D&D 5E How do you handle randomly rolling for stats

[CapnZapp]

Know what @ichabod ? I'm not here to force you to do something you clearly don't want to.

What is that? Face the facts. What you're doing is you're attempting to reduce my statements to "opinion", and that sort of anti-science stance isn't worth my time. I'm not having the "opinion" that risk is a cost, I'm telling you for a fact it is. Go check it up.

For anyone else interested:

In the short term, risk can be a cost, or it can be a benefit. That's the nature of probabilities.

No, what I'm talking about is that risk itself is a cost. I'm not just talking about the cost when the outcome is low. I'm talking about the variability itself.

Very crude example.

Imagine you're entering the dungeon and there's a goblin hidden. It will attack you first, for 2 points of damage. Then you attack and kill it. End of adventure.

Now I give you a choice: You can either start this adventure with 3 1/2 hit points or 1d6 hit points. What do you choose?

Too many people will go with the thrilling choice of possibly getting a whole six hit points, not caring about the equal possibility of getting just one.

But that's not the entire story here. In this (artificial) example, having 1 or 2 hit points is entirely worthless, because you'd be dead before you get to kill the goblin.

Which isn't so artificial after all, since having low hit points is much more of a disabler than having high hit points is a enabler. Perhaps not in all campaigns, but certainly in most.

If you only look at the cost of the outcomes, then you'd think 1d6 and 3,5 hp is equal.

But when you consider that 1d6 hp represents a 33% chance of certain doom while 3,5 hp completely avoids that, you might be able to understand what I mean: risk is a cost in itself.

In order for this risk to be worthwhile, the game needs to properly compensate for it. (In the above case, offering a choice of random hp is just plain bad - returning to D&D here) A full mathematical compensation isn't what I expect (and likely impossible to calculate) but "a little something" goes a long way to show me that the game designers are aware of this fact.

Which brings me back to the good and bad design of 5E: the default array giving you a slightly lower average than if you risk rolling the dice is a good design decision, since it (whether accidentally or intentionally) shows you're aware of this principle. I say "accidentally or intentionally" because if it was intentional, I don't think they'd gone with choosing "average" hit points rounding you up. It should definitely have meant rounding down. Not only do you not reward taking the risk (rolling for hp), you actively penalize it. This is straight-up, objectively, a bad design decision.

And to be clear: even if you could use fractional hit points, asking the player to choose between 3,5 and 1d6 hp would still fail to understand that the 1d6 choice represents a risk, and that risk needs to be compensated for - otherwise the other choice is straight up better.

I see the argument that they are convinced going for average hp leads to a better game in general, and so they want to encourage that. On the other hand, they realize they need to provide a random option. (I think 4E did away with random rolling, enough said) But that still does not excuse hanging out to dry all the people that don't care about statistics, the people that don't want to think about statistics, and the people that can't figure out statistics.

Which leads me to...

That's not my problem. Understanding the risks you are taking is part of playing a game. Understanding that a 4 is probably going to come up less often than a 5 in Settlers of Catan is part of playing the game. Understanding that rolling for hit points is probably going to leave you with less hit points at high levels is part of the game.

Thank you for so clearly showing you aren't game designer material.

Because this isn't about you. This is not your or my problem, this is, or should be, the game designer's problem.

Asking regular gamers to "understand the risks" is
1) incredibly dismissive to gamers
2) wildly underestimates how difficult the average gamer finds making proper probability calculations
3) wildly underestimates how little the average gamer even want to think about statistics, instead trusting the designer to provide him or her with reasonably weighted options
4) incredibly dismissive to the job that is game design. Asking gamers to do the risk calculation themselves is akin to "write your own scenario" where the writer just gives you a pitch and a rough outline, and leaves all the details to the DM to fill out. That's just not worth the money you pay to have someone write an adventure for you.

Lotteries and casinos take advantage of people, certainly, but not because they are hiding anything.

I guess we're different you and I.

I take it for granted that a game isn't out to fleece the unwary, the ones unable to resist thrills, and the math deficient.

When I talk about good game design, I'm talking about a game whose designers understand that (very generally speaking) a more risky (and by risky I mean variable) option needs to be given some kind of bonus, to be equal to a less risky option.

---

Finally, just to discuss one more point:

Obviously we can discuss several different examples. The more some outcomes are strictly worse than others, the more this principle applies.

Conversely, when you can't easily compare the outcomes, there is much less need to compensate for risk itself. Let's leave random ability scores for another time, and instead go for something as silly as it should be clear:

You can either roll for ice cream flavor randomly or you can choose vanilla. Assuming y'all don't have weird hangups against specific ice cream flavors , there is no real need to compensate for the "risky" choice here, since you're getting an ice cream either way.

