Mordeth said:
What I dont understand is the random factor away from combat.
You dont roll for feats.
You dont roll for skill points.
You dont roll for class abilites.
Away from combat, I "think" that is the only rolls that we do on our character sheets. I just dont feel its balanced that person can play a fighter for example, and be crippled by ba hp rolls. The HP is the meat and drink of a tank class. How many sorcs would play the class if they had to roll for how many spells the coul cast at each level? "Damn, I can only cast 1 1st level spell and 2 2nd level spells at 4th
"
Well said. This is the central issue. Why are HPs the only permanent "stat" that is randomly determined? It makes about as much as determining AC randomly (e.g. base it on a permanent roll of d20 or 3d6 instead of always using 10... If you roll 15, good for you, your AC will be always be 5 points higher than normal. If you roll 5, it's just too bad.)
Personally, I prefer a fixed number of HPs that matches the expectation of rolling except 1st level:
1st: Full HD. Even: HD/2+1. Odd: HD/2.
...except for 1st level, this is what the MM uses for monsters.
Alternatively, the randomness can be reduced by rolling three dice and taking the middle number. This gives the same average number of HPs but the variation is much lower. The chance/risk of rolling max/1 is reduced the most. (This can also be used as a variation for "take 10" on skill checks etc.)
The formula for probabilities is: Prob = [(Roll-1)*(Die-Roll)*6+(Die-1)*3+1] / [Die^3]
Except for rounding errors this gives the probabilities:
d4: 16% 34% 34% 16% (instead of all 25%)
d6: 7% 19% 24% 24% 19% 7% (not all 17%)
d8: 4% 11% 16% 18% 18% 16% 11% 4% (not all 12.5%)
d10: 3% 8% 11% 14% 15% 15% 14% 11% 8% 3% (not all 10%)
d12: 2% 5% 8% 10% 12% 12% 12% 12% 10% 8% 5% 2% (not all 8%)
d20: 1% 2% 3% 4% 5% 6% 7% 7% 7% 7% ... (not all 5%)