Good Statistical Scores?

[Tanuki Tales]

I'm curious and I'm sure the folks on here have taken the time to do all the math, so here we go:

What are the bare minimum necessary scores for a character to reliably have a chance of overcoming every level appropriate obstacle that is being placed before them?

I'm speaking mostly 3.P here, but this can be about any game system and what your opinions on the bare minimum level of "excellence" that a PC needs in their array in order to be a successful hero, who's accomplishment rely heavily on the uncaring whims of the dice, are.

[Saph]

Although I've never tried it in 3.5, I'm pretty sure the bare minimum is close to zero. It's quite possible to build a character with straight 10s in all their ability scores that can deal with any level-appropriate challenge, but it depends heavily on optimisation level and class. Some classes are much more stat-dependent than others - something like a Paladin needs multiple good stats, something like a Wizard needs one good stat, and something like a Warlock can actually manage with no good stats at all.

For Pathfinder, I know for sure that you can build a Summoner on a 0 point buy that's almost as effective as one on a 15 point buy (and better than most other classes built on far more than that)

[Tanuki Tales]

So, just throwing an array out there from the top of my head, any kind of character of any class would be able to be successfully built on an 18,16,14,13,13,13 spread for example?

Even the Monks and Paladins in the crowd?

[Tyndmyr]

I'm curious and I'm sure the folks on here have taken the time to do all the math, so here we go:

What are the bare minimum necessary scores for a character to reliably have a chance of overcoming every level appropriate obstacle that is being placed before them?

I'm speaking mostly 3.P here, but this can be about any game system and what your opinions on the bare minimum level of "excellence" that a PC needs in their array in order to be a successful hero, who's accomplishment rely heavily on the uncaring whims of the dice, are.

Class is far more notable than stats are. One sixteen and a fourteen, the rest fives? Most casters are good to go. Your paladin is crying in a corner, though.

Hell, one moderate roll is all you need to be an effective DFA.

[Saph]

So, just throwing an array out there from the top of my head, any kind of character of any class would be able to be successfully built on an 18,16,14,13,13,13 spread for example?

Yes. That's a 47 point buy if I'm counting it right, so you'd kind of expect it to be more than enough. Book NPCs are built on 25, and most people who favour point buy use 28 or 32.

[Tanuki Tales]

Yes. That's a 47 point buy if I'm counting it right, so you'd kind of expect it to be more than enough. Book NPCs are built on 25, and most people who favour point buy use 28 or 32.

Is there a median point on the curve that's been determined?

An ability score spread that would favor all the classes (Truenamer excluded of course) favorably with minimal optimization?

I saw this argued in another thread and the concept of the discussion intrigued me.

[Yora]

I think with total modifiers that sum up to +6 you've got something quite solid to work with.

[Tanuki Tales]

I think with total modifiers that sum up to +6 you've got something quite solid to work with.

So 18, 6, 12, 10, 14, 12 is just as reliable as straight 13s?

[Yora]

Depends on what you want to do with it.

But thinking of it, +6 seems actually rather high. I guess +4 or +3 would also be okay and not completely suck.

[Tyndmyr]

Is there a median point on the curve that's been determined?

An ability score spread that would favor all the classes (Truenamer excluded of course) favorably with minimal optimization?

I saw this argued in another thread and the concept of the discussion intrigued me.

Actually, if Truenamer gets a solid Int and decent con, it's happy. It's just as SAD as a wizard.

There is no real median point. Higher point buy favors MAD classes, but you have to go fairly high for it to matter at all, and it's never going to make up all the difference. You could give straight 18s, and the SAD class is still going to benefit from buying less enhancement items.

Edit: I use a 34 point buy. This is sufficiently generous to allow somewhat MAD classes to be basically functional, without going entirely crazy.

[GungHo]

All 18s. Just like me in real life.

[Thialfi]

That all really depends on how you play. The way our group plays requires very powerful characters because their opponents are always played as very aggressive and intelligent. We have been a group since the summer of 1980 and I have never seen a dungeon that could be played with the pre-rolled characters that had even a remote prayer of avoiding a TPK. It isn't a case of the DM being smarter than the players either since we all take turns being the DM and it always works out the same.

[Tyndmyr]

That all really depends on how you play. The way our group plays requires very powerful characters because their opponents are always played as very aggressive and intelligent. We have been a group since the summer of 1980 and I have never seen a dungeon that could be played with the pre-rolled characters that had even a remote prayer of avoiding a TPK. It isn't a case of the DM being smarter than the players either since we all take turns being the DM and it always works out the same.

Honestly, as my friday session has become higher and higher op, I haven't noticed any real increase in minimal stats.

[Eldan]

D&D was built on the assumption that a normal-powered game was 25 Point Buy. Or at least I seem to remember that from the DMG. 28-32 was already considered high-powered. The "normal" stat generation method of 4d6b3 gives quite low scores on average, really.

[Yora]

Really normal is "3d6", which gets you an average of 10 and average modifiers of +0.

