Best way to roll stats for new players

[Oramac]

Ok, let's assume you've got a table of all-new players. You're the only experienced person.

And let's also assume your players have decided, against all recommendation, that they absolutely want to roll their stats.

So what method do you give them to roll stats?

[Dudewithknives]

Ok, let's assume you've got a table of all-new players. You're the only experienced person.

And let's also assume your players have decided, against all recommendation, that they absolutely want to roll their stats.

So what method do you give them to roll stats?

I would go with the old roll 4D6, drop lowest.

If they reroll 1's, or roll 7 scores keeping the 6 best, that would leave them with higher scores than is normally built for in 5e.

5e is built around people working up to get an 18 and topping out at 20 later. Opens too many doors when people can start with them?

[Mortis_Elrod]

7d20 drop lowest. Each is separate ability score. Racials can't force over 20.

[Mith]

Personally, I do not think that high starting stats are terrible. The players feel awesome, and the DM can adjust encounters. The only problem comes if there is a big variance between ability scores of the players (Mortis_Elrond's suggestion would potentially create a huge variance).

With my groups last campaign, we did everyone roll an array that is placed into a 6x6 matrix. The table decides on which set of 6 they want (any row, column, or diagonal). Each player then rearranges the array as they want, add their racial bonuses, and go from there. This will lead to a higher than average stats for all the characters, but everyone is also equal to everyone else for stats in their respective classes, with slight variance due to what race is chosen. This also allows for the DM to more easily adjust encounters, as the minimal variance will keep the party close enough together that everyone feels like they are contributing.

[Trask]

In this edition, 4d6 drop the lowest straight down the line. So the first roll is str next is dex ect ect. This may be puzzling, but I have played with a LOT of new players and they all end up being stuck/hesitating/agonizing over what they should put in what stat and they end up turning to you for advise (naturally) and it just isnt a fun experience for someone new. They just want to play so they wont be moaning about the cruelty of in order stat rolling anyways. It also helps players who have no friggin idea what they want to play because everything is weird and D&D is new to sort of naturally find out what class would be good for them with the stats they got. Theres pretty much something for every combination.

[sir_argo]

The biggest problem with rolling stats is that it can be unbalancing. One player rolls lousy, while another rolls absolutely fantastic. So I like any method you want, with the caveat that each player gets to pick from any of the rolls on the table.

In other words, let's say you use the 4d6 drop the lowest; must have at least two 15's.

You have 5 players.

Each player rolls a set of stats and places them on the table.

Every player can pick from any of the 5 sets of stats. If one person rolled particularly well, it's a safe bet everyone will pick those stats, but some sets will be better for single attribute builds, while others are probably better for MAD builds. Either way, it is absolutely equal and fair since everyone has the same options.

[Trask]

The biggest problem with rolling stats is that it can be unbalancing. One player rolls lousy, while another rolls absolutely fantastic. So I like any method you want, with the caveat that each player gets to pick from any of the rolls on the table.

In other words, let's say you use the 4d6 drop the lowest; must have at least two 15's.

You have 5 players.

Each player rolls a set of stats and places them on the table.

Every player can pick from any of the 5 sets of stats. If one person rolled particularly well, it's a safe bet everyone will pick those stats, but some sets will be better for single attribute builds, while others are probably better for MAD builds. Either way, it is absolutely equal and fair since everyone has the same options.

I dont know if thats the best thing to do for very new players. Often times someone new to the game wont realize something is unbalanced or unfair unless you tell them it is, I think that method is just reinforcing in a burgeoning roleplayer that this a jealous and competitive game, when really someone rolling well doesnt penalize someone who rolled not as well. Most new players I've played with like rolling the dice and seeing what they get. And I think in 5e it doesnt really matter overmuch if one person got great stats as long as everyone has at least good stats. Just my 2 cp.

[PhoenixPhyre]

My next group of new players will either be standard array or 4d6b3, reroll if sum of bonuses is negative or highest is <=13. I prefer the standard array, but it really doesn't make that much of a difference as long as the difference between scores isn't tremendous.

[Kane0]

Every player rolls 4d6 and takes the best three. Take note of all the results. This becomes the array everyone will use.
If there are less than 6 players the DM rolls the remainder until 6 results are collected.

[SiCK_Boy]

I'd also just go with standard array.

This is the method we use all the time (new player or experienced one doesn't matter).

[Mortis_Elrod]

Personally, I do not think that high starting stats are terrible. The players feel awesome, and the DM can adjust encounters. The only problem comes if there is a big variance between ability scores of the players (Mortis_Elrond's suggestion would potentially create a huge variance).

