Greetings! I'm having some issues with finding a good median within my group for generating player character's ability scores, and part of it almost certainly relates to the fact that I took a 3 year hiatus from D&D (last playing under 4e) and I'm just making my return. The other part of the issue though is how I've typically thought of the way Adventurer's / Heroes are differentiated from NPC's as varied as commoners to hardened mercenaries and survivors of the harsh world we love to play in. Your character's are Exceptional: they are a full cut above the rabble and one part of that has nothing to do with the class they've chosen, or the current level they started the game at, it has to do with they're natural ability (everyone from commoner to king has a race so I don't consider that a part of what makes a character exceptional).
When I look at the standard array I generally have one emotional response: weak. Rolling ability scores via any number of methods (remember that back in the day, character's rolled 3d6 down the line from Strength to Charisma and thought that was fine) is too unpredictable. The point buy as suggested for 5e might as well just be the standard array because nothing that I've seen suggests there's a better distribution of your points than the 15, 14, 13, 12, 10, and 8 (not a fan of going "15, 15, 15, 8, 8, 8 (eff that)). Now, I'm currently in a space where for a few weeks I'll actually be in a sort of Assistant DM position as I re-familiarize myself with both the rules and also with the nuts and bolts of Roll20 (my groups preferred VTT). And once I start DM'ing in full - first running 'Curse of Strahd' before most likely going to a more standard home brew campaign - I'd like to have a greater confidence in the system used.
One alternative method of character generation I had contemplated was a revival of the homebrew point buy *I* used in 4e, where there was a Point Buy of 27, and the starting spread was 10, 10, 10, 10, 10, 8. Ability scores were bought in the following method: 9 cost – (1), 10 cost 0 (2), 11 cost 1 (3), 12 cost 2 (4), 13 cost 3 (5), 14 cost 5 (7), 15 cost 7 (9), 16 cost 9 (11), 17 cost 12 (14), and 18 cost 16 (18). An example array might be 17 (cost 12), 15 (cost 7), 14 (cost 5), 13 (cost 3), 10 (cost 0), and 8 (cost 0). Obviously you can do other builds but I feel that's comparative in the strength and weakness spread of the standard array to give you an idea of the difference. My players seemed fine with an array like that, but some were concerned when I also presented the example of 16, 16, 16, 10, 10, 8. Personally I think either the array of 17, 15, 14, 13, 10, 8 *or* an array of 17, 14, 14, 13, 12, 8 works well, but IDK.
A good couple of points were made as well, that when settling on any generation method, it has to be remembered that humans can get a +1 to all stats, while other races get anything from a +2, +2 to a +2, +1, +1 and so on and so forth, etc. And also true that it's not like 3rd or 3.5 or earlier editions where races were given a handicap for every bonus they got. Lastly, this isn't 3rd edition or even 4th, where there was no cap on ability scores. 95% of the time, you're not going to see an ability score go above 20, save in the case of epic level boons, nearly artifact level magic, etc. So a human fighter using the standard array of 17, 15, 14, 13, 10, 8 and taking the +1 to all ability scores, would end up at 20th level with an array of 20, 20, 20, 19, 10, 9 (in the instance where no feats were taken). By comparison, the standard 5e array would give 20, 20, 19, 13, 11, 9.
As a slight counterpoint, I'd mention it's not really a good idea to make your estimations of party strength around what they'll look like at 20th level, considering how few campaigns actually reach that point.
TL;DR? There's a split in the group on an ability score generation method where some people feel the standard 5e array is fine / good even, and others either gag or at least have their mouths twist at the idea. And the ones that feel the standard array is fine / good have a similar reaction to my proposed alternative, feeling it's too powerful. I personally like the idea of character's having a fair number of feats to help further customize them, while others (I think) feel the feats should be a bit more special and fewer.
If the community could offer some help and chime in with their own thoughts, I would certainly appreciate it as a way to help me and the group feel their way around this. (I'll be double checking with the group to ensure I presented the points of view correctly, right now this is just me getting something out that had been sticking in the back of my throat for a bit).
Honestly, all you are changing with the various options are the initial power level of the characters. The base system of either standard array or point buy is designed to give a maximum score at level 1 of 17 after racial adjustments. This is a +3 for whatever your primary stat might be. The reasoning appears to be that this is a good starting point in tier1. Having +4 or +5 in primary stats at level 1 does two things ... every to hit/damage/spell DC/save is higher making the character always more effective against tier appropriate opponents so the DM ends up having to scale up encounters a bit with higher CR creatures. The second effect is that the characters can then spend ASIs to max their stat earlier or to obtain performance enhancing feats. Both of these again advance the character on the power curve relative to typical level appropriate opponents.
Considering that the DM creates encounters that are designed/intended for the group of characters, all scaling up starting stats does is let the players look at slightly larger numbers or branch out their characters through the use of feats.
