D&D 5E Why I Am Starting to Prefer 4d6 Drop the Lowest Over the Default Array.

[Oofta]

I don't think your comparison does justice to those that prefer random rolls. There's rarely a need to find a way to die in a randomly rolled game because every DM I've seen allows rerolls for stats they feel are too low.

I don't disagree - but I've had DMs that don't allow rerolls. When the player who rolled asked to reroll or use point buy he just laughed. In his mind it was "fair" because we had all rolled ... never mind that the result of the roll was not fair.

In addition if you do allow rerolls you aren't really using the standard 4d6 drop lowest.

 

[Satyrn]

Wouldn't that make them even more "samey"?

I wondered the same thing. I think they meant that the characters across campaigns won't be so "samey" even if the characters in a single campaign are.

 

[Zardnaar]

I don't disagree - but I've had DMs that don't allow rerolls. When the player who rolled asked to reroll or use point buy he just laughed. In his mind it was "fair" because we had all rolled ... never mind that the result of the roll was not fair.

In addition if you do allow rerolls you aren't really using the standard 4d6 drop lowest.

It is fair, I give PCs the option of using point buy or the default array as we use 4d6. If you don't want to gamble use them.

I will usually allow a reroll if you get something spectacularly bad or if there is a massive disparity between 1 PC and another. Otherwise you keep what you roll or use the default array.

 

[FormerlyHemlock]

Depends on how you come up with your numbers. Since the point buy system doesn't allow for numbers outside of 8-15, you will ultimately be putting your thumb on the scale one way or another when you decide what the cost of an 18 or a 3 is.

If you limit random rolls to 8-15, it's close (within a point or two).

The only thing rolling for stats guarantees is that the results will be random. In a fairly significant percentage of groups the difference will be significant. Some characters will have better than you could have gotten with point buy, some will be worse.

I find that if you're really serious about enforcing equality between your players, it's actually best to limit both point-buy and rolled stats to a range from 10 to 10. All PCs must be single-classed half-orc thieves, no feats allowed. Every time someone rolls a d20, the number on the d20 is taken down and then everyone else gets that same d20 result on their next roll, too.

Also, all HP go into a common pool, and when that common pool reaches 0, everybody dies.

</Tongue-in-cheek> But seriously--if you limit 4d6 drop lowest to 8 to 15, it is no longer 4d6 drop lowest.

 

[Oofta]

I find that if you're really serious about enforcing equality between your players, it's actually best to limit both point-buy and rolled stats to a range from 10 to 10. All PCs must be single-classed half-orc thieves, no feats allowed. Every time someone rolls a d20, the number on the d20 is taken down and then everyone else gets that same d20 result on their next roll, too.

Also, all HP go into a common pool, and when that common pool reaches 0, everybody dies.

</Tongue-in-cheek> But seriously--if you limit 4d6 drop lowest to 8 to 15, it is no longer 4d6 drop lowest.

You're the one who keeps insisting that rolling dice gives you better results on average. All I'm saying is that it's comparing apples and oranges. I have no idea where you get your "percentile" number because there is no way to make a comparison outside of the 8-15 range.

The only fact here is that the result of rolling dice is that you get random results.

 

[FormerlyHemlock]

You're the one who keeps insisting that rolling dice gives you better results on average. All I'm saying is that it's comparing apples and oranges. I have no idea where you get your "percentile" number because there is no way to make a comparison outside of the 8-15 range.

The only fact here is that the result of rolling dice is that you get random results.

Well, for starters, I used the point-buy valuations you created and found that 4d6 drop lowest averages 31 points, vs. 27 for point-buy.

I've also mentioned repeatedly that the modal result on 4d6 drop lowest (55% of the time you get at least 16+) is outright impossible with point-buy.

Lanefan and I had a brief tangent on the lack of a total ordering for stat arrays, but you didn't seem interested in joining that conversation then and it's too late now.

It is my opinion that WotC did this on purpose, making point-buy worse than 4d6 drop lowest, so that those who gamble are usually rewarded for it. If so I think it was rather clever of WotC--it's like running a gambling game where almost everybody comes out feeling like a winner. They're unlikely to be disappointed in the proprietor of that game, and will probably come back for me. Clever, IMO, given their business goals.