With my groups last campaign, we did everyone roll an array that is placed into a 6x6 matrix. The table decides on which set of 6 they want (any row, column, or diagonal). Each player then rearranges the array as they want, add their racial bonuses, and go from there. This will lead to a higher than average stats for all the characters, but everyone is also equal to everyone else for stats in their respective classes, with slight variance due to what race is chosen. This also allows for the DM to more easily adjust encounters, as the minimal variance will keep the party close enough together that everyone feels like they are contributing.

Gonna completely agree here with everything said. What I suggested is what I use for one shots (more like half shots ) to make a character really fast , so I can show someone the basics.

In fact I use your rolling method for all of my campaigns, only I let the group decide if they want to roll individual matrices or one matrix.

[CantigThimble]

4d6 best 3, handwave some really bad or really good arrays. Is it perfect? No, but it's fast, easy and helps to introduce some variance between arrays, which I think is a good thing.

[Pex]

I would go with the old roll 4D6, drop lowest.

If they reroll 1's, or roll 7 scores keeping the 6 best, that would leave them with higher scores than is normally built for in 5e.

5e is built around people working up to get an 18 and topping out at 20 later. Opens too many doors when people can start with them?

5E is built around Bounded Accuracy. Point Buy is the variant. Dice rolling is the rule. Dice rolling allows for the lucky chance of rolling an 18. Further, you can put that 18 into a particular stat where a racial modifier makes it 20. At 1st level. It is not a crime against gamedom for a 1st level character to have an 18. The game does not cease to function.

Just because a player is new doesn't make him an imbecile needing handholding for everything. The new player should read the rules himself and learn them. Barring that, the normal roll 4d6 drop lowest arrange as desired is fine. Then talk with the player on what he envisions his character doing and help him assign his scores accordingly. Explain why a score is going into a particular spot to teach.

[McNinja]

In this edition, 4d6 drop the lowest straight down the line. So the first roll is str next is dex ect ect. This may be puzzling, but I have played with a LOT of new players and they all end up being stuck/hesitating/agonizing over what they should put in what stat and they end up turning to you for advise (naturally) and it just isnt a fun experience for someone new. They just want to play so they wont be moaning about the cruelty of in order stat rolling anyways. It also helps players who have no friggin idea what they want to play because everything is weird and D&D is new to sort of naturally find out what class would be good for them with the stats they got. Theres pretty much something for every combination.

Straight down the line is a surefire way to piss off a new player. No one, especially a new player, wants to play a warlock with 10 CHA and 17 CON or a fighter with 11 STR and 16 INT. If your players have having issues deciding where to put what stats, then they either have literally never read the book, or you haven't. I can and have told new players where to put their stats based on their class without looking at the book, although going through the book with them does help.

I remember Spoony talking about how his AD&D groups used to do 3d6, down the line. That's hardcore, but hardcore doesn't necessarily equate to fun.

[JellyPooga]

I want to suggest 3d6 "down the line"; old school, hardcore. It's actually a good way to introduce people to roleplaying because it encourages them to a) be creative with who and what they play; taking away some of the options because the stats don't fit helps newer players to make their decision ("Hmm, I've got terrible Wisdom, that rules out Cleric, Druid and Ranger, what's left?") and b) actually roleplay their stats because you inevitably end up with a good variety rather than just being all-round above average.

I would recommend offering players to roll several "lines" of stats and choose the one they like; 3 or 4 usually produces at least one or two decent arrays.

[McNinja]

I want to suggest 3d6 "down the line"; old school, hardcore. It's actually a good way to introduce people to roleplaying because it encourages them to a) be creative with who and what they play; taking away some of the options because the stats don't fit helps newer players to make their decision ("Hmm, I've got terrible Wisdom, that rules out Cleric, Druid and Ranger, what's left?") and b) actually roleplay their stats because you inevitably end up with a good variety rather than just being all-round above average.

I would recommend offering players to roll several "lines" of stats and choose the one they like; 3 or 4 usually produces at least one or two decent arrays.
First, rolling 3 or 4 arrays isn't really old school or hardcore, it's just rolling a lot of dice and hoping you don't roll crap. Second, I don't think you should he taking away options from new players.

[Blue Lantern]

Roll 3d6, drop the lowest Add 4.

It will give you very similar bell curve to the standard 4d6 drop lowest, but it avoids the pitfalls of extreme rolls on both sides.

[StoicLeaf]

standard array, if they refuse to listen to you and demand to be treated like non-newbie grown-ups, then they roll 6x(4d6 drop lowest).
Live by the dice, die by the dice.
When that inevitably happens, offer them the standard array!

[tieren]

Roll 3d6, drop the lowest Add 4.

It will give you very similar bell curve to the standard 4d6 drop lowest, but it avoids the pitfalls of extreme rolls on both sides.