Keep in mind that the power scaling in 5e is much less than previous editions. Proficiency which affects skills/to hit/spell DC increases from +2 at level 1 to +6 at level 17. A +1 difference in a stat is almost the same as a several levels in terms of proficiency increases. This is why even low level creatures can still be a threat to high level ones. Typical to hit only varies from +5 at level 1 for a 16 stat to +11 at level 20 for a 20 stat. So giving characters heroic stats from the very beginning actually amounts to a much larger relative boost to character power than in previous editions.
Magic items are another aspect of the 5e philosophy of bounded accuracy. Characters can only attune 3 magic items. They have to choose what they want and the + bonus of weapons and armor is capped at +3 which really shouldn't see play before level 15+ or so for the most part. Base 5e is a lower magic item environment than previous editions. This also means that stats can be relatively more important than previous editions.
However, allowing more powerful characters won't break the game. It just means that the DM has to create more challenging encounters.
---
Anyway, as far as stat selection goes, rolled stats I find are too random. One person could end up starting with a 20 and other high stats and another player might only have a high score of 12. Some tables that use this approach have everyone roll up a set of stats and then allow the any player to choose any of the sets rolled. They might all choose the same set in some cases if one is clearly better however some folks might want sets with a better spread if they want to make a MAD character while another player might go for the set with the biggest number for a different type of character. The approach is fair to everyone and introduces a bit of randomness and some choice.
I have used basic point buy for almost all of my characters. I find it works pretty well but if you want higher stats you could bump the point allocation to 32 from 27 (for example) and set a cap of 17 as the maximum stat before racials. This allows characters to choose to start with a 19 after racial modifications (this is a +4) and lets the characters be a bit more powerful. Your alternate point buy method sounds fine too but not much different from the PHB point buy with a higher point base for allocation.
Finally, if you are just coming to 5e from an earlier edition, I would strongly recommend avoiding any house rules until you have played the game at least through the middle of tier 2 or even tier 3. There are a lot of initial reactions from folks starting 5e from previous editions when they see a moon druid in action at level 2-3, or a paladin smiting on a crit or a rogue sneak attack (especially again at level 3/4 before martial characters get extra attack at level 5). There are some cool 5e options that have a lot of burst damage that are actually decently balanced and fun in the longer term since monsters tend to have more hit points than previous editions. So, try to hold off judgement until you've played/DMed a wide variety of classes/levels before deciding some change is needed.
I also believe more “heroic” stats are in line with how I view PCs. Here’s what we do:
Start with this array: 16, 15, 14, 13, 12, 10
Optionally you can add 1 to a single stat, and subtract 1 from another stat with the following exceptions:
16+1 is not allowed, 10-1 is not allowed, and you cannot adjust both odd stats (therefore making 6 even stats before racial adjustments).
Yeah, pretty high stats but your primary is still capped at 18 after racial adjustments. Everything David42 said was right, but higher challenges than normal just means you level up faster. My group can’t meet more than once a month, so faster advancement is a good thing. YMMV
i did read somewhere that another group had everyone roll the standard 4d6 drop lowest for a six stat array but then every player got to pick which array they wanted from the whole group’s, meaning everyone could pick the best array. It kept the fun of rolling and eliminated the risk of imbalanced characters within the party.
I think that the printed 27 Point Buy is already strong enough to build optimized characters, and doesn't need to be any more flexible. If you feel that characters need a little power level bump, a very common house rule (and one that I allow in my own games) is to also let level 1 characters start with one Feat, which can potentially give them one more +1 if they choose a half-stat feat.
But as much as Point Buy is the golden standard for discussing builds on the forums, rolling for stats is so much more fun at the table, and I feel like it leads to surprising arrays that make you think about your character as more than just "which stat am I putting all my points in?" But rolling 3d6 very often results in rolls well below the baseline standard array, and so when rolling for stats I often see DMs allow...
Roll 3d6 six times, rerolling all 1's
Roll 4d6 six times, keep the highest 3
Roll 3d6 six times, but reroll the entire array if it adds up to less than 72
Roll 4d6 six times, rerolling 1's once, keep the highest 3
I've been trying to come up with a good way of having random characters without just having some of them better than others. One option is to randomly allocate points, but even at the same points, a well chosen stat allocation is significantly better than a poorly chosen stat allocation.
The most important thing for stats is balance between the party members. Balance vs enemies can be addressed as part of encounter balancing, but its far more difficult to adjust balance between PCs. With that in mind, my preferred methods are either point buy or have everyone roll a set using 4d6 dropping lowest then letting anyone use their choice of the rolled sets.