That is all.

 

[Oofta]

Well, for starters, I used the point-buy valuations you created and found that 4d6 drop lowest averages 31 points, vs. 27 for point-buy.

I've also mentioned repeatedly that the modal result on 4d6 drop lowest (55% of the time you get at least 16+) is outright impossible with point-buy.

Lanefan and I had a brief tangent on the lack of a total ordering for stat arrays, but you didn't seem interested in joining that conversation then and it's too late now.

It is my opinion that WotC did this on purpose, making point-buy worse than 4d6 drop lowest, so that those who gamble are usually rewarded for it. If so I think it was rather clever of WotC--it's like running a gambling game where almost everybody comes out feeling like a winner. They're unlikely to be disappointed in the proprietor of that game, and will probably come back for me. Clever, IMO, given their business goals.

That is all.

If you wanted to compare apples to apples you would probably have to go back to the 3.5 rules. But even then they didn't do the low numbers so the evaluation would be limited at best.

My "guesstimate" is only that and was just something I was using to get some rough numbers for comparison.

If I were to allow people to buy up to an 18 using my numbers I would give them more points. I don't think starting that high really buys anything so I don't.

 

[cbwjm]

While everyone argues about which is superior between point buy and 4d6 drop 1, I thought I'd ask if others have used any other methods. One I thought of trying was the matrix where the players roll stats in order until you have a 6x6 matrix and players choose different lines or columns to use for their PC. Another I've considered is rolling 18d6 (or maybe 20d6) and allowing players to assign the dice to their stats. So long as no score was less than 3 or higher than 18 before race modifiers the player can assign as many dice as they like to their stats.

 

[TheCosmicKid]

Depends on how you come up with your numbers. Since the point buy system doesn't allow for numbers outside of 8-15, you will ultimately be putting your thumb on the scale one way or another when you decide what the cost of an 18 or a 3 is.

Both you and @Hemlock are right about this. The 3E point buy values, and plausible extrapolations from the 5E values, do overvalue 17s and 18s. But even if you drop those values really low, the average of the random roll comes out ahead. The high scores simply aren't probable enough for their value to have a huge effect on the average: most of that figure is determined by the 8-15 range. It's almost the same situation for the low scores. Now, because you can give them negative values and go as low as you want, it is possible to push the random average to the same point value as the base array, or lower. But because those low numbers are really improbable, you have to go really low -- indisputably putting your thumb on the scale.

However, when you stop playing with point buy values and just look at the raw probabilities, you find that the average scores of 4d6k3, from highest to lowest, are [15.66, 14.17, 12.96, 11.76, 10.41, 8.50]. That's pretty darn close to [15, 14, 13, 12, 10, 8]. Average difference of about +0.24. Yeah, rolling is better, and if I really wanted to make the standard array fair compared to rolling I'd probably round up instead of down on that 15.66. But even as-is, it's not like the standard array sucks compared to the average rolls. The gap is pretty small, and I suspect Hemlock is overinterpreting it. It's literally a rounding error.

 

[FormerlyHemlock]

(Never mind, not worth the bother.)

 

[Lanefan]

However, when you stop playing with point buy values and just look at the raw probabilities, you find that the average scores of 4d6k3, from highest to lowest, are [15.66, 14.17, 12.96, 11.76, 10.41, 8.50]. That's pretty darn close to [15, 14, 13, 12, 10, 8]. Average difference of about +0.24. Yeah, rolling is better, and if I really wanted to make the standard array fair compared to rolling I'd probably round up instead of down on that 15.66. But even as-is, it's not like the standard array sucks compared to the average rolls. The gap is pretty small, and I suspect Hemlock is overinterpreting it. It's literally a rounding error.

So if the averages are more or less the same, the main difference between point-buy/standard array and rolling becomes one of variability, in three ways:

- variability within a single set of stats for one character. There's a big difference between a range of 8-15 and a range of 3-18 - someone else can do the math on how often a 4d6k3 roll will give a 3-7 or 16-18 result but I suspect many of us would find a 6 and 17 more fun to have as a couple of stats than 8 and 15. I also suspect that of the three variability types this is the least contentious.

