I'm starting my first stint as a DM, in DnD 4E, and one of my players insisted on rolling dice for his stats. I was ok, provided, as I said, "that the character was balanced with the rest of the group".

Problem: his rolls were something like '13, 13, 14, 14, 15, 17' (final stats, after discarding the worst value of 7 4d6 rolls). The other two players who rolled today, both newbies to 4E and, in my wife's case, newbie to role-playing, were "normal", with a few 11s, 12s, 14s, and 16s (I misremembered. I must have been really worried about this).

What I wanted to know is: should I make him reroll, or the other characters reroll or select some "better" values from standard arrays or using the point-buy system? Or am I worrying over nothing? He's playing a striker, but apart from possibly doing tons of damage, I'm worried that every skill challenge will be too easy to him. I was thinking of applying some hard DCs to his rolls, because hey, You Asked For It You Got It.

Bonus points on detailed advice on how to handle it. I am, as I said, new at this, and he's an experienced player who talked a lot. He might be a problem player (at least for me), or I might be wrong. I'm a bit afraid of becoming a doormat or overreacting.

As it is possible that the particular player might read this: please don't take offence. It's my first time at this, we just met, and I'm just trying to create the best game I can. As I said, it's possible that I am completely wrong.

  • 11
    \$\begingroup\$ I'm curious as to why you permitted this player — or any of the players — to roll for their stats in this game. The rules for 4E no longer consider this a valid method of character generation. \$\endgroup\$
    – Jadasc
    Commented Feb 7, 2011 at 12:44
  • \$\begingroup\$ @Jadasc Well, in the Essentials RC, rolling dice for stats are allowed, only with a suggestion that if the rolls are too low or high the DM might at his discretion make the player reroll or choose another way to do the stats. \$\endgroup\$ Commented Feb 7, 2011 at 12:55
  • \$\begingroup\$ @Adriano In that case, I would probably establish a baseline of "no stat at 16 or above" as "too low." \$\endgroup\$
    – Jadasc
    Commented Feb 7, 2011 at 13:15
  • 5
    \$\begingroup\$ @Jadasc, Are you sure? IIRC, the PHB recommends point-buy, but shows standard arrays and dice rolling as other options. \$\endgroup\$
    – Brian S
    Commented Nov 20, 2013 at 15:13

8 Answers 8


I don't think this character will be overpowered. In fact, I suspect that the other characters will be underpowered — the math of D&D 4E is engineered for characters who have at least a 16 in the score most related to their class: Strength for fighters, for example.

As for the question of whether this character will be a "Mary Sue," the question is difficult to answer in terms of roleplaying games — characters in RPGs often are hypercompetent and well-loved by the populace, even those hard to convince. It's the portrayal, not the mechanics, that determines Mary or Marty.

  • \$\begingroup\$ Thanks for your answer. It's true that the 'mary sue' bit of the question is unanswerable based on just this fact, and in fact belongs in a different question. \$\endgroup\$ Commented Feb 7, 2011 at 12:56
  • \$\begingroup\$ what is a mary sue? \$\endgroup\$
    – DForck42
    Commented Dec 19, 2011 at 20:59
  • 2
    \$\begingroup\$ An impossibly perfect, beloved-by-all heroine. It comes originally from Star Trek fandom, but has grown to a commonly used trope in the SF/fan community. tvtropes.org/pmwiki/pmwiki.php/Main/MarySue \$\endgroup\$
    – Jadasc
    Commented Dec 19, 2011 at 21:24
  • 1
    \$\begingroup\$ Regarding the fact that the other characters will be a bit underpowered, if things get too bad, I would seriously suggest trying to find a way to shore up their abilities so they can handle the challenges and keep up with the the "big man". How you go about this will vary, but carefully selected/crafted magic items is one route and presenting them with skill checks their characters can pass (but aren't too out of the ordinary) is a another (they have to be good at something). I would also consider house ruling in-game bonuses (temporary or permanent) - but sparingly. \$\endgroup\$ Commented Jul 31, 2014 at 17:30

Using stat arrays, my two preferred arrays are:

18 14 11 10 10 8

16 16 12 12 10 8

The Gamma World game, which is based off of D&D 4e, and uses similar math, starts characters with an 18 in their primary stat, and a 16 in their secondary stat, and asks all other stats to be rolled.