Lunali, but not every class is equally stat intensive. The casters and ranged shooters can kick ass with one high attack stat, and only need one more for survivability as a backup to the front liners doing their job. But your Barbarian is going to be a wet noodle without 3 high physical stats, and he'll need a middling fourth one if he wants to be able to growl at people or track a deer. Balance between the party may at times be best served by making sure that the MAD classes have their crack at a stat rolling method that can give them some meaty attributes.
Lunali, but not every class is equally stat intensive. The casters and ranged shooters can kick ass with one high attack stat, and only need one more for survivability as a backup to the front liners doing their job. But your Barbarian is going to be a wet noodle without 3 high physical stats, and he'll need a middling fourth one if he wants to be able to growl at people or track a deer. Balance between the party may at times be best served by making sure that the MAD classes have their crack at a stat rolling method that can give them some meaty attributes.
The dice do not know which characters need meaty attributes. If you want to do something that helps the MAD characters, use a different version of point build. However, I'd kind of disagree with barbarians being unusually MAD. Pretty much every class wants its primary ability, Con, and either Dex (no/light/medium armor) or Str (heavy armor). Fighters and Rogues only have two because their primary fighting ability is Str or Dex and fighters are proficient with heavy armor; barbarians, monks, and rangers don't get that benefit because they use a secondary stat for powering special abilities.
The easiest MAD fix would probably be to just say that ability gains on level-up must be to two different stats.
4d6, drop the low roll, arrange them how you like. If the rolls are no good, that guy decides to go be a farmer and you do another set. If you need some sort of objective measure of if they are good, at least one score 15 or higher, and the total of all modifiers must be greater than 1. Even then sometimes, it’s not a good character, and we end up with another farmer. yes, sometimes one player ends up with better stats than another. It’s not a big deal. The players are on the same team, there’s nothing wrong with having an all-star or two on the team. Everyone still gets to play.
Thing is, as a DM, if you want the characters to uniformly be better, just lower the DCs of the skillchecks and throw weaker or fewer enemies at them. The starting stat options aren't about comparing the characters to how good they are vs the world (since the "difficulty" of the "world" is on-the-fly adjustable by the DM) they're about how the experience of character creation works and balance of characters vs each other.
1) Standard array - everyone starts at basically the same arrangement. At level 1, their primary stat gets +3 bonus, their first and second ASI can get them to +4 and +5 so they can't "max out" early. Everyone has decent but not great secondary stats - they can start with either +3 +2 +2 or +3 +3 +1 depending on how many scores they care about, so two-attribute-dependent characters can start with two good scores. Characters are limited to only one "bad" score (the single -1) so everyone is decently well-rounded.
2) Point buy - for people who want to optimize. You can push characters to be more unbalanced, with multiple dump stats - it doesn't let you get above +3 for your starting modifier ever, but you can dump multiple stats and optimize multiple stats. It generates more specialized characters and generally fits better into long-term "builds".
3) Rolling. There's a million ways of rolling stats. The main reason there's a million of them is because people want the experience of getting randomly generated stats and getting "lucky" or "unlucky", BUT nobody wants to be left with an unusably bad character and honestly it's not that fun if someone starts with an overpoweredly good one either, so all the methods are "randomness" but with some bounds - mainly bounds on the lower end, because nobody wants to be the jerk DM that says "hey, you rolled 18 18 18, so we'll make you reroll" (even though a campaign where one person started with a stat of 20 at level 1 and someone else didn't get above a 16 could be pretty bad.)
So I don't really have an answer to your question, because to me it seems that if you just want the characters to "feel more powerful" you really don't need to homebrew a new stat rolling rule! Just start them at level 2 or 3 or whatever. Or give them an extra feat. (A feat is roughly similar in power to a +2 in a stat, since feats can be taken in lieu of +2 in a stat later.) Or do neither and just use the narrative to make the characters be more powerful. (After all, if you ACTUALLY power up the characters by giving them more stats and levels, as DM you'll just increase the difficulty of the challenges they face so that they feel challenged, right?)
When I look at the standard array I generally have one emotional response: weak.
IMO this discussion all comes down to that emotion.
The reason for it is that 5e, the way the balance works out, numbers are a little lower across the board than they were in 4e. This applies to both the characters AND the monsters and is balanced. But since you're still used to 4e balance, you look at 5e numbers for the PCs and they look low to you. IMO the best thing to do is recalibrate to thinking about 5e numbers instead of 4e numbers, instead of trying to make 5E PCs with 4E numbers (and then having to rebalance the rest of the game around that change too).
Roll 6*n stats, then going around the table clockwise each player can pick a single stat. Once everyone has one, go around counter clockwise starting with whoever picked the last stat for their second pick. Repeat untill everyone has 6 stats.
That way everyone gets more or less the same statsum, it allows for everyone to have a strength and a weakness.