- variability of the average stat for a single character. Do the 6 stats average out at 10, or 12, or 14? The standard array gives an average just over 12. Point-buy has a rather narrow possible range for averages; rolling is much wider. (another way of viewing this variable is by just looking at the total of the six stat rolls) Some people don't like playing characters whose total/average is significantly below the norm; and some even dislike playing with a total/average significantly above the norm. In the long run, however, it should all even out - you'll get your share of duds and your share of demigods - so one can always take the bad with the good and just go with it.

- variability of the total/average stats between different characters in the same party. This one seems to be the most contentious as some see it as leading to or directly causing imbalance within the party. This is probably of greater concern to those who like to fine-tune the party numbers to the nth degree to squeeze out every possible advantage, to those who play in a mechanics-first style, or to those who play in no-death and-or no-retirement games where one is stuck with one's initial character throughout; and for these point-buy or similar might be the better - though dull - option.

Lanefan

 

[Barolo]

I like to roll. I like the challenge of making effective characters out of mediocre stats, and I like the opportunity of making some "outrageous" stat-race-class combinations when I get lucky with the rolls (str & int 16+ orc warlocks FTW!!).

But mostly, I like the randomness that follows.

 

[TwoSix]

So if the averages are more or less the same, the main difference between point-buy/standard array and rolling becomes one of variability, in three ways:

- variability within a single set of stats for one character. There's a big difference between a range of 8-15 and a range of 3-18 - someone else can do the math on how often a 4d6k3 roll will give a 3-7 or 16-18 result but I suspect many of us would find a 6 and 17 more fun to have as a couple of stats than 8 and 15. I also suspect that of the three variability types this is the least contentious.

I think if I was looking for a method that was broadly ecumenical (impossible, I know, simply going by this thread!), this is the area I would focus on. A system to target a certain baseline of total stats, with the dice used to distribute the points. (I posted a method many pages back that does that.) My other personal favorite method uses Beyond the Wall's playbooks to create a truly random character background, complete with random (but balanced) stats.

 

[Oofta]

- variability of the total/average stats between different characters in the same party. This one seems to be the most contentious as some see it as leading to or directly causing imbalance within the party. This is probably of greater concern to those who like to fine-tune the party numbers to the nth degree to squeeze out every possible advantage, to those who play in a mechanics-first style, or to those who play in no-death and-or no-retirement games where one is stuck with one's initial character throughout; and for these point-buy or similar might be the better - though dull - option.

Lanefan

No. Just no. The reason I don't like variability in party members is not because of all the mildly insulting reasons you list. I'm trying to "squeeze out every possible advantage", "play in a mechanics-first style", and so on.

I do have to laugh at the "is stuck with" their character. Seriously? I'm "stuck with" a character I really enjoy playing? Oh heavens to Betsy how will I ever survive!

If you look at the people playing on a professional (American) football team, you have different skills. The quarterback is going to have different skills from the linebacker or a running back. But they are all going to be professionals. They are all going to be better than the average person in their chosen field. If you roll for stats you can get Tom Brady and "Bob", the guy that was a second string on his team in high school.

It's about letting every player pretend to be someone who is just a little better than average, that they can all contribute to the team on even footing. That there's no reason to force someone to play a character who has the stats of a commoner.

It's about playing the character I want to play. About player empowerment to build to a vision, not having that vision forced on them by a random result.

If I wanted random, I'd write up some arrays and roll for which array to use.

If I wanted a "power group" (which can be fun) I'd use the old 3.5 rules that let you buy up to an 18 and use the high power array of 32 points. I don't because I think the math of 5E works seems to work better with a 27 point buy.

I will say that if you deviate from the roll 4d6 drop lowest to give people characters they want to play you are "upping" the average value for most groups. Not all groups of course. I know some people love it when their highest stat is a 12.