Perhaps this point is best carried across by contrasting random hit points (where a 1 is strictly worse than a 6) with randomly selecting your character class, as discussed previously. You can't really say playing a bard or sorcerer is "unplayable". Is paladin really better than cleric? Is ranger really better than rogue?

In this case, there's no real need to compensate for the risk, since "fighter" isn't really strictly worse than "wizard" in the sense that you can't have a fun game and enjoy yourself. However, when you're at 0 hit points, most players aren't really enjoying themselves, and this is much more likely if you roll a 1 instead of a 4 or 6.

With my game dev hat on, I would STILL grant the player choosing to roll for random character class a token benefit (WFRP4 was brought up earlier) but I wouldn't call it actively bad design to not do that.

With something much more clearly risky, such as hit points, I definitely would. And do.

 

[Lanefan]

This isn't luck like rolling an attack. You do tons of rolls like that, so they average out over time.

This is luck that affects your character forever,

Well, for as long as the character lasts; which in an even moderately-lethal game might not be all that long.

so the impact of this one set of rolls is hugely magnified, and creates ongoing inequity. It's well and good to argue that this doesn't bother you personally, but it does bother a lot of other players, and it's not surprising. They are being punished over and over for one instance of bad luck, while the person sitting beside them is being rewarded over and over for one instance of good luck.

If one decides to focus on that, sure. But if one instead largely ignores that and just enjoys playing the character for what it is, it stops mattering.

It's a particular pain in the butt when you're working with a lot of teenagers, who are very quick to point out the structural inequity. Because they're right.

I'll risk saying that learning to deal with such inequities might be of benefit to them.

I switched to standard array about five years ago and am never going back. Contrary to some claims, it does not produce homogenous characters, because what makes each character unique and interesting comes from the player, not the stats. It just makes it so every character is starting from the same place mechanically.

That's just it, though: in real life, not everyone starts from the same place; and I want the game to reflect that. Some people just simply have more going for them than do others (and I don't mean socio-economically, that's a different issue); you know, the sort who have looks and brains and physique and talent, etc., as opposed to the rest of us normal schlubs who might have one or even two of those but not all of them.

 

[Lanefan]

No, what I'm talking about is that risk itself is a cost. I'm not just talking about the cost when the outcome is low. I'm talking about the variability itself.

Very crude example.

Imagine you're entering the dungeon and there's a goblin hidden. It will attack you first, for 2 points of damage. Then you attack and kill it. End of adventure.

Now I give you a choice: You can either start this adventure with 3 1/2 hit points or 1d6 hit points. What do you choose?

Too many people will go with the thrilling choice of possibly getting a whole six hit points, not caring about the equal possibility of getting just one.

But that's not the entire story here. In this (artificial) example, having 1 or 2 hit points is entirely worthless, because you'd be dead before you get to kill the goblin.

Which isn't so artificial after all, since having low hit points is much more of a disabler than having high hit points is a enabler. Perhaps not in all campaigns, but certainly in most.

If you only look at the cost of the outcomes, then you'd think 1d6 and 3,5 hp is equal.

But when you consider that 1d6 hp represents a 33% chance of certain doom while 3,5 hp completely avoids that, you might be able to understand what I mean: risk is a cost in itself.

While I get what you're saying with this analysis, I just gotta ask: what if that goblin is going to do 4 points damage instead of 2?

'Cause now all those who took the average are dead*, along with those who didn't roll 5 or 6 on the die. But those who took the risk and then rolled 5 or 6 are away to the races.

* - and if everyone did, that's the fastest TPK ever.

Thank you for so clearly showing you aren't game designer material.

Because this isn't about you. This is not your or my problem, this is, or should be, the game designer's problem.

The game designer's solution should be to point out in no uncertain terms that at its core D&D is a game based largely on luck, and due to that not everything is going to come up roses.

When I talk about good game design, I'm talking about a game whose designers understand that (very generally speaking) a more risky (and by risky I mean variable) option needs to be given some kind of bonus, to be equal to a less risky option.

If both options are available, I kind of agree.

My tack, though, would be to remove the no-risk option entirely.

Finally, just to discuss one more point:

Obviously we can discuss several different examples. The more some outcomes are strictly worse than others, the more this principle applies.

Conversely, when you can't easily compare the outcomes, there is much less need to compensate for risk itself. Let's leave random ability scores for another time, and instead go for something as silly as it should be clear:

You can either roll for ice cream flavor randomly or you can choose vanilla. Assuming y'all don't have weird hangups against specific ice cream flavors , there is no real need to compensate for the "risky" choice here, since you're getting an ice cream either way.