Try standard array and add two to every stat, so 17, 16, 15, 14, 12, 10. If the player wants to take a negative in a stat (8), I let them have a racial feat from Xanathars. I've used this method a few times, mostly when I want players to create characters without me(DM) at the table to roll stats. Players seem happy with it and I don't have any problems with the power curve, I can change anything I want as DM to make the game more challenging.
I actually use The Holy Grail devised by Wraith1020 over on Reddit. It strikes my ideal balance between randomness and fairness. Essentially, you use a deck of cards to ensure that there is some randomness, but also nobody ends up with an epic score while somebody ends up with a peasant score, because the 'rolls' come from a pool. A lot of people who encourage rolling for stats like to have characters with definite strengths and weaknesses - this method ensures that if you get 18/16/13 you're gonna have some pretty interesting weaknesses.
The basic method is this:
You create the deck with 4x 2's, 3x 3's, 4x 4's, 3x 5's, and 4x 6's (2, 2, 2, 2, 3, 3, 3, 4, 4, 4, 4, 5, 5, 5, 6, 6, 6, 6).
Shuffle your deck
Draw three cards. That's your first stat
Discard those three cards.
Repeat steps 3 and 4 until you get six stats (which conveniently equals the amount of cards.
I encourage you to check out this Iterary made by theqwert and draw a couple sample arrays to get an idea. I 'rolled' a dozen characters or so and thought each one was encouraging to play but easy to DM for. You can also pretty easily tweak it if you want slightly stronger characters (ie, replace one of the 2s with a 5).
I used it with my group (who really wanted rolled stats when I wanted standard/point buy) as a compromise and they were pretty happy with it. Players that got a high of 14 looked at the player with a high of 17 and could see clear tradeoffs in their stats (ie, the 14 had a low stat of 11, the 17 had a 7.)
Plus, you can turn it into a dynamic part of character creation, like the tarokka reading in Curse of Strahd.
Rollback Post to RevisionRollBack
Another medical problem. Indefinite hiatus. Sorry, all.
Thanks everyone, for all your responses! Currently this (last) week we decided for a strange array where everyone starts with 8's but have 32 Points to buy their stats with and no cap of 15 (which is a BS bit of arbitrariness from wizards). It was fine by me, though I still argue that 5e has a poor allowance for it's point buy, and would like to point to the stats you see on Critical Role character's by example. I could have gone with their rolling method (think it's 4d6, minus the lowest die, done 6 times and reroll if the total of all 6 comes to less than 70), but I generally *hate* randomness impacting anything until the table comes into play.
Critical Role roll for abilities though I don't know (and I think they keep secret) the method they use. The stats over all the characters make me convinced they do something more powerful than the 4d6 drop 1 RAW option which is slightly better that standard array on average but not much (something like 16,14,13,12,11,8 taking average as you get one score in the middle of the top 16.7%ile of the distribution, another in the middle of the 2nd 16.7%ile and so on.). Critical role has a skilled DM who is able to balance encounters to noe standard strength characters and needs to be entertaining (and feats are generally more entertaining that ASIs and high ability characters allows more room for feats.
One very simple way of rolling while ensuring everyone in the party is reasonably balanced in rolls 4d6D1 THREE times. Your other 3 stats are 24 minus what you rolled so if you have 18,18,18 you have to deal with 6,6,6 (if is very unlikely to be as extreme as this).
Card draw to replace the die roll produces some chars that are distinctly better than others; 12/12/12/12/12/12 and 18/12/12/12/10/8 are not equally good.
Critical Role roll for abilities though I don't know (and I think they keep secret) the method they use. The stats over all the characters make me convinced they do something more powerful than the 4d6 drop 1 RAW option which is slightly better that standard array on average but not much (something like 16,14,13,12,11,8 taking average as you get one score in the middle of the top 16.7%ile of the distribution, another in the middle of the 2nd 16.7%ile and so on.). Critical role has a skilled DM who is able to balance encounters to noe standard strength characters and needs to be entertaining (and feats are generally more entertaining that ASIs and high ability characters allows more room for feats.
One very simple way of rolling while ensuring everyone in the party is reasonably balanced in rolls 4d6D1 THREE times. Your other 3 stats are 24 minus what you rolled so if you have 18,18,18 you have to deal with 6,6,6 (if is very unlikely to be as extreme as this).
Matt Mercer also creates very powerful magical items. On a whole the Critical Role campaigns are harder than the WotC ones but the characters have access to more powerful items. It averages out because their characters do die.
Card draw to replace the die roll produces some chars that are distinctly better than others; 12/12/12/12/12/12 and 18/12/12/12/10/8 are not equally good.
Sure, but less often than rolling for the same characters. I feel like the gap is also closer in cases where it does. Your 'worst case' here still has a Point Buy equivalent of 24, which is damn close to the 27 point benchmark, and the 'best case' scores a 33 on Chicken Dinner.
