Yeah, I have my player reroll anything less than half average (so a barbarian can reroll a 6 or less, a fighter 5 or less, etc.). This gives them about 75% max HP, and I have no problems doing the same for many of the monsters they face. Running a group through WLD right now and have tallied a dozen PC deaths to date, so I've not really had problems with them being overpowered (a few extra HP wont save you if you make a blunder, are rololing bad, and IM rolling hot, as my PCs have learned repeatedly).
Hit point generation methods
- Thread starter Liminal Syzygy
- Start date
CalicoDave
Explorer
My group plays that you can either roll or take the average. We allow re-rolls of 1's until something other than a one comes up. Becuase of the re-roll 1's thing the average goes to 3 for a d4, 4 for a d6, etc. Out of about seven players only one or two regularly roll, most just take the average.
Jeff Wilder
First Post
Does anyone use a method involving rerolling all Hit Dice at each level? Maybe with some other mechanism involved?
One method I considered was using normal HP rolling, but allowing players to reroll all their HD (except 1st level, which is still max) any time they gain a level, if they choose to.
With this method players are highly unlikely to stay HP-screwed for more than a level or two, but it otherwise has a very mild effect on "HP inflation," which, IME, is a fairly serious problem with some of the more liberal methods suggested here.
One method I considered was using normal HP rolling, but allowing players to reroll all their HD (except 1st level, which is still max) any time they gain a level, if they choose to.
With this method players are highly unlikely to stay HP-screwed for more than a level or two, but it otherwise has a very mild effect on "HP inflation," which, IME, is a fairly serious problem with some of the more liberal methods suggested here.
Jondor_Battlehammer
First Post
Simillar to an idea i've had.Fundin Strongarm said:
D4 = D4.......2.5
D6 = D4+2....4.5
D8 = D6+2....5.5
D10 = D6+4....7.5
D12 = D8+4....8.5
or
D4 = 2D2
D6 = 2D3
D8 = 2D4
D10 = 2D5
D12 = 2D6 or 3D4
to wilder_jw,
As for re-rolling ALL HP each level, I hope you mean to see if they gain more HP. Not to gain a new total, even if lower than the original.
This would put players on the track to average HP, unless they had lucky rolls to begin with. Not a bad thing, but it would be easier to achive by giving out average HP to start.
The real thing to worry about is how powerful you want magic to be. Monsters can be given more HP too, but a Fireball does a static amount of damage, and you only get so many per day. A 5th level PC with average HP off of a D8 and a 12 CON would be 31. A 5th level Fireball does an average of 18 damage. Does this look high or low to you? Or, By HP inflation, do you mean that you want a 10D6 Fireball to still mean something at higher levels?
I suppose allowing a reroll at a -1 to the max roll, a second reroll at a -2 to the max roll, etc, could keep things from getting to crazy. A barbarian rolls a 2, decides to reroll, and gets a 12. But because it was on a reroll, his max is 11, (an 11 or lower would have been unmodified). Ofcourse, rerolling using the next lower dice already does this.
Last edited:
Here are the nonstandard methods I've used or seen used:
1. Roll twice. Take the better roll. (Raises the mean but keeps the same theoretical minimum and maximum).
2. Roll twice. If either roll is above the mean, take the better roll. If both rolls are below the mean, add them up. (Raises the mean more and almost abolishes below-average results; raises the minimum from 1 to 2.)
3. Reroll all hit points at each level. If the total is less than or equal to your previous hit point total, add one to your previous total instead. That's not one + con, it's just one, period. (Uniquely out of these methods, should keep the mean about the same; at second level and possibly for a little while afterwards, it actually *LOWERS* it. I suspect it actually raises the mean a tiny bit at high levels, as a really good roll is more likely to have a lasting impact than a really bad one, but overall it should keep you pretty close to the mean. Same theoretical minimum and maximum, but makes it even less likely that you'll actually be anywhere near either.)
4. Roll once. If your roll is below half the maximum (or to put it another way, less than mean - .5) treat it as though you had rolled half the maximum. (Obviously raises the mean and minimum, probably about the same mean overall as method 2 but I haven't checked in detail.)