Compared to either of the above, the off stats for the one character are well above average, but the primary stats and secondary stats look reasonably balanced.

I would echo Jadasc's concerns that if the other characters are topping out at 14, that they will be underpowered.

[edit] I would recommend letting that player have his fun - in this case at least, he's good but not great, but also allowing the other players in the group use a standard array, as that will ensure that they have characters with ability scores sufficient to allow them to excel at what their characters need to do.

In addition, by pointing them to a method likely to give them a low stat, you can encourage the other players to role-play a weaknesses. [/edit]

  • \$\begingroup\$ Is the -1 modifier to a skill (for the 8) a bother to inexperienced players? \$\endgroup\$ Commented Feb 7, 2011 at 15:03
  • 2
    \$\begingroup\$ Probably not... I've not played with inexperienced players in years now, but I would imagine that the knowledge that "there are some things that my character isn't good at" is an easy and natural enough notion that it shouldn't be a problem. \$\endgroup\$ Commented Feb 7, 2011 at 15:52

I am not particularly familiar with 4e, so I will not attempt to answer how much stats impact play. This answer will focus on whether the players were better off rolling.

Point-buy v. 4d6, drop the lowest

  1. 4d6, Drop the lowest

    I just wrote a little script (which may not be perfect) that rolls 4d6 and drops the lowest.

    The average of 10,000,000 stat rolls was 12.243. Let's assume that as our average (if anyone could do the math to give me the actual expected value of 4d6 drop the lowest, that would be great).

    Using this average for each stat gives this array: 12.243, 12.243, 12.243, 12.243, 12.243, 12.243. This sums to 73.458 stat points.

  2. Point buy ($22)

    I will refer to the currency in point buy as $.

    Explanation: You start with the array 10, 10, 10, 10, 10, 8. You have $22 to improve those skills. From 8 to 13, each stat point costs $1. From 14 to 16, each additional stat point costs $2. The 17th stat point costs $3 and the 18th stat point costs $4.

    Application: Buying stats most efficiently (spending as few $ as possible per stat) results in the array 14, 13, 13, 13, 13, 13, which totals to 79 stat points. Buying points least efficiently ( highest $/stat ratio ) results in the array 18, 14, 11, 10, 10, 8, which totals to 71 stat points.

    Point-buy thus results in a range of 71-79 stat points.


  1. The experienced player

    The experienced player's stats are 13, 13, 14, 14, 15, 17. The sum of these stats is 86 stat points, almost 13 points above the average roll.

    By choosing to roll, the experienced player chose an option that would provide an average ~73 stat points. Rolling an 86 was actually high for that roll scheme.

  2. The other rollers

    The original post says the other players had arrays "with a few 11s, 12s, 14s, and 16s". At most, this means a stat array of 16, 16, 14, 14, 12, 11, which totals 83 points. At least, this means a stat array of 16, 14, 12, 12, 11, 11, which totals 76 points. We will assume, then, that the other rolling players were in the range 76-83.

    If the other rolling players rolled the estimated range (76-83), they are down 3-10 points from the other player, but still above the average by 3-10 points.

  3. The point-buy players

    The players that chose point-buy ended up in the range 71-79. This ranges from 2 points below the rolling average to 6 points above it. However, it offers predictability (no rolls way below average) and choice (how to assign stats). With a $22 start, this seems fairly balanced against the roll.


The experienced player took a gamble that paid off, but the choice was not systematically unfair to the other players.

Further Questions

Since the systems' center values are fairly close, they seem balanced mathematically. But there are other questions that this answer doesn't cover:

  1. Is the opportunity to cripple your character fun? One set of low rolls at the beginning of the game can negatively affect your character for the rest of play.