Rolling for stats can (theoretically) have a character with 3/3/3/3/3/3 in a party with 18/18/18/18/18/18. Such a huge gap is impossible when using a pool (ie, card draw). In any game you can pretty quickly point out the character with the 'best' stats and the 'worst' stats and usually by a bigger margin than the card draw.
Rollback Post to RevisionRollBack
Another medical problem. Indefinite hiatus. Sorry, all.
To post a comment, please login or register a new account.
Greetings! I'm having some issues with finding a good median within my group for generating player character's ability scores, and part of it almost certainly relates to the fact that I took a 3 year hiatus from D&D (last playing under 4e) and I'm just making my return. The other part of the issue though is how I've typically thought of the way Adventurer's / Heroes are differentiated from NPC's as varied as commoners to hardened mercenaries and survivors of the harsh world we love to play in. Your character's are Exceptional: they are a full cut above the rabble and one part of that has nothing to do with the class they've chosen, or the current level they started the game at, it has to do with they're natural ability (everyone from commoner to king has a race so I don't consider that a part of what makes a character exceptional).
When I look at the standard array I generally have one emotional response: weak. Rolling ability scores via any number of methods (remember that back in the day, character's rolled 3d6 down the line from Strength to Charisma and thought that was fine) is too unpredictable. The point buy as suggested for 5e might as well just be the standard array because nothing that I've seen suggests there's a better distribution of your points than the 15, 14, 13, 12, 10, and 8 (not a fan of going "15, 15, 15, 8, 8, 8 (eff that)). Now, I'm currently in a space where for a few weeks I'll actually be in a sort of Assistant DM position as I re-familiarize myself with both the rules and also with the nuts and bolts of Roll20 (my groups preferred VTT). And once I start DM'ing in full - first running 'Curse of Strahd' before most likely going to a more standard home brew campaign - I'd like to have a greater confidence in the system used.
One alternative method of character generation I had contemplated was a revival of the homebrew point buy *I* used in 4e, where there was a Point Buy of 27, and the starting spread was 10, 10, 10, 10, 10, 8. Ability scores were bought in the following method: 9 cost – (1), 10 cost 0 (2), 11 cost 1 (3), 12 cost 2 (4), 13 cost 3 (5), 14 cost 5 (7), 15 cost 7 (9), 16 cost 9 (11), 17 cost 12 (14), and 18 cost 16 (18). An example array might be 17 (cost 12), 15 (cost 7), 14 (cost 5), 13 (cost 3), 10 (cost 0), and 8 (cost 0). Obviously you can do other builds but I feel that's comparative in the strength and weakness spread of the standard array to give you an idea of the difference. My players seemed fine with an array like that, but some were concerned when I also presented the example of 16, 16, 16, 10, 10, 8. Personally I think either the array of 17, 15, 14, 13, 10, 8 *or* an array of 17, 14, 14, 13, 12, 8 works well, but IDK.
A good couple of points were made as well, that when settling on any generation method, it has to be remembered that humans can get a +1 to all stats, while other races get anything from a +2, +2 to a +2, +1, +1 and so on and so forth, etc. And also true that it's not like 3rd or 3.5 or earlier editions where races were given a handicap for every bonus they got. Lastly, this isn't 3rd edition or even 4th, where there was no cap on ability scores. 95% of the time, you're not going to see an ability score go above 20, save in the case of epic level boons, nearly artifact level magic, etc. So a human fighter using the standard array of 17, 15, 14, 13, 10, 8 and taking the +1 to all ability scores, would end up at 20th level with an array of 20, 20, 20, 19, 10, 9 (in the instance where no feats were taken). By comparison, the standard 5e array would give 20, 20, 19, 13, 11, 9.
As a slight counterpoint, I'd mention it's not really a good idea to make your estimations of party strength around what they'll look like at 20th level, considering how few campaigns actually reach that point.
TL;DR? There's a split in the group on an ability score generation method where some people feel the standard 5e array is fine / good even, and others either gag or at least have their mouths twist at the idea. And the ones that feel the standard array is fine / good have a similar reaction to my proposed alternative, feeling it's too powerful. I personally like the idea of character's having a fair number of feats to help further customize them, while others (I think) feel the feats should be a bit more special and fewer.
If the community could offer some help and chime in with their own thoughts, I would certainly appreciate it as a way to help me and the group feel their way around this. (I'll be double checking with the group to ensure I presented the points of view correctly, right now this is just me getting something out that had been sticking in the back of my throat for a bit).