I was going to suggest 3d4+6, should give a decent array with very few negatives but also very hard to get an 18.

[Trask]

First, rolling 3 or 4 arrays isn't really old school or hardcore, it's just rolling a lot of dice and hoping you don't roll crap. Second, I don't think you should he taking away options from new players.

In my experience new players are not concerned with options. They just want to get in and play, and become frustrated when it takes a long time or they have to make a bunch of decisions. This isnt them being stupid or lazy, its just their expectations of what this game would be. Its hard for us to see because experienced players can make a new character like a breeze and know all the stats by heart. New people just arent as receptive to the character creation process and its small comfort if you tell them "dont worry, im preserving your ability to customize your character and give you options!" (ESPECIALLY people who are not only new to rpgs but new to games in general). I think youre conflating a little too much of the importance of GitP optimizer would put on options and customization than a totally new player would who wants to play a half elf wizard or a tiefling fighter and isnt thinking about whats "best".

I think it could actually be damaging to a new player to redirect their attention to the best "options" all the time and not let them figure it out on their own.

[LtDarien]

Straight down the line is a surefire way to piss off a new player. No one, especially a new player, wants to play a warlock with 10 CHA and 17 CON or a fighter with 11 STR and 16 INT. If your players have having issues deciding where to put what stats, then they either have literally never read the book, or you haven't. I can and have told new players where to put their stats based on their class without looking at the book, although going through the book with them does help.

I remember Spoony talking about how his AD&D groups used to do 3d6, down the line. That's hardcore, but hardcore doesn't necessarily equate to fun.

Well, if you were going to do 4d6 down the line I would have players choose race/class after rolling stats.

[Blue Lantern]

I was going to suggest 3d4+6, should give a decent array with very few negatives but also very hard to get an 18.

It can work as well, the bell curve top just slightly above mine. The thing I don't like too much about your method is that it would be almost impossible to get a negative score, and the minimum is a 9. Personally I like dealing with, and especially roleplaying, low scores. But is a matter of personal preferences I guess.

[KorvinStarmast]

The biggest problem with rolling stats is that it can be unbalancing. They aren't playing against each other.

[hymer]

How about rolling 1d2+7, 1d2+9, 1d2+11, 1d2+12, 1d2+13, and 1d2+14? Arrange as you please.


They aren't playing against each other.

That doesn't mean lack of some sort of party balance won't cause trouble. For new players it may warp their mind sets about the whole game, causing them sit around hoping to get a powerful array the next time around so they can get theirs. A lack of balance fosters power play and accentuates differences between PCs, and so can make your statement less clear cut, however admirable it is.

[PhoenixPhyre]

They aren't playing against each other.

But it can produce resentment if one is much more effective than another. I once rolled a character: 7 7 7 9 10 13. Made him a warlock. He sucked. Rolling super low while others have super high scores creates a perception of unfairness. Even if in practice the differences are small, the perception is toxic.

[Blue Lantern]

How about rolling 1d2+7, 1d2+9, 1d2+11, 1d2+12, 1d2+13, and 1d2+14? Arrange as you please.

The famous coin shaped dice.

[hymer]

The famous coin shaped dice.

I suppose a d1 would make it even more balanced. :smallsmile: But I think I'll amend my previous suggestion, and instead make it 1d2+7, 1d2+9, 2 x 1d2+11, and 2 x 1d2+13

[Cap'm Bubbles]

Well, first off, NEVER force a new player to roll down the line. If they want to because they are completely indecisive about what to play, fine, but never force it. If my DM had done that to me when I first started, I would have walked out, because I could end up either playing a crappy version of what I want to play, or find that the stats only work for something I'd have no interest in. If down the line gives you INT and CON, your new player might be overwhelmed with all that a wizard has to keep track of, or will find that they are doing worse than their peers in anything else they try to play. It's one thing to choose a sub-optimal race for a class, but it's another thing entirely to be told that the class you wanted to player will suck because these stats can't be moved around.

Personally, I'd say keep it simple, and always let players use any of the rolls made by anyone else during character creation. 4d6 and keep the best of 3 usually works, but can still end up lousy. I personally would say, just for a brand new group, to ignore the stat cap of 20 and try something more adventurous like 6 d20's or 3d8. Are the players overpowered? Technically yes, but experience will show you that it will be some time before they know how to abuse that. Adjust your encounters accordingly, and play until they find they want a new character or to hand off the DM seat to someone else (for variety's sake). Don't kid yourself that you're going to take an entire group of new players from level 1 to level 20 in this day and age; you'll play until interests change, and then you can all agree on what to do for stats next time.

[KorvinStarmast]

But it can produce resentment if one is much more effective than another. That's weird, and makes me wonder what's wrong with players who take this attitude.