  2. Is an unusually high roll unfair for the rest of the players? Should pure rolling benefit one character for the duration of play?

  3. Is a low roll actually penalizing? If the GM's style allows for dying and re-rolling, the player could sacrifice themselves during the first few sessions to allow a re-roll. If the GM allows the player to start at the same level, the problem is even more pronounced.

  4. What is your goal for rolling? Is it just gambling, or does it provide other benefits? Is that the best way to obtain those benefits?

  • 1
    \$\begingroup\$ 12.24 is the accurate average of 4d6 drop 1 according to anyDice, however the most commonly rolled result is 13. See for yourself at by calculating on the phrase "output [highest 3 of 4d6]" \$\endgroup\$
    – dpatchery
    Commented Dec 20, 2011 at 20:53
  • \$\begingroup\$ Note that when comparing the scores, each estimated 6 points means you're ~5% better at everything, and do ~13% more damage. A difference of 12 points means ~10% better at almost everything, and do ~27% more damage. Don't make the mistake of thinking that 76/86 -> 12% better. (If that appears close to the actual values, that's pure coincidence.) \$\endgroup\$ Commented Jul 31, 2014 at 20:08

Probably Not

In 4e, stats are one of the least important areas of a character. So long as you have 16-20 in your primary and 14-20 in your secondary (for most classes) you'll be fine. the 17/15 stat array seems underpowered when it comes to a standard array (at least it's a bad combination) but nothing that I'd take exception to.

A better way of measuring power is to look at the DPR for each character. Damage Per Round is a key indicator of cheese because it is the value directly applied against the enemy hitpoint total. Therefore it factors in multi-attacks from powers, changes from feats, and associated class features.

After accuracy is factored in, I would only consider a character "overpowered" if they could do more than half a monster of their level's HP in a single round. (Round rather than turn due to some off-turn attacks being critical parts of some builds).

If you post the levels, classes, important feats, and items of each player, we will be able to assess whether the power-balance is significantly out of whack.

To calculate DPR, look here. The best way to handle it, if this is a worry to you, is to force all of your players to use point-buy. Player skill, ultimately, determines the success or failure or the character, rather than stats.

In general, I've had great success with group character creation. Look here for an example of how it plays out and here for the theory.

  • 1
    \$\begingroup\$ Since Wave will cease to exist sometime in the near future (I forget when), you may want to consider finding a way to preserve your example. If you have a blog, maybe you could post it on there and update the link? \$\endgroup\$
    – user1637
    Commented Dec 19, 2011 at 20:45

I don't allow rolling for stats anymore. I had a player, whom I watched roll, get 4 18s. It just isn't fair to the rest of the group. I much prefer either giving everyone the same stats to assign at will, or a point-buy system to allow a little more customization.


In a group of experienced players, who have most likely gotten over the "twink" phase of role-playing, rolling for stats is fair-game and likely to create more interesting situations, than not. With my more advanced groups, we would use very draconian rules for getting stats for just these reasons. Wide stats ranges make challenging situations and combos more easy to find.

When starting up, with mostly inexperienced players, especially in a D&D4 game, I'd say go with standard arrays. Force the advanced player to try and show his skills at playing through better role-playing, and by teaching his fellow role-players the rules, rather than taking advantage of them.

You might also want to read up a bit on role playing game theory, there are some abstract concepts which can help you deal with some of these kinds of problems (for example, http://www.darkshire.net/jhkim/rpg/theory/threefold/faq_v1.html)


In this instance the character is not overpowered in the least. He just took a reasonable gamble and it paid off. The other players on the other hand are kinda underpowered but that can be fixed with either a re-roll of one random key stat. You might be thinking using point buy might fix this problem but really it won't.