Honestly, all you are changing with the various options are the initial power level of the characters. The base system of either standard array or point buy is designed to give a maximum score at level 1 of 17 after racial adjustments. This is a +3 for whatever your primary stat might be. The reasoning appears to be that this is a good starting point in tier1. Having +4 or +5 in primary stats at level 1 does two things ... every to hit/damage/spell DC/save is higher making the character always more effective against tier appropriate opponents so the DM ends up having to scale up encounters a bit with higher CR creatures. The second effect is that the characters can then spend ASIs to max their stat earlier or to obtain performance enhancing feats. Both of these again advance the character on the power curve relative to typical level appropriate opponents.
Considering that the DM creates encounters that are designed/intended for the group of characters, all scaling up starting stats does is let the players look at slightly larger numbers or branch out their characters through the use of feats.
Keep in mind that the power scaling in 5e is much less than previous editions. Proficiency which affects skills/to hit/spell DC increases from +2 at level 1 to +6 at level 17. A +1 difference in a stat is almost the same as a several levels in terms of proficiency increases. This is why even low level creatures can still be a threat to high level ones. Typical to hit only varies from +5 at level 1 for a 16 stat to +11 at level 20 for a 20 stat. So giving characters heroic stats from the very beginning actually amounts to a much larger relative boost to character power than in previous editions.
Magic items are another aspect of the 5e philosophy of bounded accuracy. Characters can only attune 3 magic items. They have to choose what they want and the + bonus of weapons and armor is capped at +3 which really shouldn't see play before level 15+ or so for the most part. Base 5e is a lower magic item environment than previous editions. This also means that stats can be relatively more important than previous editions.
However, allowing more powerful characters won't break the game. It just means that the DM has to create more challenging encounters.
---
Anyway, as far as stat selection goes, rolled stats I find are too random. One person could end up starting with a 20 and other high stats and another player might only have a high score of 12. Some tables that use this approach have everyone roll up a set of stats and then allow the any player to choose any of the sets rolled. They might all choose the same set in some cases if one is clearly better however some folks might want sets with a better spread if they want to make a MAD character while another player might go for the set with the biggest number for a different type of character. The approach is fair to everyone and introduces a bit of randomness and some choice.
I have used basic point buy for almost all of my characters. I find it works pretty well but if you want higher stats you could bump the point allocation to 32 from 27 (for example) and set a cap of 17 as the maximum stat before racials. This allows characters to choose to start with a 19 after racial modifications (this is a +4) and lets the characters be a bit more powerful. Your alternate point buy method sounds fine too but not much different from the PHB point buy with a higher point base for allocation.
Finally, if you are just coming to 5e from an earlier edition, I would strongly recommend avoiding any house rules until you have played the game at least through the middle of tier 2 or even tier 3. There are a lot of initial reactions from folks starting 5e from previous editions when they see a moon druid in action at level 2-3, or a paladin smiting on a crit or a rogue sneak attack (especially again at level 3/4 before martial characters get extra attack at level 5). There are some cool 5e options that have a lot of burst damage that are actually decently balanced and fun in the longer term since monsters tend to have more hit points than previous editions. So, try to hold off judgement until you've played/DMed a wide variety of classes/levels before deciding some change is needed.
I also believe more “heroic” stats are in line with how I view PCs. Here’s what we do:
Start with this array: 16, 15, 14, 13, 12, 10
Optionally you can add 1 to a single stat, and subtract 1 from another stat with the following exceptions:
16+1 is not allowed, 10-1 is not allowed, and you cannot adjust both odd stats (therefore making 6 even stats before racial adjustments).
Yeah, pretty high stats but your primary is still capped at 18 after racial adjustments. Everything David42 said was right, but higher challenges than normal just means you level up faster. My group can’t meet more than once a month, so faster advancement is a good thing. YMMV
i did read somewhere that another group had everyone roll the standard 4d6 drop lowest for a six stat array but then every player got to pick which array they wanted from the whole group’s, meaning everyone could pick the best array. It kept the fun of rolling and eliminated the risk of imbalanced characters within the party.
I think that the printed 27 Point Buy is already strong enough to build optimized characters, and doesn't need to be any more flexible. If you feel that characters need a little power level bump, a very common house rule (and one that I allow in my own games) is to also let level 1 characters start with one Feat, which can potentially give them one more +1 if they choose a half-stat feat.
But as much as Point Buy is the golden standard for discussing builds on the forums, rolling for stats is so much more fun at the table, and I feel like it leads to surprising arrays that make you think about your character as more than just "which stat am I putting all my points in?" But rolling 3d6 very often results in rolls well below the baseline standard array, and so when rolling for stats I often see DMs allow...
dndbeyond.com forum tags
I'm going to make this way harder than it needs to be.
I've been trying to come up with a good way of having random characters without just having some of them better than others. One option is to randomly allocate points, but even at the same points, a well chosen stat allocation is significantly better than a poorly chosen stat allocation.
