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

[FormerlyHemlock]

I'm pretty sure I'm remembering that right. It stood out last time I browsed the RC because I thought you were able to reduce any ability. I might have to reread to refresh my memory though.

Sent from my SM-G925I using EN World mobile app

Sorry, I didn't mean to imply I was doubting you. I was just literally answering a question about "who remembers", meaning to imply that "I think Red Box was somewhat like that, but slightly different, IIRC."

 

[cbwjm]

Sorry, I didn't mean to imply I was doubting you. I was just literally answering a question about "who remembers", meaning to imply that "I think Red Box was somewhat like that, but slightly different, IIRC."

Haha, no worries. I wasn't inferring that you were doubting me, it's just one of those things where you say something and someone say they don't remember it so it just makes you doubt yourself a bit. I'm still at least 90% certain that was a rule in the RC.

Sent from my SM-G925I using EN World mobile app

 

[Oofta]

(1) Please quote someone claiming that rolling will "always" give you "better" numbers. That's so obviously untrue that it's hard to imagine anyone even writing that, let alone it being the "biggest [most common?] lie on this forum."

(2) The math says the opposite of what you say: on average you will get better numbers than with point buy. 56% of the time, rolled stats will get at least one number better than ANY PHB point buy PC has. This is a well-known and provable mathematical result (http://anydice.com/articles/4d6-drop-lowest/). Furthermore, when I took your very own point-buy valuations and did a Monte Carlo sim rolling up a million PCs (https://repl.it/I9Sd/1), I consistently get rolled PCs who, according to your own valuations, are worth 31 points on average. You lose 4 points by using point-buy.

Are you claiming that 27 point-buy and 31 point-buy are about the same "on average"? (Of course, the rolled stats are allocated less efficiently, which is part of the point of using rolled stats--less min-maxing occurs.)

That you can claim with a straight face that "the math says" rolled stats should be "about the same" as point buy characters, and that rolled stats which aren't comparable to point buy are signs of cheating, is remarkable. Both closed-form math and Monte Carlo approaches agree: the math says the exact opposite.

I'm confused. You seem to be stating
A) No one claims that rolling for stats gives you better numbers.
B) Rolling for stats gives you consistently better numbers.

Which is it?

The very premise of this thread is that if you want a "good" character you have to roll for it. Over in the war cleric thread, it's stated that you "have to" start with 2 18's, and that to get it you have to roll. Many, many people justify rolling by saying that they want higher numbers. It probably works for them because they know most DMs will let them reroll, allow them to roll multiple characters, reroll low numbers. Or they'll ignore that 4 intelligence and state that their character is "a little slow" when in reality they should just about as mentally handicapped as is possible for a person to be barring illness or injury.

The only thing rolling dice for stats guarantees is that you will have random starting attributes for your character. Some people will win the stat lottery, others will lose. Rolling gives a slightly wider spread, but on average the numbers are about the same.

It's great that you like random stats. Yippee for you. I won't join your game, but that's OK.

 

[FormerlyHemlock]

I'm confused. You seem to be stating
A) No one claims that rolling for stats gives you better numbers.
B) Rolling for stats gives you consistently better numbers.

Oh, so you're just conflating "consistently" with "always"? The mean for 4d6 drop lowest is higher than point buy, and about 55% of rolled arrays (using 4d6 drop lowest) will have at least one 16+, which is better than the best stat in 100% of point buy arrays--but it's obviously incorrect that 4d6 drop lowest would "always" give you better rolls. You can roll all 3s on 4d6 drop lowest (and it would be awesome), which would constitute sometimes not being better than point buy.

Which is it?

It's (B). No one that I know of claims that rolling for stats "always" gives you better numbers (call that A*), per your original claim quoted below, but plenty of people including me claim that point buy gives you worse stats than rolling does, in the general case.

The biggest lie I see on this forum is that rolling will always give you "better" numbers. The math says otherwise. On average you should get about the same numbers with rolling as with point buy. Sometimes you will get better, sometimes you will get worse.

Now that I know how you meant "always", I can state unequivocally that you are just wrong here. The math says rolling will [consistently] give you better numbers, full stop. Anyone who says otherwise either doesn't know how to do math, or is making an obscure point about the inherent ambiguity of the word "better" in a system with no recognized total ordering. I don't think you're making an obscure point in the latter vein. Are you?

 

[Lanefan]

I don't like Ability Score Increases because this encourages generic character's that all end up with the same stats, especially when combined with the Point Buy or Standard Array systems. I am somewhere between Old School D&D that had no way to increase abilities, and New School which allows far too many increases. I am still trying to find the right balance.

What we did ages ago was take the percentile stat increment idea first seen in Cavaliers in the 1e Unearthed Arcana and apply it to all classes.

The write-up on how this works in my game is here: http://www.friendsofgravity.com/gam...blue-book-in-html/decbluebook3.html#increment

You can easily tweak the average rate of stat advancement by changing what dice get rolled each level.

Hope this helps!