[Tyndmyr]

D&D was built on the assumption that a normal-powered game was 25 Point Buy. Or at least I seem to remember that from the DMG. 28-32 was already considered high-powered. The "normal" stat generation method of 4d6b3 gives quite low scores on average, really.

I calculated the average point buy given by 4d6b3 once. I don't recall it exactly, but it was in the 28-32 range.

I suspect they didn't do their math well.

Edit: Also, 3d6 averages 10.5, not 10.

[tyckspoon]

I calculated the average point buy given by 4d6b3 once. I don't recall it exactly, but it was in the 28-32 range.

I suspect they didn't do their math well.


28 is the rough average equivalency, IIRC. A lesser value in point-buy is considered of the same power because it can be more precisely and optimally employed; you don't waste random points on odd stat values that you have no intention of raising, for example.. a line of all odd values could be 5-20 points more expensive than the even values of the same modifier, but performs exactly the same in-game.

[Tyndmyr]

28 is the rough average equivalency, IIRC. A lesser value in point-buy is considered of the same power because it can be more precisely and optimally employed; you don't waste random points on odd stat values that you have no intention of raising, for example.. a line of all odd values could be 5-20 points more expensive than the even values of the same modifier, but performs exactly the same in-game.

It depends if you take the point buy of the average roll, or take the average point buy of rolls. They're notably different, and the second is substantially higher(because you get a greater point buy increase from a lucky good roll than you loose from an equally unlucky poor roll).

Note that all the really terrible sets are rerolled as well, as per standard rules, so this further increases the average.

So, while what you say is true, point buy is more useful for a given value than a random allocation, the standard rolling is not particularly low powered. In particular, you'll often see people say that straight 10s is "average". Not only is this not true for adventurers, who would immediately reroll if they got this, it's not even true for 3d6 with no rerolls.

[Morithias]

All 18s. Just like me in real life.

Are you an elf? Cause that's pretty much the most elvish arrogant claim I've ever seen.

[Chilingsworth]

shouldn't this thread be in the 3.5 forum? :smallconfused:

[Haarkla]

What are the bare minimum necessary scores for a character to reliably have a chance of overcoming every level appropriate obstacle that is being placed before them?

An ability score spread that would favor all the classes ... favorably with minimal optimization?

I recon 14, 10, 10, 8, 8, 6 is playable for any class.

[GungHo]

Are you an elf? Cause that's pretty much the most elvish arrogant claim I've ever seen.
Don't hate me just cause mom and dad had a good run at the table.

Did I mention charisma was my dump stat?


shouldn't this thread be in the 3.5 forum? :smallconfused:
Actually 4E has stats. The same ones. So do many games, and it's a good question to consider. What do you really need to succeed/feel successful?

A lot of games have a more controlled array (WoD), for example, but given your purpose, there a some arrangements more optimal than others.

Other threads to consider: What games de-emphasize main stats? What games over-emphasize main stats? Do the stats make the man, or are they a go-by?

[jvluso]

For those discussing math since I did this recently, these are the statistics for rolling Xd6b3 for values of X from 1-11. Spoilered for size.
{table]Dice rolled:| 1| 2| 3| 4| 5| 6| 7| 8| 9| 10| 11
mean: | 3.5| 7| 10.5| 12.2446| 13.4302| 14.2738| 14.9028| 15.3888| 15.7748| 16.088| 16.3466
std dev:| 1.87083| 2.41523| 2.95804| 2.84684| 2.60347| 2.36365| 2.15049| 1.96485| 1.80351| 1.66271| 1.53898
1|16.66666| 0| 0| 0| 0| 0| 0| 0| 0| 0| 0
2|16.66666|2.777777| 0| 0| 0| 0| 0| 0| 0| 0| 0
3|16.66666|5.555555|0.462963|0.077160|0.012860|0.0 02143|0.000357|5.954E-0|9.923E-0|1.654E-0|2.756E-0
4|16.66666|8.333333|1.388888|0.308642|0.064300|0.0 12860|0.002500|0.000476|8.931E-0|1.654E-0|3.032E-0
5|16.66666|11.11111|2.777777|0.771604|0.192901|0.0 45010|0.010002|0.002143|0.000446|9.096E-0|1.819E-0
6|16.66666|13.88888|4.629629|1.620370|0.527263|0.1 67181|0.052869|0.016849|0.005427|0.001766|0.000579
7| 0|16.66666|6.944444|2.932098|1.157407|0.443672|0.1 67538|0.062633|0.023219|0.008542|0.003119
8| 0|13.88888|9.722222|4.783950|2.186214|0.958076|0.4 07593|0.169324|0.068944|0.027593|0.010881
9| 0|11.11111|11.57407|7.021604|3.806584|1.959019|0.9 80581|0.483265|0.236224|0.115072|0.056037
10| 0|8.333333| 12.5|9.413580|6.044238|3.594393|2.052969|1.146214| 0.631215|0.344481|0.186780
11| 0|5.555555| 12.5|11.41975|8.551954|5.799897|3.718349|2.301002| 1.389692|0.824312|0.482072
12| 0|2.777777|11.57407|12.88580|11.32973|8.862740|6.4 86839|4.557172|3.118222|2.097184|1.394786
13| 0| 0|9.722222|13.27160|13.56738|12.10133|9.994784|7.8 68941|6.001996|4.479016|3.290861
14| 0| 0|6.944444|12.34567|14.85339|15.12345|14.03070|12. 26661|10.29226|8.379759|6.668173
15| 0| 0|4.629629|10.10802|14.28755|16.65380|17.43184|17. 05753|15.95246|14.45099|12.79136
16| 0| 0|2.777777|7.253086|12.02417|16.18441|19.32441|21. 36369|22.39842|22.60014|22.15758
17| 0| 0|1.388888|4.166666|7.844650|11.86342|15.76110|19. 21939|22.05539| 24.1937|25.63525|26.43006563
18| 0| 0|0.462963|1.620370|3.549382|6.228566|9.577546|13. 48468|17.82595|22.47732|27.32248
avg pb|-61|-13.66667|15|28.23611|37.96682|45.65432|52.00686|57 .38566|62.004|66.00537|69.49429
[/table]
explanation of each row
dice rolled: X in Xd6b3. I assume that 1d6b3=1d6 and 2d6b3=2d6
mean: The mean, or average
std dev: The Standard Deviation
#: the chance of getting # when you roll Xd6b3
avg pb: The average point buy you get when you roll Xd6b3 6 times. The numbers are worth the following:
{table]roll|point worth
1|-19
2|-15
3|-11
4|-8
5|-5
6|-3
7|-1
8|0
9|1
10|2
11|3
12|4
13|5
14|6
15|8
16|10
17|13
18|16[/table]
If you want me to recalculate it with a different downward extrapolation, just ask. I'll probably soon have more stats about these things, for varying definitions of 'soon.'