We didn't have these problems in the old editions. Seriously. We had fun no matter what we rolled up. We played in parties that tried to work together. (The game was lethal enough without having dysfunction in the party).

If the character was well below average (3d6 top to bottom, or 4d6d1 top to bottom, or 4d6d1 mix and match) most DM's would have them roll again.

Also, in OD&D and in Basic you could trade points in one stat on a 2 for 1 or 3 for 1 basis if you wanted to.

The Best Way to roll stats for a new player is exactly what is says in the Basic Game. (Chapter 1, page 7)

You generate your character’s six ability scores randomly. Roll four 6-sided dice and record the total of the highest three dice on a piece of scratch paper. Do this five more times, so that you have six numbers.

[PhoenixPhyre]

That's weird, and makes me wonder what's wrong with players who take this attitude.

We didn't have these problems in the old editions. Seriously. We had fun no matter what we rolled up. We played in parties that tried to work together. (The game was lethal enough without having dysfunction in the party).

If the character was well below average (3d6 top to bottom, or 4d6d1 top to bottom, or 4d6d1 mix and match) most DM's would have them roll again.

Also, in OD&D and in Basic you could trade points in one stat on a 2 for 1 or 3 for 1 basis if you wanted to.

Part is natural. Being less effective due to random chance you can do nothing about is not too much fun. Although if the mean lifetime of a character is short, it's not a big deal.

Part is a difference of expectations/style. You expected to roll up lots of characters. That way you win some, lose some. That's not the mode most people use these days. Neither is wrong, just different and incompatible.

[Capt Spanner]

With my new-player-heavy game I used a point buy. It keeps the party together, in power levels, without any outliers. (I do roll down the line for NPCs, but the players don't need to know that.)

See: even if everyone gets reasonable rolls (or you fudge the rolling system to avoid bad rolls) the issue of one player who has rolled very well, and can dominate the party's effectiveness, still creates issues.

I think the experience of playing a character who has rolled very well or very badly is an interesting challenge for an experienced player, but for a new player it's an extra level of complexity (i.e. "how do I mitigate lots of low rolls, and still be effective despite another character being statistically much better than me?") which might make learning the game harder. Also, more experienced players who roll very well will (hopefully!) be more aware of how to play that effectively without dominating the spotlight. (Although I've certainly read plenty of horror stories to suggest otherwise.)

Once players are familiar with the core rules of the game, then I'm happier to let them roll, and deal with the results.

[Cybren]

Yes he just... roll stats normally. If someone seems like they got boned by the dice, don't be a jerk, but if everyone is happy with their result, then let them take their result

[MrFahrenheit]

If your entire party is new, I'd recommend the 4d6 method, but add up the players' individual totals, and then look for whoever got closest to 75, without going below it. Have everyone use that person's array. If multiple people got the same total above, but closest to, 75, then there can be multiple options.

(75=12+12+12+13+13+13, which is the highest total you can get from standard point buy).

[Jophiel]

That's weird, and makes me wonder what's wrong with players who take this attitude.

We didn't have these problems in the old editions. Seriously. We had fun no matter what we rolled up. We played in parties that tried to work together. (The game was lethal enough without having dysfunction in the party).
I jumped right from 1st edition to 5th so I couldn't tell you where the change happened but 1st edition (in my experience) relied a lot less on rolling against ability scores. These days you have a bunch of skills and DMs seem to make use of them having you regularly roll against your Survival or Perception or Persuasion. Back in the day, you might have one or two ability checks per session -- roll your Dex or under to cross a slippery bridge, for example -- but now you're often rolling DC checks on a regular basis. So it stands out more when one character has bad stats across the board because he's getting tested against those stats much more often.

Agreed about working together but it's harder, and less fun, to feel like a part of a well-oiled adventuring machine when you have little to contribute from a stats standpoint, especially if the player has good ideas but not the numbers to pull it off (deceiving a guard, sneaking into a building, scaling a wall...)

[MrFahrenheit]

I jumped right from 1st edition to 5th so I couldn't tell you where the change happened but 1st edition (in my experience) relied a lot less on rolling against ability scores. These days you have a bunch of skills and DMs seem to make use of them having you regularly roll against your Survival or Perception or Persuasion. Back in the day, you might have one or two ability checks per session -- roll your Dex or under to cross a slippery bridge, for example -- but now you're often rolling DC checks on a regular basis. So it stands out more when one character has bad stats across the board because he's getting tested against those stats much more often.

Agreed about working together but it's harder, and less fun, to feel like a part of a well-oiled adventuring machine when you have little to contribute from a stats standpoint, especially if the player has good ideas but not the numbers to pull it off (deceiving a guard, sneaking into a building, scaling a wall...)