Lanefan

 

[FrogReaver]

Please quote me in context. I was the one who said the 2 18's about the warcleric. But that was in the context of that stat spread making his abilities being as strong as other cleric subclass abilities.

I'm confused. You seem to be stating
A) No one claims that rolling for stats gives you better numbers.
B) Rolling for stats gives you consistently better numbers.

Which is it?

The very premise of this thread is that if you want a "good" character you have to roll for it. Over in the war cleric thread, it's stated that you "have to" start with 2 18's, and that to get it you have to roll. Many, many people justify rolling by saying that they want higher numbers. It probably works for them because they know most DMs will let them reroll, allow them to roll multiple characters, reroll low numbers. Or they'll ignore that 4 intelligence and state that their character is "a little slow" when in reality they should just about as mentally handicapped as is possible for a person to be barring illness or injury.

The only thing rolling dice for stats guarantees is that you will have random starting attributes for your character. Some people will win the stat lottery, others will lose. Rolling gives a slightly wider spread, but on average the numbers are about the same.

It's great that you like random stats. Yippee for you. I won't join your game, but that's OK.

 

[FrogReaver]

Maybe elaborate on what kind of point buy you are using? I assumed it was the kind that costed an amount to proceed to the next stat which is different than the phb method?

Statistically they will be stronger than regular point buy and weaker than 4d6 drop lowest. I just did 8 rolls to prove it, but you are welcome to do more testing if you so desire. Here are the results of the rolls, before allocating the 17 remaining points.

9, 13, 6, 13, 10, 9

8, 14, 11, 12, 10, 14

8, 12, 10, 11, 12, 10

14, 10, 8, 11, 14, 10

8, 10, 8, 9, 5, 16

9, 15, 16, 13, 5, 13

16, 14, 13, 12, 12, 9

13, 9, 13, 6, 10, 11

Only one of those rolls ends up worse off than standard point buy, but it also had a pre racial 16 which is not achievable through point buy.

 

[Lanefan]

Now that I know how you meant "always", I can state unequivocally that you are just wrong here. The math says rolling will [consistently] give you better numbers, full stop.

In one way, yes - as someone pointed out earlier, it'll give an average point-buy value of 31 vs 27, or something like that.

But you know what? I don't give a flippity frak about point buy values.

What rolling will give is a wider variance in individual results (range of 3-18 vs. 8-15) and a lesser but still significant variance between one character and the next.

What I want to know is whether the average sum total of the 6 numbers is more, less, or the same whether rolled or point-bought; and by how much. This is what tells me how powerful overall a set of stats are and whether, on average, rolled will be more or less or the same amount of powerful. If I care.

Then there's numeric variance within the 6 stats; with - generally speaking - greater variance giving greater interest and opportunity for characterizations.

12-12-12-12-12-12 (total 72) is kinda dull. 15-14-13-12-10-8 (total 72) isn't much better but at least there's something to grab for. 18-15-11-11-10-7 (total 72) - now there's something I can play! And I can only get it by rolling.

Lan-"and in 5e as written the stat increments by level make this all moot after a while anyway"-efan

 

[CTurbo]

Since this thread has ballooned to 16 pages now. I'm gonna throw my 2 cents in again. I really like 4d6 drop lowest a lot, but it has it's issues. The main issue is that you can get wildly unbalanced power levels between characters. This is a terrible thing for everybody including the DM.

I have played in campaigns when every player was way OP and it was a lot of fun.

I have played in campaigns when every player was well below average and it was also fun.

As long as the characters are roughly equal, it's all good. The DM can adjust up or down as needed. I have never seen a campaign with unbalanced characters work out very well. The trick is to figure out a way to level the playing field. I have seen some very good ideas in here about how to do that.

 

[FormerlyHemlock]

In one way, yes - as someone pointed out earlier, it'll give an average point-buy value of 31 vs 27, or something like that.

But you know what? I don't give a flippity frak about point buy values.

What rolling will give is a wider variance in individual results (range of 3-18 vs. 8-15) and a lesser but still significant variance between one character and the next.

What I want to know is whether the average sum total of the 6 numbers is more, less, or the same whether rolled or point-bought; and by how much. This is what tells me how powerful overall a set of stats are and whether, on average, rolled will be more or less or the same amount of powerful. If I care.

Then there's numeric variance within the 6 stats; with - generally speaking - greater variance giving greater interest and opportunity for characterizations.

12-12-12-12-12-12 (total 72) is kinda dull. 15-14-13-12-10-8 (total 72) isn't much better but at least there's something to grab for. 18-15-11-11-10-7 (total 72) - now there's something I can play! And I can only get it by rolling.

Lan-"and in 5e as written the stat increments by level make this all moot after a while anyway"-efan

Heh. It was for people like you that I added the aside about 'the inherent ambiguity of the word "better" in a system with no recognized total ordering.'

FWIW, I share your preference for 18-15-11-11-10-7 over 12-12-12-12-12-12 or 15-14-13-12-10-8; I think it is quite an interesting array and requires interesting choices to be made.