As always, I don't get why people keep insisting that there is only "one true way" to play D&D. That as one person stated "rolling for stats is a good way of weeding out bad players" because bad players want to use point buy. Like to roll? Great! I just don't think it's inherently or numerically superior to the other system.

Anyway, [MENTION=6683613]TheCosmicKid[/MENTION], thanks for the analysis. I make no attempt to try to divine intent of the people who wrote the rule or came up with the numbers, I always assumed they did the math.

 

[Soul Stigma]

No. Just no. The reason I don't like variability in party members is not because of all the mildly insulting reasons you list. I'm trying to "squeeze out every possible advantage", "play in a mechanics-first style", and so on.

I do have to laugh at the "is stuck with" their character. Seriously? I'm "stuck with" a character I really enjoy playing? Oh heavens to Betsy how will I ever survive!

If you look at the people playing on a professional (American) football team, you have different skills. The quarterback is going to have different skills from the linebacker or a running back. But they are all going to be professionals. They are all going to be better than the average person in their chosen field. If you roll for stats you can get Tom Brady and "Bob", the guy that was a second string on his team in high school.

It's about letting every player pretend to be someone who is just a little better than average, that they can all contribute to the team on even footing. That there's no reason to force someone to play a character who has the stats of a commoner.

It's about playing the character I want to play. About player empowerment to build to a vision, not having that vision forced on them by a random result.

If I wanted random, I'd write up some arrays and roll for which array to use.

If I wanted a "power group" (which can be fun) I'd use the old 3.5 rules that let you buy up to an 18 and use the high power array of 32 points. I don't because I think the math of 5E works seems to work better with a 27 point buy.

I will say that if you deviate from the roll 4d6 drop lowest to give people characters they want to play you are "upping" the average value for most groups. Not all groups of course. I know some people love it when their highest stat is a 12.

As always, I don't get why people keep insisting that there is only "one true way" to play D&D. That as one person stated "rolling for stats is a good way of weeding out bad players" because bad players want to use point buy. Like to roll? Great! I just don't think it's inherently or numerically superior to the other system.

Anyway, [MENTION=6683613]TheCosmicKid[/MENTION], thanks for the analysis. I make no attempt to try to divine intent of the people who wrote the rule or came up with the numbers, I always assumed they did the math.

This is just it - nothing is superior when it comes to char gen, just different. I prefer random generation, but that's mostly because it harkens back to the old D&D I grew up on, which put stat generation first. After stat generation, you could determine with racial bonuses, etc., whether your character met the requirements of a given class.

I don't agree (having matured, I hope) with dice rolls preventing you from playing the class you want to play, so I appreciate the focus shift over the years to a "class first" mentality.


Sent from my iPhone using EN World mobile app

 

[MoonSong]

[MENTION=29398]Lanefan[/MENTION] [MENTION=6801845]Oofta[/MENTION] I think it is about splitting the difference, both methods are different and both have good and bad parts. On the one hand I like the idea of rolling, because it is unpredictable and there is the chance of low scores that just cannot happen under point buy and there is the chance of more variety, but at the same time, rolls tend to be just too high under 4d6, and a lot of the time I get a "bland" group of scores, without clear flaws or strengths. Last time I just asked if I could roll 3d6 instead, but ended up with too many high scores anyway.

Honestly I just want a character that is quite weak, not very resilient, kinda clumsy, not really smart, naive and somewhat likeable. Neither method really can give me that.

 

[Wulffolk]

People with both preferences have made valid points for their opinions ... and made them vigorously and repeatedly.

What I conclude from this conversation is that once DNA manipulation is a real possibility the entire human race will end up having the equivalent of the genetic standard array because nobody will want to play the genetic lottery and risk having any disadvantage. Everybody will end up all "same-y". Sure we might all start with 12's in every stat, but with the inevitable power-creep of this edition of Humans&History we will all eventually end up with 18's in everything.