The most important thing for stats is balance between the party members. Balance vs enemies can be addressed as part of encounter balancing, but its far more difficult to adjust balance between PCs. With that in mind, my preferred methods are either point buy or have everyone roll a set using 4d6 dropping lowest then letting anyone use their choice of the rolled sets.
Lunali, but not every class is equally stat intensive. The casters and ranged shooters can kick ass with one high attack stat, and only need one more for survivability as a backup to the front liners doing their job. But your Barbarian is going to be a wet noodle without 3 high physical stats, and he'll need a middling fourth one if he wants to be able to growl at people or track a deer. Balance between the party may at times be best served by making sure that the MAD classes have their crack at a stat rolling method that can give them some meaty attributes.
dndbeyond.com forum tags
I'm going to make this way harder than it needs to be.
The dice do not know which characters need meaty attributes. If you want to do something that helps the MAD characters, use a different version of point build. However, I'd kind of disagree with barbarians being unusually MAD. Pretty much every class wants its primary ability, Con, and either Dex (no/light/medium armor) or Str (heavy armor). Fighters and Rogues only have two because their primary fighting ability is Str or Dex and fighters are proficient with heavy armor; barbarians, monks, and rangers don't get that benefit because they use a secondary stat for powering special abilities.
The easiest MAD fix would probably be to just say that ability gains on level-up must be to two different stats.
4d6, drop the low roll, arrange them how you like. If the rolls are no good, that guy decides to go be a farmer and you do another set.
If you need some sort of objective measure of if they are good, at least one score 15 or higher, and the total of all modifiers must be greater than 1. Even then sometimes, it’s not a good character, and we end up with another farmer.
yes, sometimes one player ends up with better stats than another. It’s not a big deal. The players are on the same team, there’s nothing wrong with having an all-star or two on the team. Everyone still gets to play.
Thing is, as a DM, if you want the characters to uniformly be better, just lower the DCs of the skillchecks and throw weaker or fewer enemies at them. The starting stat options aren't about comparing the characters to how good they are vs the world (since the "difficulty" of the "world" is on-the-fly adjustable by the DM) they're about how the experience of character creation works and balance of characters vs each other.
1) Standard array - everyone starts at basically the same arrangement. At level 1, their primary stat gets +3 bonus, their first and second ASI can get them to +4 and +5 so they can't "max out" early. Everyone has decent but not great secondary stats - they can start with either +3 +2 +2 or +3 +3 +1 depending on how many scores they care about, so two-attribute-dependent characters can start with two good scores. Characters are limited to only one "bad" score (the single -1) so everyone is decently well-rounded.
2) Point buy - for people who want to optimize. You can push characters to be more unbalanced, with multiple dump stats - it doesn't let you get above +3 for your starting modifier ever, but you can dump multiple stats and optimize multiple stats. It generates more specialized characters and generally fits better into long-term "builds".
3) Rolling. There's a million ways of rolling stats. The main reason there's a million of them is because people want the experience of getting randomly generated stats and getting "lucky" or "unlucky", BUT nobody wants to be left with an unusably bad character and honestly it's not that fun if someone starts with an overpoweredly good one either, so all the methods are "randomness" but with some bounds - mainly bounds on the lower end, because nobody wants to be the jerk DM that says "hey, you rolled 18 18 18, so we'll make you reroll" (even though a campaign where one person started with a stat of 20 at level 1 and someone else didn't get above a 16 could be pretty bad.)
So I don't really have an answer to your question, because to me it seems that if you just want the characters to "feel more powerful" you really don't need to homebrew a new stat rolling rule! Just start them at level 2 or 3 or whatever. Or give them an extra feat. (A feat is roughly similar in power to a +2 in a stat, since feats can be taken in lieu of +2 in a stat later.) Or do neither and just use the narrative to make the characters be more powerful. (After all, if you ACTUALLY power up the characters by giving them more stats and levels, as DM you'll just increase the difficulty of the challenges they face so that they feel challenged, right?)
IMO this discussion all comes down to that emotion.
The reason for it is that 5e, the way the balance works out, numbers are a little lower across the board than they were in 4e. This applies to both the characters AND the monsters and is balanced. But since you're still used to 4e balance, you look at 5e numbers for the PCs and they look low to you. IMO the best thing to do is recalibrate to thinking about 5e numbers instead of 4e numbers, instead of trying to make 5E PCs with 4E numbers (and then having to rebalance the rest of the game around that change too).
Pondering a variant random allocation: roll 15d6, group by number, and then add one die of your choice. Determine stats as follows:
This is basically a point allocation using 16 points instead of 27 (it's possible to generate the standard array) with a median stat of 12.