5. As 4 above, except that you still get half the maximum even if Con penalties would reduce you to less than that. (Con *bonuses*, however, are treated as in 4.) Moreover, at first level, you ignore Con penalties entirely. (The same as 4 for most characters, but it makes a low Con somewhat playable.)
The last game I was a player in used 3, and I briefly did as well; both games I currently DM use 5.
EDIT - forgot these:
6. 1d4 + (normal maximum - 4). So for example, D6 would become D4+2, while D10 would become D4+6. This was originally from a document available on the net called Enhanced Dungeons & Dragons, which also allowed you to trade points from the fixed number part (but not the D4) for other benefits. (What I don't like about this one is that it doesn't help the characters at the low end, who need it most, very much if at all; on the other hand, for characters at the high end this is even more generous than 2, 4 and 5. As someone mentioned earlier high variance is harder to deal with than raising the mean, and this *increases* variance.)
7. 1d4 + (normal maximum -2). Uniquely out of all these methods, this not only raises the minimum and the mean, but breaks the taboo on raising the maximum. In fact, your mean result with this method is greater than the maximum result using the book rule. Obviously extremely generous; and the flat two extra points for everyone somewhat reduces the variance problem too (technically there's just as much, but it's easier for a 45 hp character and a 105 hp character to adventure together than if you replace those numbers with 25 and 85).
1. Roll twice. Take the better roll. (Raises the mean but keeps the same theoretical minimum and maximum).
2. Roll twice. If either roll is above the mean, take the better roll. If both rolls are below the mean, add them up. (Raises the mean more and almost abolishes below-average results; raises the minimum from 1 to 2.)
3. Reroll all hit points at each level. If the total is less than or equal to your previous hit point total, add one to your previous total instead. That's not one + con, it's just one, period. (Uniquely out of these methods, should keep the mean about the same; at second level and possibly for a little while afterwards, it actually *LOWERS* it. I suspect it actually raises the mean a tiny bit at high levels, as a really good roll is more likely to have a lasting impact than a really bad one, but overall it should keep you pretty close to the mean. Same theoretical minimum and maximum, but makes it even less likely that you'll actually be anywhere near either.)
4. Roll once. If your roll is below half the maximum (or to put it another way, less than mean - .5) treat it as though you had rolled half the maximum. (Obviously raises the mean and minimum, probably about the same mean overall as method 2 but I haven't checked in detail.)
5. As 4 above, except that you still get half the maximum even if Con penalties would reduce you to less than that. (Con *bonuses*, however, are treated as in 4.) Moreover, at first level, you ignore Con penalties entirely. (The same as 4 for most characters, but it makes a low Con somewhat playable.)
The last game I was a player in used 3, and I briefly did as well; both games I currently DM use 5.
EDIT - forgot these:
6. 1d4 + (normal maximum - 4). So for example, D6 would become D4+2, while D10 would become D4+6. This was originally from a document available on the net called Enhanced Dungeons & Dragons, which also allowed you to trade points from the fixed number part (but not the D4) for other benefits. (What I don't like about this one is that it doesn't help the characters at the low end, who need it most, very much if at all; on the other hand, for characters at the high end this is even more generous than 2, 4 and 5. As someone mentioned earlier high variance is harder to deal with than raising the mean, and this *increases* variance.)
7. 1d4 + (normal maximum -2). Uniquely out of all these methods, this not only raises the minimum and the mean, but breaks the taboo on raising the maximum. In fact, your mean result with this method is greater than the maximum result using the book rule. Obviously extremely generous; and the flat two extra points for everyone somewhat reduces the variance problem too (technically there's just as much, but it's easier for a 45 hp character and a 105 hp character to adventure together than if you replace those numbers with 25 and 85).
Last edited:
Heres another idea I've occasionally considered:
Roll 1d6 and whatever you would roll under the normal rules, and take the higher result. Cuts down on really poor rolls and really materially helps mages, but won't have much effect on the people at the high end.
Any number of variants are possible. For example, you could make the second die depend on race; dwarves might get a d8 or their normal die, whichever is better, while halflings would get a mere d4 for their "plan B" die. I suspect this is unbalanced under 3E, (it helps races that are already good more than ones that are currently a little weak, under the most obvious implementation) but it's a theme that could be played around with.