Perhaps this point is best carried across by contrasting random hit points (where a 1 is strictly worse than a 6) with randomly selecting your character class, as discussed previously. You can't really say playing a bard or sorcerer is "unplayable". Is paladin really better than cleric? Is ranger really better than rogue?

In this case, there's no real need to compensate for the risk, since "fighter" isn't really strictly worse than "wizard" in the sense that you can't have a fun game and enjoy yourself. However, when you're at 0 hit points, most players aren't really enjoying themselves, and this is much more likely if you roll a 1 instead of a 4 or 6.

Isn't the potential reward offered by the risk in itself enough compensation?

 

[Ogre Mage]

I prefer only 1, maybe 2 odd scores at start, too many instances to "fix" otherwise
17,15,14,12,12,10 would be preferred out of those two, same point buy budget as 17,15,14,13,11,10.

I remember that in both campaigns the person whose rolls got picked was not mine. The first time we did this I remember one player rolling badly and his set of scores were disregarded immediately, lol.

 

[CapnZapp]

Isn't the potential reward offered by the risk in itself enough compensation?

No, while the prospect of rolling a 6 mathematically compensates for the scenario where you roll a 1, it doesn't compensate you for taking the risk in the first place if you dont have to.

 

[CapnZapp]

If we agree rolling a 1 is more bad than rolling a 6 is good, what we're saying is the average isn't really 3,5 but lower.

But I'm talking about the fact that risk is in itself a cost.

For example, if we agree the "true" average of rolling a d6 for HP is only 3.1 and not 3.5 because a 1 is more disabling than a 6 is enabling, I would argue changing the choice to "either roll 1d6 or take 3 HP" might not be enough, because it sets the price of risk at only 0.1.

Not saying 0.1 can't be the proper cost of risk in this case, just that I'm talking about a different factor than either the mathematical average or the weighted average; I'm talking about the inherent cost of risk itself.

 

[CapnZapp]

That's just it, though: in real life, not everyone starts from the same place; and I want the game to reflect that. Some people just simply have more going for them than do others (and I don't mean socio-economically, that's a different issue); you know, the sort who have looks and brains and physique and talent, etc., as opposed to the rest of us normal schlubs who might have one or even two of those but not all of them.

While you're obviously free to like what you want, the argument here is that a game set in, say, Middle-Earth is much more suited to this line of thought.

While an Elf is assumed and even expected to be plain superior to a Man in many and varied regards, this simply does not hold true for D&D.

WotC has expended considerable energy through the years at leveling the field between characters. Having one character be clearly inferior (stats-wise) to another works against all this effort.

While most if not all games inspired by Tolkien (including all editions of WFRP) make it a point to make the Elf choice clearly mechanically superior.

My advice if you allow random rolling and therefore variable ability score totals? Make it clear that the shining all-capable heroes are supposed to contribute more effort than the one's saddled with various handicaps and deficiencies.

Just like it didn't need to be said out loud that Legolas and Gimli assumed a much larger share of danger than the Hobbits in LotR.

 

[Horwath]

@CapnZapp

I agree with you game of chance and risk and reward, but it should not apply to determining ability scores and/or HPs for D&D

D&D is already a big game of chance(risk) with d20, and d20 is treacherous enough without randomizing what bonuses/penalties come after that d20.

take a deck of cards for poker,
point buy/standard array is the standard deck of 52 cards.

and let's say that we roll for stats and HPs,
I roll very good and you roll very bad.

then we translate that into poker game in a way that because I have better stats/HP that the average or fair amount, I get to redraw every 2, 3 or 4 that I draw from the deck for free and you with lower stats/HPs get to redraw every ace, king or queen that you draw.

you cannot win in the long term, sure you can win a hand or two here and there because of numbers law, but in the long term(a D&D campaign) you will lose most of the hands.

now for D&D it might not matter to you, because it's all fun and roleplay, but most people do not like to play a game where they know that the deck is stacked against them from the start.

 

[Seramus]

I let my players roll, then every set of rolls becomes an option for every player. This definitely results in more powerful characters, but that is something I can easily account for as a DM. It also gives everyone the fun of rolling, but protects them from vast differences between sets (unless they intentionally pick a very low set for some reason).

 

[ezo]

You usually end up seeing one of three trends, in my experience.

For point-buy I always do 14, 14, 14, 10, 10, 10 or 15, 14, 14, 10, 10, 8.

This way I'll either have +3,+2,+2,+0,+0,+0 or +3,+3,+2,+0,+0,-1.