dice rolling creates the illusion of randomness, but some sort of social contract existed to mitigate that as an actual result while retaining some of the end values that randomness was the methodology used to create. "Throw out hopeless characters" so that the actual methodology wasn't Method 1, but rather Method 1 plus some number of rerolls until the player or table as a whole is happy with the character is one example of that. But, "Keep rolling until we get what we want.", as any reflective DM knows that has rerolled random encounter tables is, "Pretend to be random, but actually choose the outcome."

If you start looking at the stated primary goals of random character - the actual values we are prizing when we apply the methodology - it turns out that we have the methodology now for creating pretty much any of that while retaining balance.

For example, many have suggested that the like randomness for sparking diversity and preventing cookie cutter builds. Which is fine, but from a purely objective position, if what we valued most was diversity surely choice would produce more diversity than randomness would. After all, with choice you can always choose to play something different, but with random results you could be given a set of numbers similar enough to the last one to be effectively identical. If a player says, "I prize diversity over optimization and balance, but if I'm given a choice to create what I want, I always prioritize optimization over diversity.", what are we to make of that but at best and most charitably, "I like diversity, but the temptation to optimize is too great for me to over come so I need methodology that forces me to not do it". However, I've actually suggested methodology that would spark diversity and yield balance, yet this methodology remains unattractive. Clearly diversity as the end goal of randomness is at most secondary.

Others have said that the thrill of gambling is the reason that they like rolling dice. This is the theory that rolling dice is fun, so that's why random chargen is preferred. But the problem with that is that for most people, gambling is fun only when they win and that chargen carries no real cost. So in actual play, players didn't gamble once and accept losing as a happy result. They kept up the gambling until they produced some sort of 'winner'. The discarded the "hopeless" characters through some methodology. And so while we can honestly accept "the thrill of gambling" as being part of the attractiveness of the mechanic and part of the reason people honestly liked it, when we examine the actual impact on play that this thrill of gambling produced it was to produce a non-random set of playable "winners" who were perforce more balanced than randomness itself would have actually produced. Again, this isn't actual random, and we can't overlook the end state when discussing why people actually liked the methodology that they actually used.

I think there is something here besides, "Oh, i just like randomly determining stats"

As a DM, absolutely its fun to just randomly determine stuff - even stuff you'll never going to lose. There is definitely exploratory pleasure in randomly making stuff. But equally, absolutely there was and is more going on with random chargen than actual randomness. If there wasn't, there wouldn't be such elaborate table agreements and social constructs around protecting players from it.

Again, one of the most dominate aspects of random chargen is how much of the rerolling gets thrown out with the short term memory.

 

[Bedrockgames]

I think if you feel the need to ascribe hidden or false motives to people because they say they like rolling because it is random that shows a lack of respect in the conversation. This is a big problem in our hobby where it simply isn't enough to disagree over preferences but we also assign motivations to people that paint them as sinister or foolish. That is an easy thing to do. It is a simple thing to dissect a person's opinion in order to 'prove' that they want something other than they claim. In most cases I think, and here I will assign sinister motives, it isn't truly part of any effort to arrive at the truth but a product of kbe's own bias on the issue: the other side isn't just wrong---they are bad or stupid. Either way, I do not have a. Interest in continuing to humor that segment of the hobby (even those I agree with on issues) by responding to their point by points.

 

[Bedrockgames]

Again, mostly that matters (particularly for 1e) if in fact no rerolls are allowed. Once you start throwing out the garbage results, it's just more likely to generate high numbers, plus perhaps also some low numbers at the same time. Low numbers plus no high numbers typically means the character is discarded through some sort of social agreement or metagame activity.

I am only responding to this because this is part of the misunderstanding. People are re-rolling hopeless characters. That is they are re-rolling characters that who have an entire stat set deemed non-survivable in the game. They are not re-rolling individual results. There is a big difference between chucking a character because it has four bad results and rolling a set but re-rolling each result you don't like. Even if you are chucking hopeless characters, you are still getting the spikes and dips associated with a random rolling method.

 

[Bedrockgames]

I'm not doubting Bedrockgames when he states, he enjoys the randomness, but in my experience what most players enjoy is the gamble, the hope to get something better than 'average'.

I appreciate the good faith here. I would just add to this, these are not mutually exclusive things. I've already said part of the attraction is the hope of a good roll. I want to roll well, absolutely. Getting an 18 is exciting. But part of the requirement of that excitement is that bad results also be allowed to stand. I think the gambling comparison is a good. That is in fact one of the key draws to the random method for me. But it ceases to be exciting if I can re-roll every bad result I get. I'm fine with 4d6 drop the lowest. I am even okay with doing two sets if that is what the group really wants to do. But anything beyond that and the excitement starts to diminish considerably for me because the higher results become more of a foregone conclusion. So it isn't just about the high results. I want them, but for the excitement of the roll to mean anything to me, there needs to be the possibility of doing poorly. And when a bad result happens I can take it in good spirits and work it into the fun of the game (it is actually one of the unique challenges to rolling stats that sometimes you get stuck with this terrible number and need to make sense of it---having a 4 Dex is very different flavor wise than an 8).

I think these discussions are a lot more productive when folks do not assume people in favor of one thing or the other are not trying to pull a fast one.