Or if you dislike variance and want to reduce the difference betwen classes somewhat: Roll a d6, a d10 (or d12) and whatever you would normally roll, and take the middle result.
Roll 1d6 and whatever you would roll under the normal rules, and take the higher result. Cuts down on really poor rolls and really materially helps mages, but won't have much effect on the people at the high end.
Any number of variants are possible. For example, you could make the second die depend on race; dwarves might get a d8 or their normal die, whichever is better, while halflings would get a mere d4 for their "plan B" die. I suspect this is unbalanced under 3E, (it helps races that are already good more than ones that are currently a little weak, under the most obvious implementation) but it's a theme that could be played around with.
Or if you dislike variance and want to reduce the difference betwen classes somewhat: Roll a d6, a d10 (or d12) and whatever you would normally roll, and take the middle result.
Gez
First Post
My own method is pretty simple.
Max HD at first level, for PCs and "elite" NPCs.
For each level thereafter, the player makes a public HD roll, and the DM makes a secret HD roll. The player can then choose his own (known) result, or the DM's (unknown) result.
For elite NPCs, it means that (if I do not simply pick and choose the NPC's total) I can reroll any and every of his HD rolls once.
As an added rule, I sometime use "level 0" characters, building upon the old 3e's "apprentice level" for 1st-level multiclass. An apprentice is considered level 0 or 1/2, depending on what's the most practical, for capacities based on level (including BAB and saves). They have (Base Skill Points + Int bonus) times 3 skill points, with a max rank of 3. Average apprentices have 1 hit point, elite apprentice roll for hit points (with the DM secret roll). Upon becoming full-fledged 1st level, they max out their HD, and gain BSP+IB skill points.
Max HD at first level, for PCs and "elite" NPCs.
For each level thereafter, the player makes a public HD roll, and the DM makes a secret HD roll. The player can then choose his own (known) result, or the DM's (unknown) result.
For elite NPCs, it means that (if I do not simply pick and choose the NPC's total) I can reroll any and every of his HD rolls once.
As an added rule, I sometime use "level 0" characters, building upon the old 3e's "apprentice level" for 1st-level multiclass. An apprentice is considered level 0 or 1/2, depending on what's the most practical, for capacities based on level (including BAB and saves). They have (Base Skill Points + Int bonus) times 3 skill points, with a max rank of 3. Average apprentices have 1 hit point, elite apprentice roll for hit points (with the DM secret roll). Upon becoming full-fledged 1st level, they max out their HD, and gain BSP+IB skill points.
Ant
First Post
The players can choose from:
Average rounded down
or
A roll off against the DM (who, of course, keeps his roll secret)
So far no-one's gone for the former option but we've only been trying this mix of methods for a few levels. Previously the latter choice was the core rule of take what you get with one roll (making the former option a lot more attractive.)
I really wanted to retain the random element of hit-points but also wanted to tip the odds in the player's favour. So far it's working very well (ie fun for me and the player's are ending up with higher than average HPs)
Average rounded down
or
A roll off against the DM (who, of course, keeps his roll secret)
So far no-one's gone for the former option but we've only been trying this mix of methods for a few levels. Previously the latter choice was the core rule of take what you get with one roll (making the former option a lot more attractive.)
I really wanted to retain the random element of hit-points but also wanted to tip the odds in the player's favour. So far it's working very well (ie fun for me and the player's are ending up with higher than average HPs)
Jeff Wilder
First Post
Another possibility: HP bribery.
Roll HP normally. Any player who writes a log for at least half the sessions of his last level gets the option to reroll.
No, I meant to gain a new total. Obviously, players would only choose to do so if they were below average, which is exactly the effect I want from an alternate HP system. I want to eliminate long-term below average results, while (a) still maintaining some randomness, and (b) while not overly inflating HP totals.
Why do you say you "hope" I didn't mean that?
Roll HP normally. Any player who writes a log for at least half the sessions of his last level gets the option to reroll.
Jondor_Battlehammer said:
No, I meant to gain a new total. Obviously, players would only choose to do so if they were below average, which is exactly the effect I want from an alternate HP system. I want to eliminate long-term below average results, while (a) still maintaining some randomness, and (b) while not overly inflating HP totals.
Why do you say you "hope" I didn't mean that?
Similar Threads