One of the main reasons I don't use point buy system is that every character turns out generic or min/maxed. So you want to be a (insert class here) for the fifth time in a row? Okay just use the exact same stats from the other four (insert class here) you played and just change the name. Oh you want a fighter with max str and con? Then everything else just becomes a dump stat. Oh and don't get me started on racial bonuses, I have never seen a friend using point buy ever choose human as a race. And besides if you never had to play a class you're not used to then things would get really dull really fast. Oh and point buy basically means you have generic pre made stat builds to choose from in every class.

Rolling stats has its rewards and it's downfalls.

Good: like in this instance the person got lucky and rolled some good stats. Sometimes you can even get a character that has higher stat totals than something point buy could get you.

Bad: you can get a very low stat roll and be stuck with it. Rolled a 5 for strength? Okay get over it. But this isn't really a bad thing, if you are concerned for the role playing potential at all (but if you are just trying to get a generic class that's easily replaceable then use point buy) you will find that the biggest flaws make for the most fun.

Character development can be based around the flaws of a character. You can even go as far as explaining a low stat roll in a epic backstory. One human fighter I had got a charisma roll of 4 and was the ugliest S.O.B you've ever seen. I made this awesome backstory about how his face was mangled in a tragic arson incident by a mercenary corp. the D.M really liked it and even made it work in the campaign!

So in conclusion the kid got lucky, don't penalize him for it. If it becomes a problem just buff up the other players or go to the old "generic and replaceable" point buy characters to even things out(not recommended).


I will try to get into the game mechanics a little bit more than the other answers, showing you what those values really mean in D&D 4e.

We need something to compare these values to so I'll use a small subset of the possible point-buy arrays, the ones that grant us the best tactical advantage:

  • The unbalanced array, 18 14 11 10 10 8, very useful for strikers and characters who prioritize hitting above high defenses.
  • The balanced array, 16 16 12 12 10 8, or the very similar 16 16 13 11 10 8, useful for all other builds

Some other combinations are possible, but they get only used in special cases where characters need at least a 13 in four different stats, because of feat requirements. While the rolled character can't ever compete with the unbalanced array (which is the favourite of many strikers), I will include it for comparison.

All six abilities contribute to at least one skill. But having a +1 or a +2 in a wider array of skills is not game-unbalancing.
This character is gonna get a +4 after racial modifiers at best, which is in line with the balanced array. Supposing the balanced array gets used at its best, with 16s matching the racial +2s, this means he will get a +3 instead of a +4 in some skills and a character-wide +1 boost on all others. Very not game-breaking.
He's way better than the unbalanced array but you see, that array is not even looking at its skills. It only cares for hitting figures.

Three different abilities are used to calculate defenses. Sometimes a class feature moves the AC-determining stat around, but you still need a third ability to keep reflexes up.
Hence, we're looking at the three better stats only.

  • Unbalanced: +4 +2 +0 or
  • Balanced: +3 +3 +1
  • His: +3 +2 +2

As you can see, he someway covers his weak point. After racials (and supposing racial bonuses don't both cover the same defense) we have:

  • +5 +3 +0 or +5 +2 +1 for defense-wary strikers.
  • +4 +4 +1
  • +4 +3 +2

...which means he still has some weakness.

Now, your powers usually ask you to raise two stats that contribute to the same defense.
This means your second best value is not available for defenses, and we get these figures:

  • +5 +0 +0
  • +4 +1 +1
  • +4 +2 +2

As you can see in this case the rolled character fares better, totalling a net +1 increase on both weak defenses. Calculations for human characters are a little bit different and we'd need to factor in the +1 to all non-AC defenses (very useful to unbalanced array characters).

So your array characters have a weak point, a low defense that is destined to become wider and wider in time as most builds go glass cannon and stack all their ability bonuses in the same two abilities. This character has a slightly more protected weak point, at the cost of slightly less protected other defenses. This of course is advantageous to him, but not that much.

Now, if the other players rolled at least two 16s they're not running behind the expected figures, but a single 16 is often enough to get a decent character, because riders (the effects based on a secondary ability) can be ignored for a good number of builds.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .