Why I Think Rolling For Hit Points is a Bad Thing

[Szatany]

Fighter: +4+1d4
Cleric: +2+1d4
Rogue: +1+1d4
Wizard: +0+1d4

I never saw the need why any class should be more random on HP than another.

I would like something a bit different. Each class rolls d6 and adds a number based on its original HD. CON grants cumulative advantage or disadv.

Wizard 1d6/level
Cleric/Rogue 1d6+1/level
Fighter 1d6+2/level
For each point of CON bonus, roll an extra d6 and choose highest.
For each point of CON penalty, roll an extra d6 and choose lowest.

So a CON 12 wizard will roll 2d6 and choose best at each level, a sickly cleric with CON 8 will roll 2d6 and choose lowest, and then add 1. A CON 17 fighter will roll 3d6, choose best and then add 2 at each level.

 

[FireLance]

Actually, all these complicated rolling methods reminded me of one reason why I like static hit points.

If I ever lose the character sheet, I don't have to try and remember what my hit point rolls were. I can just re-calculate them from scratch.

EDIT: The same logic also applies to point buy ability scores and having a fixed, level-dependant schedule of magic item acquisition and allowing the player to choose what items the PC gains.

 

[Fifth Element]

Yes but you do not get to use the sum of your attack rolls and saving throws.

In some sense you do, because each roll counts.

But my point is, you get to make many individual d20 rolls for attacks, saves, checks etc, enough so that luck will even out over time.

But you get one hp roll per level, and that one roll can have a great effect on the character throughout that level. Every time that character takes damage, that one roll comes into play.

It might be enormous, it might be not. Depends on the level. At level one, a fighter with 15 hp has a minor advantage over a fighter with 10 hp.
At high level, a fighter with 150 hp has indeed a strong advantage over a fighter with 100.

At level one, 5 hit points is a pretty big advantage. If monster cause 5 hp damage, you can take 3 hits instead of 2.

Since dangers scale roughly with level, a 50% advantage will always be a big advantage. It can vary by level, but I don't think it's ever a negligible advantage.

But why would you build a game to use one set of maths (ie rolled hp) but then test it on a different basis (ie fixed hp)? That doesn't make much sense to me.

Because we're not playtesting advancement mechanics?

Also, note that the hp values they present can result from die rolls. So it's not like they're presenting characters that are impossible according to the random die roll rules. Just pretend they rolled those numbers.

 

[lutecius]

I strongly disagree.

If every character of X class gained the same number of hit points every level, it would be boring--characters of the same class, theme, and background would look depressingly the same. Rolled HP gives us yet another means of differentiating one character from another.

it makes some characters objectively, mechanically better than others with the same class and backround and not even that different story-wise. how interesting is that?

The randomness of attack rolls can mean that a party is wiped.
The randomness of saves can mean that someone dies when they really shouldn't.

some unpredictability in the outcome of each action makes the game exciting. knowing that your character will have an overall handicap or undeserved advantage for one level or more, not so much.

to each his own, I suppose.

 

[IronWolf]

Here's the thing... WotC can offer multiple options in the core rules so that each gaming group can decide for themselves how to generate hit points. In fact, I'd be shocked if they didn't do this. It seems like such a common sense set of options to me. Just offer guidance to the DMs about each option so that they can make an informed choice about which variant is right for their group.

Yeah - this is one of the easiest issues to solve within your own group. Hit point generation is pretty easy to deal with through any number of house rules.

 

[KidSnide]

To me, it is poor game design for one character to be vastly weaker than another just because of a few unlucky die rolls. Sure, lots of things in the game come down to the luck of the dice, but as I pointed out, with things like ability scores, the difference is nowhere near as great. And characters have means of recovering from failed saving throws or missed attack rolls. At the very worst, they die and can have their character resurrected. This is a far cry from having a character who is so irreperably fragile that he is practically worthless.

I strongly recommend that in DnDNext that players gain a set number of hit points per level based on their class.

Essentially, you are 44.987% likely to get hit points within the 11-point 50-60 range, and 91.208% likely to get hit points within the 31-point 40-70 range.

Given that average hit points are 55, that's a 91.208% chance that randomly rolled hit points are within 15 points of the average.

The math errors shouldn't obscure the OP's essentially correct point: HPs are too important to character balance to be random by default. Yes, there is a 91% chance that characters will be in the 40-70 range, but there is a huge difference in effectiveness between a 40 hp fighter and a 70 hp fighter.

But the math for this is even worse than this would suggest because balance is a party problem, not a character problem. If there is a 9% chance that any one character has abnormally high or low hit points, there is a 37.5% chance that any five person party will have at least one member with this problem.

Even worse, the distribution evens out somewhat at higher levels, but new players almost always learn the game at low levels. One or two capricious die rolls can have a profound impact on whether someone finds D&D fun or not. Pitty the poor newbie who is trying to play a front line fighter, but rolls a 1 and a 3 on his first two hit point rolls.

Rolling HPs clearly needs to be an optional module, but it's not a good default rule for new players.

-KS

 

[Cadfan]

You can't fix the negative effects of randomization by using random skew. That just makes it all the more tragic when that one player actually fails to roll well, even though you sweetened the odds for him.

Take rolling for ability scores. You're right that the odds of rolling truly badly are fairly low. You get to roll 3 dice and add them up, so its bell curved. You get to assign the scores where you please, so you can bury the worst scores. In 3e, you even threw out ability score sets that were beneath certain thresholds. But you know what? Even with all of that, there's still going to be one guy out there who rolls the minimum legal set of ability scores, and there's still going to be one guy who rolls ridiculously high.

And take hit points. Someone in the thread suggested rolling twice and taking the better roll. You can do that. It will reduce the odds of someone rolling 1s repeatedly as they level up. But you know what? There's still going to be someone out there who repeatedly rolls snake eyes when they level up. It will happen less often, but it will still happen.

So you have to ask- is randomization of these character elements really adding anything that 1) justifies the cost, and 2) can't be replaced with some other mechanism?

I think the answer to both questions is no.

 

[Cadfan]

This is wrong. After first level they are subject to bell curve dynamics.

The life expectancy of a male residing in the US is 75.6.

The life expectancy of a 50 year old male residing in the US is substantially higher, because he has failed to die before attaining the age of 50.

The average hit points of a level 10 D&D character might be X.

But the average hit points of a level 10 D&D character who rolled a 1 for hit points at level 2 is Y, where Y is less than X by the difference between 1 and the mid point of his hit die.

Dice have no memory. If you roll badly early, you shift your expected long term value to the left.

 

[Steely_Dan]

I guess the pregen characters took average.

It should be a choice, I know one guy who loves to roll for HP (in 3rd Ed he would consistently roll 3s and 4s for his Psion's HP), and I have another buddy who practically breaks out in a nervous sweat when rolling for HP (he would roll poorly so he always put a high score in Con), so better for him to just take average.

 

[Walking Dad]

Rolling this (take 3 best dice):

STR: 4d6
DEX: 4d6
CON: 5d6
INT: 7d6
WIS: 4d6
CHA: 4d6

Resulted in this:


STR: 10
DEX: 11
CON: 10
INT: 14
WIS: 12
CHA: 12


And another time I rolled this for a dwarf (3d6 in order, I'm currently unable to find the link:


Str: 18 (+2 level)
Dex: 15
Con: 18 (+2 racial)
Int: 12
Wis: 14
Cha: 6 (-2 racial

Don't tell me about bell curves...