As someone who never played before 3.5, is your experience with 1st because it was extremely combat heavy, or was there just un-statted rp?

[Jophiel]

As someone who never played before 3.5, is your experience with 1st because it was extremely combat heavy, or was there just un-statted rp?
Little bit of both. More was handled via straight up role-playing. If a character tried to trick a guard, the DM would usually weigh it by what the player said than demand a charisma check. If a Druid wanted to identify a plant, the DM would usually just allow it rather than rolling against wisdom or intelligence. I mean, there definitely WERE checks for stuff but they were often handled via charts (like the thief tables for lockpicking) or standard X-in-Y style rolls like a 1-in-6 chance of finding a secret door or being surprised versus blowing Perception checks with a low wisdom. Straight up rolling against your attributes was fairly uncommon. Not rare, but uncommon.

There was also -- again, just my experience -- more combat and dungeon-clearing going on. That could very well be a function of the lack of defined skills; once DMs felt like they had the tools, they could test characters in different ways. Or maybe DMs just got into a trap of thinking every time you picked an apple off a tree, it deserved a Survival check. I don't really have a strong opinion on which is "better". At this point the two games are far enough removed that they're different entities.

[MrFahrenheit]

Little bit of both. More was handled via straight up role-playing. If a character tried to trick a guard, the DM would usually weigh it by what the player said than demand a charisma check. If a Druid wanted to identify a plant, the DM would usually just allow it rather than rolling against wisdom or intelligence. I mean, there definitely WERE checks for stuff but they were often handled via charts (like the thief tables for lockpicking) or standard X-in-Y style rolls like a 1-in-6 chance of finding a secret door. Straight up rolling against your attributes was fairly uncommon. Not rare, but uncommon

I think the combat-heavy dungeon crawling aspect persisted till this edition. I find it far easier to run an rp-heavy campaign than ever before (my experience includes 3.5, 4th, Earthdawn and GURPS).

So here's where I have a disconnect: one of my players is a lawyer irl. He's very persuasive. But if he has a character with low int and low cha, it wouldn't seem plausible in-game that he can trick/convince NPCs on a regular basis to do what he wants. For me, that's where 5e gives me the ability to say "That's great, and you make a convincing argument. However, your character is far less intelligent and eloquent than you are. roll (relevant check)." I may give him advantage, but that'd be about it.

[Jophiel]

Oh yeah, well obviously every DM is different, etc. I would suspect that most would question the half-orc with a 6 Charisma sweet talking the elf patrol regardless of what the player said. But that's more likely handled with a "Haha nope" from the DM than a Charisma check. More executive decision making on the DM's part rather than dice rolling. Which might be why you did more combat -- less work for the DM and less arguments when you're rolling D20's against an attack matrix.

Anyway, without getting too deep into the merits of either situation, my primary point was that you felt less crippled with a poor stat array in 1st edition because you were leaning on your stats less often and less likely to have a good idea get dorked because you have a -1 to your DC check. You still didn't want to play a fighter with 8 strength or anything but it was less blatant if you had average or below average stats and someone else had above average rolls.

[Rhedyn]

Ok, let's assume you've got a table of all-new players. You're the only experienced person.

And let's also assume your players have decided, against all recommendation, that they absolutely want to roll their stats.

So what method do you give them to roll stats?
Point buy.

Tell them no.

[PhoenixPhyre]

Oh yeah, well obviously every DM is different, etc. I would suspect that most would question the half-orc with a 6 Charisma sweet talking the elf patrol regardless of what the player said. But that's more likely handled with a "Haha nope" from the DM than a Charisma check. More executive decision making on the DM's part rather than dice rolling. Which might be why you did more combat -- less work for the DM and less arguments when you're rolling D20's against an attack matrix.

Anyway, without getting too deep into the merits of either situation, my primary point was that you felt less crippled with a poor stat array in 1st edition because you were leaning on your stats less often and less likely to have a good idea get dorked because you have a -1 to your DC check. You still didn't want to play a fighter with 8 strength or anything but it was less blatant if you had average or below average stats and someone else had above average rolls.

And this is one reason why I strongly recommend that "experienced" players (experienced with previous editions) put their knowledge aside and play 5e for what it is. Not importing stuff from previous editions. Not trying to play 1e with 5e rules (or 3.X with 5e rules). Even though things have the same name, they're different and the differences matter.

In 1e (and 2e to a large extent), ability scores meant little because they had only a small amount of mechanical weight and weren't used directly much. Player knowledge/skill mattered a whole lot more. Random-rolled stats (including 3d6 straight down) worked well here--gave randomness but had little mechanical import.

3.X changed that paradigm--now stats matter quite a lot and scale to significant levels. Everything was now tied to ability scores. Rolling skill checks became a constant thing (roll to get out of bed. Now roll to open the door. Point buy (optimize that array!) became a huge thing.

4e carried this "stats matter" trend along, but did so in a completely different way where only a couple stats matter for each class and scores also scale with levels. Still rolling quite a bit. Point buy is still king.

5e moved back toward 1e and 2e a bit--now scores don't scale as much and the d20 has a bigger effect. It's still noticeable and there are still lots of things tied to the ability scores. Skill checks can (should) be less frequent, as there aren't defined DCs for everything and the DMG suggests that most things shouldn't take rolls. Climbing, for example. Or jumping. Most of the time, those just happen (at half speed for climbing, up to a fixed distance based on STR for jumping). 4d6b3 works well, as does the standard array, as does point buy. Unlike 1e and 2e, large disparities can still make a big difference here. Just not as much of one as 3.X or 4.

[KorvinStarmast]

In 1e (and 2e to a large extent), ability scores meant little because they had only a small amount of mechanical weight and weren't used directly much. Nope. I think you are referring to OD&D.
By AD&D 1e, the +/- to various game things: hit, damage, armor class, saving throws, HD, morale adjustments, etc. were applied to all six ability scores. There were a host of mechanical advantages and disadvantages.
The original one was +1 to missile attacks for a dex greater than 12.
Greyhawk provided: +defense for Fighting Men based on high dex (above 13)
The strength bonus numbers
The con bonus numbers.
Ability checks were made as far back as 1e using a d20 versus a score. Roll below and succeed at what you were trying ... common at a lot of table I played at.
2e went far deeper with non weapons proficiencies, mostly I think due to fan feedback and IIRC quite a few articles in Dragon.

[PhoenixPhyre]

Nope. I think you are referring to OD&D.
By AD&D 1e, the +/- to various game things: hit, damage, armor class, saving throws, HD, morale adjustments, etc. were applied to all six ability scores. There were a host of mechanical advantages and disadvantages.
The original one was +1 to missile attacks for a dex greater than 12.
Greyhawk provided: +defense for Fighting Men based on high dex (above 13)
The strength bonus numbers
The con bonus numbers.
Ability checks were made as far back as 1e using a d20 versus a score. Roll below and succeed at what you were trying ... common at a lot of table I played at.
2e went far deeper with non weapons proficiencies, mostly I think due to fan feedback and IIRC quite a few articles in Dragon.

But overall, the differences between a X and an X+2 weren't so big and (more importantly) didn't have a continuous scale. They were more binary. Also, the bonuses started much higher (and the penalties started much lower). 7-14 were pretty much no difference. That makes 3d6 straight down much better than having more granular bonuses/penalties. The big effect was controlling what class you got into, since there were minimums.

There was also a cap--for most things 20 was as high as it went under normal conditions. Much like 5e, and very different from 3.X and 4.

[KorvinStarmast]

But overall, the differences between a X and an X+2 weren't so big and (more importantly) didn't have a continuous scale. They were more binary. Also, the bonuses started much higher (and the penalties started much lower). 7-14 were pretty much no difference. That makes 3d6 straight down much better than having more granular bonuses/penalties. The big effect was controlling what class you got into, since there were minimums.

There was also a cap--for most things 20 was as high as it went under normal conditions. Much like 5e, and very different from 3.X and 4. Good point on the minimums for entry.

IIRC, 19 was the cap, not 20. (Without a magical item like frost giant strength girdle ..) You could get a 19 by rolling an 18 in a score and being an elf ... (Dexterity)
elven characters odd a bonus of + 1 to their initial dexterity score.


7-14 no difference? That's just plain wrong. Did you actually play AD&D 1e?
Str 7 -1 to hit
Str 14 no penalty, but better open doors, +200 weight allowance, 7& lift gates/bend bars versus 7's zero.

7 Int: you can't be a magic user
14 Int: you can cast up to 7th level spells.

And so on.

[Tanarii]

Ok, let's assume you've got a table of all-new players. You're the only experienced person.

And let's also assume your players have decided, against all recommendation, that they absolutely want to roll their stats.

So what method do you give them to roll stats?For new players:
Select class and race first per the PHB order.
4d6 best 3, 6 times. Or choose standard array. Decide before rolling. Arrange as desired.
No feats or multiclassing.

[orange74]

2d6, drop highest, in order, pick race and class first. That'll learn 'em to deviate from the ONE TRUE PATH.

[smcmike]

For new players:
Select class and race first per the PHB order.
4d6 best 3, 6 times. Or choose standard array. Decide before rolling. Arrange as desired.
No feats or multiclassing.

I'd let them roll, then allow standarf array as a backstop against poor rolls.

Ban variant human, and allow feats at level 4, as normal. Allow multiclassing, but don't encourage it.

[Theodoxus]

For new to me players, regardless of their claimed bonafides (because really, it's hard to prove how long you've actually played the game), I always use standard array, simply because I don't have to worry if Johnny has a 16 in his primary stat, or possibly a 20. It makes running from the screen side of the game easier - AC, to Hit, Saves and Save DCs are a lot easier to remember/track when using standard array.

However, for an established group, I'm beginning to use the average of 4d20D1, x6 for stats. But then, I'm on a 3d20 kick, loving the averaged rolls and simplicity for determining advantage and disadvantage. It's been a fun side project.

[Tanarii]

I'd let them roll, then allow standarf array as a backstop against poor rolls.I don't because:
1) That's not what the rules say.
2) That results in above intended stats.


Ban variant human, and allow feats at level 4, as normal. Allow multiclassing, but don't encourage it.My game is no feat, no multiclassing. So that's actually the rules for all players, not just new players.

[smcmike]

I don't because:
1) That's not what the rules say.
2) That results in above intended stats.

My game is no feat, no multiclassing. So that's actually the rules for all players, not just new players.

Fair enough. Not saying your way is wrong! I just don't like the idea of a new player taking the risk of rolling the dice, then regretting it. Feats and multiclassing are fun, but to each their own.

[KorvinStarmast]

Our first 5e campaign was:
roll 4d6d1
Arrange
all PHB races/classes available
No variant human
No feats until the first ASI: 4th level.

We started again with the same DM in a shared world where we now have two, and are working on 4 DM's.

Same as above.
Variant Human is fine.
Exception: A few PC races in my brother's world are forbidden: Tieflings, Dragonborn, Drow. Since he began the world, I doubt any of us will pick one of those races anyway ... it might be a world wide ban, or just a ban on the continent where we started.

[Tanarii]

Fair enough. Not saying your way is wrong! I just don't like the idea of a new player taking the risk of rolling the dice, then regretting it. Feats and multiclassing are fun, but to each their own.
I don't like that idea much either. I strongly encourage players who are entirely new to D&D to take the standard array. But it's their choice, and choices should have consequences. Just like when the game is being played.

[JungleChicken]

When I was new to playing we would roll 4D6 drop the lowest. Depending on how brave we all felt we would either go down the line for stats or roll then place where we wanted.

We had a rule that the sum total of the stats average couldn't be below 13. This way if someone rolls really bad then can re roll and by and large you should end up with a minimum of +6 total attribute bonuses

[Waterdeep Merch]

My rules are 4d6 drop the lowest, with the following caveats-

If you play anything but a variant human or half-elf, two of your stats are rolled at 5d6 drop the two lowest.

If you play a half-elf, one of your stats is rolled at 5d6 drop the two lowest.

If you dislike your stats, you can renege and take standard array.

All dice are rolled where I can see them. You may not make a video of it and send it to me. You may not snatch up dice I haven't gotten to look at, or I will assume you rolled a 5 and standard array just got closed off to you. Any further attempts at cheating will result in your expulsion from the game, as you cannot be trusted to play fairly.

[Demonslayer666]

I disagree with everyone saying don't do this, don't do that.

Work out what works best at your table by letting them choose how to generate their character. If someone wants to point buy, let them. If someone wants the standard array, let em.

We do 4d6 drop the lowest, arrange as you choose, and you may pick your class and race after you roll. If your total stat modifier bonus is +1 or less, you can reroll. Plus you can point buy at 27 points or take the standard array.

If someone rolls incredibly well, give the other players a minor starting bonus (stat boost, feat, skill, item, etc).

[Laserlight]

My table are all bright people but several of them are....challenged by arithmetic. And apparently keeping track of "A 16 gives me a +3 and that costs X, and a 12 gives me a +1 and that costs Y, and my running total of X and Y is..." is more than they can handle. They have unanimously asked for rolls.

We use 2d6+6.

[PhoenixPhyre]

My table are all bright people but several of them are....challenged by arithmetic. And apparently keeping track of "A 16 gives me a +3 and that costs X, and a 12 gives me a +1 and that costs Y, and my running total of X and Y is..." is more than they can handle. They have unanimously asked for rolls.

We use 2d6+6.

Heck, I'm not mathematically challenged and point buy is too much tedium for me to want to do without a computer doing the work.

Edit: here's a random idea I just came up with for group ability generation: go around the group, each either rolling 4d6b3 or rerolling the previous person's score. Each score can only be rerolled once and you take the second roll. Go until you have 6 numbers, everyone uses this but can arrange as they see fit.

[Bobkin]

1. Primary Stats(3): 4d4+6, drop the lowest = +4, +3, +2, +1, + 6 = ~15 on average.

2. Secondary Stats(3): 3d4+3 = ~9 on average.

3. If (1)Secondary > (1)Primary, you may reroll both until Primary > Secondary or exchange them as desired.

4. No more than one 18. Reroll one stat if (2)18.

5. No less than 8 on multiple stats. If more than once you get # < 8 reroll one until you get >= 8.

[Whit]

1. It is up to the DM how he wants players is his her campaign.
2. Players are suppose to be cream of the crop of lvl 1
3. The DM can always modify encounters
4. Players should play their stats. If they want that low stats then play it.
5. We prefer the 4d6 per stat dropping the lowest.
6. However, you can get 1 with fantastic stats and 1 with garbage stats.
7. To remedy this you can decide what to do.
8. Yiu could also do a 1 for 2 after stats are rolled.
9. Yiu want your group to have fun. But also to accept rules.
10. We are not a fan of stats below 10 unless it's a strict campaign or a players decision for character concept.

1.4d6 drop lowest
2.4d6 7 times dropping lowest stat
3.4d6 drop lowest modify 1 for 2 points Once or up to you.
4. Pre rolled stats in phb. Add 1 to each before racial

[guachi]

For new players who want to roll I'd go with 4d6 drop lowest down the line. There are other die rolling methods I like better but this is simple.

The 3d6 drop lowest add 4 is also pretty good. It gives an average of 12.46.

The scores look like this (chances out of 216)
6--1
7--3
8--8
9--12
10--19
11--27
12--34
13--36
14--34
15--27
16--16

So the six most common are 10,11,12,13,14,15. It's a little higher than 4d6 drop lowest.


Personally, I'd just stick with the basic 4d6 drop lowest. Anyone who wants to change scores can do so if they want but put a time limit on creating a character so you can get playing. Let players change later if they want as well.

But rolling down the line is the easiest way to make a character and start. I suspect most new players are open to playing whatever the dice dictate.

[greenstone]

For new players, character creation is boring. It's the stuff you do before you get to play the actual game. It is the D&D equivalent of laying out the board, choosing tokens, and putting all the Monopoly money in neat piles.

For new players, give them pregens for the first couple of sessions. "What do you want to play? A smack-things fighter? A sneaky mage? Here. Lets get in to it!"

Once they've had a couple of hours of play (and I repeat, chargen is not play), then give them simple chargen. "What's mot important from STR, DEX, etc? Put 15 there. Next one, put 14. And so on."

Once someone has played the game for a while and got a handle on the rules then they can start digging in to the complexities of custom character generation, if they want. Some people don't want to do that, and that's fine.

[dropbear8mybaby]

Roll 1d65 using a random rolling app from somewhere, like http://donjon.bin.sh/

Consult this chart:

https://drive.google.com/open?id=1NWGjpBej0ikmZtjrhxfy2eheg502B6FzUcUR0tzDN Fg

Then roll 1d6 to determine which of the six scores is used for Strength. Then 1d5 for Dexterity, 1d4 for Con, and so on.

This way, you get a random character while still working within the balance of the point-buy system.

[Randomthom]

Standard array would be my first thought, it keeps things fair and simple. Also allows racial modifiers to create some distinction between the players.

If you really want to roll dice to determine the scores and don't want to risk some ultra-low scores there are a few good alternatives my groups have used down the years.

4d6, drop lowest x7 then drop the lowest stat.

4d6, drop lowest x6, do it twice, pick one array or the other.

6+2d6. (average score with this is 13.5 so higher than normal but not grossly-so).

[Logosloki]

For rolling, I go by the book. 6 x 4d6, drop lowest and arrange as is wanted. I do offer standard array after a roll attempt but only because I'm a softy. I will also sometimes run standard array with +2 to everything for short campaigns or where I'm doing a one-off with curveballs. I also also run standard array -2 on everything...for goblin/kobold adventures. I find that goblin/kobold fun times are best when you have a stack of adventurers that are dying like flies.

[Ionsniper]

Ive never DMed before but if I did I would most likely use the standard 4D6 x6. I would never do 3D6 down the line for a game that wasnt strictly designed for it and my players knew beforehand and agreed to it. Yes it does make for some interesting roleplaying and disastrous scenarios but it is not something I would use for every game.

Ive also thought about using 10 + 1D10, with 10 being Zero and 9 representing 8 plus a boon. Also thought of 6 + 2d6 reroll 1's. Stat array is also fairly basic with 18, 16, 14, 12, 12, 10. It lets people build what they want to build without having to antagonize over single stat build chars or MAD charachters. Id never want to limit my players choices unless I was upfront about it. Just my 2 copper pieces.