tldr: mean point buy for 4d6b3 is 28.23611

[Valameer]

Since this isn't in the 3.P thread, I'll try to handle a few other systems.

AD&D: You need at least a 9 in strength. Other stats don't even need to be above 6. Become the DM's most liked player (whether by being fun and interesting or by buying him lots of pizza) and play smart and cautiously and you're good to go. Be a fighter, because fighters rock right off the bat and always have stuff to do later on. Hope for the gauntlets of ogre's strength and the cool +5 sword (the DM loves you, remember?) Get a cool ring or two to blast like a spell-caster. Win at AD&D. Who had high stats in this game anyway? (answer: the cheaters did, but they still died :smalltongue:)

GURPS: Max one trick, dump the rest. Middle ages, no magic? Boost strength and cut heads off. Middle ages, high magic? Boost intellect and deathtouch or fire breath. Unfortunately, combat-wise, the game favours one trick ponies over well rounded characters.

Star Wars: Saga: On the right build, one 16, drop the rest will do very well. The most MAD characters are Jedi, but even they can be boiled down to wisdom and charisma only builds (or strength and con only). If I had to streamline, I would make a 16 int tech-specialist with maxed mechanics and use computer. There's a lot of stuff you can do when you have the power to control all technology at your fingertips.

Palladium (any): Forget stats, when OCCs and equipment are all that matter. Heck, they even dictate your stats sometimes, too.

[Eldan]

Actually, a caster with 12 in his casting stat should be more or less playable in 3.5. Put a point at it on level 4, so you can get third level spells, then hope for stat-boosting items.

[sonofzeal]

If you want an array that anyone can make use of but isn't innordinately useful for anyone, try 16 / 16 / 14 / 14 / 12 / 12. MAD classes love you, but they're usually the weakest. And many casters would prefer 18 / 14 / 10 / 8 / 8 / 8, but they don't need the help.

[Jay R]

What are the bare minimum necessary scores for a character to reliably have a chance of overcoming every level appropriate obstacle that is being placed before them?

Slightly above average intelligence, and an especially quick mind.
Good, detailed knowledge of the rules.
Clear, quick ability to do arithmetic in your head, especially simple probabilities.
Basic psychology, focused on what appeals to the DM and his characters.
Good genre knowledge and genre savviness (two very different things) that are consistent with the DM's. (This includes knowing the period well in any historical game.)
Good understanding of strategy and tactics, especially the uses of the abilities of the party.

The above is far more important than anything on your character sheet, unless it's the supercannon that blows up all your opponents trivially.

[ko_sct]

I'm surprise it hasnt been mentioned that warlocks can be built to work quite well whit 8 in all their stats.

[Tyndmyr]

I'm surprise it hasnt been mentioned that warlocks can be built to work quite well whit 8 in all their stats.

Actually, I can build a truenamer the same way. So long as I'm still getting at least 1 skill point/level, I'm in business.

[Tengu_temp]

D&D was built on the assumption that a normal-powered game was 25 Point Buy. Or at least I seem to remember that from the DMG. 28-32 was already considered high-powered. The "normal" stat generation method of 4d6b3 gives quite low scores on average, really.

Thought it was also built on the assumption that a typical cleric will heal and a typical wizard will throw magic missiles and fireballs, so the result might "slightly" differ from the creators' intentions here.