D&D General Math Question re: 3d6 vs. 4d6 drop one

[Blue]

Two reminders.

1. The shape of the distribution is different. So adding a static number is a poor fit to simulate it. And it can raise the cap.

2. 5e rewards specialization. The difference between the average of 4d6K3 and 3d6 is irrelevant for this discussion. The primary point of interest is between the highest of six of each, then comparisons of the second highest of six from there, and then some minor quibbles, like how many penalty modifiers. Looking at it from the average roll will not mimic each other at all in practical terms within 5e.

 

[DND_Reborn]

Confirmed. My simulator was right, I just plain fat-fingered the result.

It happens.

 

[jgsugden]

I like my system, but it can only be used when players start a new campaign together:

• Tell your players to bring a dollar in small change, mostly pennies.
• Then, secretly roll (4d6 best 3) 8 times for each PC, with each score placed on a different index card (so you have 8 different cards for each player with a number from 3 to 18 on each).
• Now, add 5 to 20 more index cards. On each of these cards, write a fun little boon that a PC might have, like a low powered feat, an ally, a rarely used uncommon magic item, a +5 bonus to speed, a wild psionic talent, etc... You might allow an option of 2 boons to give some variability, or just put SECRET on it and let players discover the secret after they win the bid (I once did this to give a player an artifact at level one that they had to figure out how to protect, and how to activate).
• Shuffle the cards. Seat the players in a circle. Randomly pick a player to start bidding.
• Reveal a card. The player that starts the bidding must bid at least $0.01 for the revealed card. Go in a circle and bid per normal (fail to bid and you're out for that round). If you win the bid, you must be the starting bid for the next card if you have any cash left and are eligible to bid.
• You can bid on cards only so long as you do not have 6 numbers.
• If you ever run out of $ (and do not have 6 numbers), you get $0.01 more.

 

[Matt1984]

I am asking players to roll 3d6 4 times and 4d6 drop one 4 times and then dropping the lowest and highest of these sets (less than 7 more than once gets re rolled ) . In testing it gives a good balance of stats for a campaign where I don’t want them all starting out being good at everything .