HPs gained per level

[Manzanita]

I have some thoughts on this that haven't been mentioned.

1. Doing your HPs by the "Greyhawk" method may result with each die class getting a different percentage of maximum HPs, but it does retain the average differential between dice, which is important. (By this I mean, the average for a d6 is 3.5; the average for a d8 is 4.5, a difference of 1, which is retained if you take a 4 for d6 and a 5 for d8). If the percentage difference bothers you, you could round up at odd levels, then down on even levels, which would give everyone basically 50% of max for levels after 1. (So for wizards, they would get 4HPs@1st , 3 @2nd, 2 @3rd, 3 @4th etc.

2. The designers of the game had to make some sort of assumption of how many HP/die would result for game balancing purposes. For instance. If you were to give 75% HPs/die or max HPs/die, a d12 Barbarian will come up 2 HPs ahead of a fighter per level, which comes to 4 HPs for 2 levels. A fighter gains a feat every other level. If you say a feat is equivalent to 3 HP, then not only is the barbarian getting 4/3 the feats as a fighter, but also the various other advantages of their class.

Thus from a game balance perspective, I think whatever HP method you use, the difference between the average gained from a d12 should be 1 HP higher than for a d10, and likewise for the other die classes.

I realize some groups don't put as much stock in game balace as others. I'm of the school that I like my players to be pretty well balanced, or at least have that oportunity. HP rolls can really make or break a PC more than I care to risk. But if you keep it stricktly random, then you're not disturbing the inherent checks and balances of the game.

 

[Curugul]

Not

Entirely true imo.

"But if you keep it stricktly random, then you're not disturbing the inherent checks and balances of the game."

We frequently have seen people with a barbarian roll 1, 3, 2, 2. The rogue AND the wizard had higher hitpoints. It was, frankly, stupid. Take a look at how much the ranger gained from 3.0 to 3.5, and the balance? D10 to d8 hitpoints. THATS how much of a balance issue the game designers apparently feel hitpoints are. Yet they're stupidly random.

Does a rogue roll 1d8 skill points every level? Does a cleric roll to see if he gains all his new spells? No? They why should a Fighter have a CORE part of his power so randomly determined? Hitpoints are at the core of a fighter character. That a rogue can easily have more than a barbarian due to rolls is STUPID.

This is something I was hoping 3.5 would fix.


Curugul

 

[A'koss]

You can add my name to list of people who think random hit point rolls can really suck the fun out of a game for players. I've seen exactly the same situation as Curugul time and again, and if it's bad for a barbarian, rolling 3 straight 1's for a wizard - you might as well retire.

For years we've used static HP values and generally on the high side. Wizards get 4, Rogues 5, Clerics 6, Fighters 7, Barbarians 8. This helps a little at higher levels by keeping the inevitable HP disparities between characters a little more managable.

 

[Norfleet]

I find this re-roll all method intriguing though. I assume that you still get max hits for first level and that you re-roll your new current level as well (these were not mentioned).

That's correct. Although you don't "re-roll" your new current level, because re-rolling something requires that you have rolled it before, and if you're just levelling in it, you're rolling it for the first time. Basically, your hitpoints are regenerated as if you were being created at that level, but you will always gain at least one hit point over your previous level.

For example, I am risk adverse with money. I wouldn't accept a 50% chance to win $20,000 if the chance cost $10,000 -- I value the $10k I have more than the $10k I stand to gain. If I were risk-neutral, I might well take that chance -- and I certainly would take it if I were risk-loving.

Well, there's more to being risk-adverse or loving than just the raw quantity stated that's involved. Consider: You have $10K. You owe the mob $20K. You have a 50% chance of winning $20K, if you wager your $10K. Regardless of whether you have $0(tried+lost) or $10K(didn't try), if you don't cough up that $20K tonight, you're going to sleep with the fishes. Go for it?

It all comes down to what that money, or hitpoints, means. Ultimately, it may turn out to be a binary outcome: A chance to live, or certain death. It rather handily alters the equation if, while on the surface, you may be risking the loss of your $10K, on a higher level, you really have nothing to lose.

Even if the probability of winning may be small, if the effect of not trying and failing ultimately amounts to the same thing, even the most risk-adverse may be faced with no real choice in the matter, like a barbarian with 89 hitpoints levelling up before facing down a PWK: Take the 6, or roll them bones?

 

[CRGreathouse]

Well, there's more to being risk-adverse or loving than just the raw quantity stated that's involved. Consider: You have $10K. You owe the mob $20K. You have a 50% chance of winning $20K, if you wager your $10K. Regardless of whether you have $0(tried+lost) or $10K(didn't try), if you don't cough up that $20K tonight, you're going to sleep with the fishes. Go for it?