Pointbuy or standard array + a free ASI/Feat
Or
Roll 6*n stats, then going around the table clockwise each player can pick a single stat. Once everyone has one, go around counter clockwise starting with whoever picked the last stat for their second pick. Repeat untill everyone has 6 stats.
That way everyone gets more or less the same statsum, it allows for everyone to have a strength and a weakness.
Attack: [roll]3d20dl2+6[/roll] Damage: [roll](if([roll:-1:critical], =0, 1d8, 2{1d8}.expand()))+4[/roll] Sneakattack: [roll](if([roll:-2:critical], =0, 2d6, 2{2d6}.expand()))[/roll]Try standard array and add two to every stat, so 17, 16, 15, 14, 12, 10. If the player wants to take a negative in a stat (8), I let them have a racial feat from Xanathars. I've used this method a few times, mostly when I want players to create characters without me(DM) at the table to roll stats. Players seem happy with it and I don't have any problems with the power curve, I can change anything I want as DM to make the game more challenging.
I actually use The Holy Grail devised by Wraith1020 over on Reddit. It strikes my ideal balance between randomness and fairness. Essentially, you use a deck of cards to ensure that there is some randomness, but also nobody ends up with an epic score while somebody ends up with a peasant score, because the 'rolls' come from a pool. A lot of people who encourage rolling for stats like to have characters with definite strengths and weaknesses - this method ensures that if you get 18/16/13 you're gonna have some pretty interesting weaknesses.
The basic method is this:
You create the deck with 4x 2's, 3x 3's, 4x 4's, 3x 5's, and 4x 6's (2, 2, 2, 2, 3, 3, 3, 4, 4, 4, 4, 5, 5, 5, 6, 6, 6, 6).
Shuffle your deck
Draw three cards. That's your first stat
Discard those three cards.
Repeat steps 3 and 4 until you get six stats (which conveniently equals the amount of cards.
I encourage you to check out this Iterary made by theqwert and draw a couple sample arrays to get an idea. I 'rolled' a dozen characters or so and thought each one was encouraging to play but easy to DM for. You can also pretty easily tweak it if you want slightly stronger characters (ie, replace one of the 2s with a 5).
I used it with my group (who really wanted rolled stats when I wanted standard/point buy) as a compromise and they were pretty happy with it. Players that got a high of 14 looked at the player with a high of 17 and could see clear tradeoffs in their stats (ie, the 14 had a low stat of 11, the 17 had a 7.)
Plus, you can turn it into a dynamic part of character creation, like the tarokka reading in Curse of Strahd.
Another medical problem. Indefinite hiatus. Sorry, all.
Thanks everyone, for all your responses! Currently this (last) week we decided for a strange array where everyone starts with 8's but have 32 Points to buy their stats with and no cap of 15 (which is a BS bit of arbitrariness from wizards). It was fine by me, though I still argue that 5e has a poor allowance for it's point buy, and would like to point to the stats you see on Critical Role character's by example. I could have gone with their rolling method (think it's 4d6, minus the lowest die, done 6 times and reroll if the total of all 6 comes to less than 70), but I generally *hate* randomness impacting anything until the table comes into play.
Critical Role roll for abilities though I don't know (and I think they keep secret) the method they use. The stats over all the characters make me convinced they do something more powerful than the 4d6 drop 1 RAW option which is slightly better that standard array on average but not much (something like 16,14,13,12,11,8 taking average as you get one score in the middle of the top 16.7%ile of the distribution, another in the middle of the 2nd 16.7%ile and so on.). Critical role has a skilled DM who is able to balance encounters to noe standard strength characters and needs to be entertaining (and feats are generally more entertaining that ASIs and high ability characters allows more room for feats.
One very simple way of rolling while ensuring everyone in the party is reasonably balanced in rolls 4d6D1 THREE times. Your other 3 stats are 24 minus what you rolled so if you have 18,18,18 you have to deal with 6,6,6 (if is very unlikely to be as extreme as this).
Card draw to replace the die roll produces some chars that are distinctly better than others; 12/12/12/12/12/12 and 18/12/12/12/10/8 are not equally good.
Matt Mercer also creates very powerful magical items. On a whole the Critical Role campaigns are harder than the WotC ones but the characters have access to more powerful items. It averages out because their characters do die.
Sure, but less often than rolling for the same characters. I feel like the gap is also closer in cases where it does. Your 'worst case' here still has a Point Buy equivalent of 24, which is damn close to the 27 point benchmark, and the 'best case' scores a 33 on Chicken Dinner.
Rolling for stats can (theoretically) have a character with 3/3/3/3/3/3 in a party with 18/18/18/18/18/18. Such a huge gap is impossible when using a pool (ie, card draw). In any game you can pretty quickly point out the character with the 'best' stats and the 'worst' stats and usually by a bigger margin than the card draw.
Another medical problem. Indefinite hiatus. Sorry, all.