 

[Oofta]

[MENTION=29398]Lanefan[/MENTION] [MENTION=6801845]Oofta[/MENTION] I think it is about splitting the difference, both methods are different and both have good and bad parts. On the one hand I like the idea of rolling, because it is unpredictable and there is the chance of low scores that just cannot happen under point buy and there is the chance of more variety, but at the same time, rolls tend to be just too high under 4d6, and a lot of the time I get a "bland" group of scores, without clear flaws or strengths. Last time I just asked if I could roll 3d6 instead, but ended up with too many high scores anyway.

Honestly I just want a character that is quite weak, not very resilient, kinda clumsy, not really smart, naive and somewhat likeable. Neither method really can give me that.

What's funny is that one of the reasons I don't like rolling is because of the chance of getting the super-high-stat character. They're kind of boring.

For the most part I don't base my flaws on stats because I come up with character concept first, everything else second. But ... if I have the concept of the absent-minded wizard who is brilliant but obtuse it's not going to work as well if my "low" stat is a 14 instead of an 8. I just don't go the extreme you seem to want.

 

[C4]

While everyone argues about which is superior between point buy and 4d6 drop 1, I thought I'd ask if others have used any other methods. One I thought of trying was the matrix where the players roll stats in order until you have a 6x6 matrix and players choose different lines or columns to use for their PC. Another I've considered is rolling 18d6 (or maybe 20d6) and allowing players to assign the dice to their stats. So long as no score was less than 3 or higher than 18 before race modifiers the player can assign as many dice as they like to their stats.

I'm a fan of point buy, but I get why others like randomization. So in my Points of Light game, I give players three options: a randomized array, an assigned array, or point buy. All three are equivalent numbers-wise, so it makes stat-gen a truly positive-sum game for everyone. My group loves it.

Really, the only reason that stat-gen is such a hot-button topic in D&D is because a lot of fans want not only randomized stats, but stats randomized in this specific way which demands unfair results. That's the aspect of D&D fandom which drives this zero-sum game, and the evergreen argument over point buy v. rolling. Until and if the fandom gets over the "It's not D&D unless I can roll a fistful of d6s for stat-gen" attitude, stat-gen will always be an argument regardless of which method is default or how many methods the PHB allows players to choose from.

Didn't mean to get on my soapbox there, just writing out loud.

 

[OB1]

If you look at the people playing on a professional (American) football team, you have different skills. The quarterback is going to have different skills from the linebacker or a running back. But they are all going to be professionals. They are all going to be better than the average person in their chosen field. If you roll for stats you can get Tom Brady and "Bob", the guy that was a second string on his team in high school.

It's about letting every player pretend to be someone who is just a little better than average, that they can all contribute to the team on even footing. That there's no reason to force someone to play a character who has the stats of a commoner.

It's about playing the character I want to play. About player empowerment to build to a vision, not having that vision forced on them by a random result.

While I understand what you are getting at, and I certainly don't want to tell you how to play your game how you want, but I'm not sure this analogy works at all. On every professional sports team, the range of players ability to contribute even among starters varies wildly. Take a look at the NBA finals game last Friday, where the Warriors top Starter scored 35, and the 5th place starter scored 6 and the average was 16.8. The Caveliers top starter had 40, the bottom 5 and the average for starters was 22.8.

Those differences don't come from training, they come from innate ability, and star players are just better than average players and way better than below average players in the NBA, but it still takes 5 players working together to win a game.

When I play with a team of randomly rolled players, I expect that sometimes the Star player will help keep my character alive and reaching goals and other times I'll save her bacon. Alone, even the super star would be toast.

I'm not saying that one style is "better" than the other. There are times when I like that feeling, as was described earlier in the thread, of a special forces team that has been bred for battle. Other times, I like the band of misfits who succeed despite not having all of the tools at their disposal that others do. Sometimes I like having the superstar character that has to worry about keeping the rest of the party alive, sometimes I like being the guy who needs a little help from his friends to get by.

I do find it interesting that the initial decision to roll these stats came out of a time when the draft was used extensively to supply the army during wars (WWI, WWII, Korea, Vietnam) and the expectation was that not every soldier was going to be exceptional, and as that has changed over the last 4 decades to a volunteer, dedicated army that recruits the best (Gulf War, Afghanistan, Iraq), it has been reflected in the playstyle of RPGs.