It all comes down to what that money, or hitpoints, means. Ultimately, it may turn out to be a binary outcome: A chance to live, or certain death. It rather handily alters the equation if, while on the surface, you may be risking the loss of your $10K, on a higher level, you really have nothing to lose.

Even if the probability of winning may be small, if the effect of not trying and failing ultimately amounts to the same thing, even the most risk-adverse may be faced with no real choice in the matter, like a barbarian with 89 hitpoints levelling up before facing down a PWK: Take the 6, or roll them bones?

In both examples, I would say that the normally risk-adverse person is temporarily risk-loving through circumstances. That is, in both cases he values his first (hp or $) less than his (101st hp or 20,000th $).

Of course, if the person might well return to his/her risk-adverse ways after 101 hp or $20k, but that's beside the point -- until he/she has the threshhold, he/she's risk-loving.

 

[Jeff Wilder]

In both examples, I would say that the normally risk-adverse person is temporarily risk-loving through circumstances.

Fair enough, but that falls apart in the D&D context, because the circumstances that make your "risk-neutral" HP choices actually mistakes don't change. They're the very rules of the game.

You can say the player who thus makes more cautious choices than the math you presented would indicate -- because he is cognizant that D&D punishes low HP totals more that it rewards high HP totals -- is being "permanently temporarily risk-averse," if you like, but that's a little too cumbersome for me. I'm just gonna call him "smart."


Jeff

 

[CRGreathouse]

Fair enough, but that falls apart in the D&D context, because the circumstances that make your "risk-neutral" HP choices actually mistakes don't change. They're the very rules of the game.

I'm not sure I agree. In the short term, risk-adverse behavior is rewarded, but in the long-term the extra hit points a risk-neutral player gains should make a cushion of sorts.

What numbers would you tend to stay on, as a player? I'll run some numbers for you, short-term and long-term.

 

[Jeff Wilder]

What numbers would you tend to stay on, as a player? I'll run some numbers for you, short-term and long-term.

Hmmm. Well, it really does depend on a lot of factors (not least being how previous levels have gone), but since you said "tend" ...

d12, I'd stay on a 5. d10, I'd stay on a 4. d8, I'd stay on a 4 (though it's close). d6, I'd stay on a 3. d4, I'd stay on a 3.


Jeff

 

[CRGreathouse]

d12, I'd stay on a 5. d10, I'd stay on a 4. d8, I'd stay on a 4 (though it's close). d6, I'd stay on a 3. d4, I'd stay on a 3.

I'm assuming that you can't roll down from a d4 to a d3 or something, so I'm disregarding the d4 stay on 3.

Your averages:
d4 2.5
d6 3.833333333
d8 5.1875
d10 6.45625
d12 7.81875

These are substantially similar to my numbers. The d4, d6, and d8 are exactly the same (since we stay on the same numbers). Over 12 levels, it will only average 3hp behind the risk-neutral average.

The risk-neutral method is significantly more likely to get high numbers (76% vs. 65% to be in the upper half) for the d12, but more likely to get low numbers as well -- twice as likely to have a 1 or 2 (1.5% vs. 0.7%).

I can post the whole breakdown, if you'd like.

 

[CRGreathouse]

I have an example that might help. Consider three feats (two of which I just made up, natch): Lesser Toughness (does nothing), Toughness (+3 hp, can be taken multiple times), Greater Toughness (+5 hp, can be taken multiple times, requires Lesser Toughness)

On first level, the extremely low-Con elven wizard takes Toughess, since it boosts his hitpoints fourfold. On level three, he takes Toughness again, since he wants his hit points to go up -- he needs them! The same thing happens on levels 6, 9, and 12.

An identical wizard takes Lesser Toughness on level 1 and Greater Toughness for the remaining levels. His hit points are worse up to level 9, but they're better from there on -- every three levels he gains 2hp on the other wizard.


With the risk-neutral character vs. a risk-adverse character, the r-n's hit points aren't always going to be lower (at every point, they tend to be higher) -- but at high levels, they're all but certain to be better.


Edit: I just ran some quick calculations on the various stratagies, by the way. There may be mistakes, but it looks good at inspection. I found 65 rational risk-adverse stratagies (yours is one of them), 1 risk-neutral stratagy, 335 rational risk-loving stratagies (for example, discarding all but the best result), 3465 mixed rational stratagies (staying on a 3 for a d12 but on a 4 on a d10), and 2295 irrational stratagies (for example, rerolling a 12 on a d12).

I'm not sure what that has to do with anything, but it's kind of cool.