Quote from: VisionStorm on December 17, 2023, 10:34:20 PM

Quote from: migo on December 17, 2023, 06:34:55 PM

Quote from: VisionStorm on December 17, 2023, 05:59:22 PM*snip*

With random generation, you're not expecting to be able to make a character according to your wishes. So the second point is automatically a non-issue. The first point, yeah, you can have character imbalance with random generation, but there you only have one issue to solve, not two.

Even with random generation, you might still want to build a character a certain way, it's just that the system doesn't allow you to, or limits your options. So it can still be an issue (which I've personally had, or dealt with player who had it), it's just you can do nothing about it. Ever.

But with point buy you at least have more control over your character, even if you can't get 100% what you want out of the gate. But you might still get it eventually. And the GM might even make adjustments or concessions to get it right away.

Sometimes you have to manage your expectations. And it's unrealistic to expect a system to automatically accommodate every conceivable concept out of the box without adapting it to a particular setting or circumstance (maybe the GM could hand out extra points specifically for non-combat/adventuring "background" abilities, for example). Or waiting till you have enough points to get every ability you want.

The "issue" here is ultimately that you want something that you can actually eventually have. But you want to have it right away. That's a much better issue to have than not being able to get it ever.

Sure it's an inherent problem of random generation, but it's not an inherent problem of random generation failing at its stated goal. Point buy, on the other hand fails at its stated goal. You move from random generation to point buy being expected to create the exact character you envisioned, and you still can't.

Quote from: migo on December 18, 2023, 11:10:42 AM

Quote from: VisionStorm on December 17, 2023, 10:34:20 PM

Quote from: migo on December 17, 2023, 06:34:55 PM

Quote from: VisionStorm on December 17, 2023, 05:59:22 PM*snip*

With random generation, you're not expecting to be able to make a character according to your wishes. So the second point is automatically a non-issue. The first point, yeah, you can have character imbalance with random generation, but there you only have one issue to solve, not two.

Even with random generation, you might still want to build a character a certain way, it's just that the system doesn't allow you to, or limits your options. So it can still be an issue (which I've personally had, or dealt with player who had it), it's just you can do nothing about it. Ever.

But with point buy you at least have more control over your character, even if you can't get 100% what you want out of the gate. But you might still get it eventually. And the GM might even make adjustments or concessions to get it right away.

Sometimes you have to manage your expectations. And it's unrealistic to expect a system to automatically accommodate every conceivable concept out of the box without adapting it to a particular setting or circumstance (maybe the GM could hand out extra points specifically for non-combat/adventuring "background" abilities, for example). Or waiting till you have enough points to get every ability you want.

The "issue" here is ultimately that you want something that you can actually eventually have. But you want to have it right away. That's a much better issue to have than not being able to get it ever.

Sure it's an inherent problem of random generation, but it's not an inherent problem of random generation failing at its stated goal. Point buy, on the other hand fails at its stated goal. You move from random generation to point buy being expected to create the exact character you envisioned, and you still can't.

This kind of problem stems not from a point buy system but rather from player desires not aligning to the power level of the campaign. If a player is told to create a 150 point character, but the super duper character they have envisioned requires 275 points to build then the player has to adjust the desired concept to something 150 points will buy. Some players always want more than the campaign starting level gives their characters regardless of creation method. In point buy they never have enough points. In random generation the rolled stats are not high enough for the character that they envision. NO creation method will help with that.

Quote from: Kyle Aaron on December 17, 2023, 10:03:59 PM

QuoteThere's almost no reason for that to have been in the book, the one merit is that it prevents the strongest woman from being as strong as the strongest man, a piece of realism that all later versions have left out.

Ah, you're one of those guys. Realism in a game about... dungeons and... dragons. Rightyo. Start shaving your neck, mate, and please discard the fedora.

No one is giving you a social credit point here for this, so just don't bother. Women are physically weaker than men, and the last version of D&D that recognized this was AD&D 1e. Me pointing this out in passing isn't actually a prompt for you to do a big feminism.

Quote
In my experience, any male bringing up this "issue" (and it's always a male, though rarely a very masculine one) has nothing else intelligent to contribute to the conversation. And further perusal of your comments confirms this is so, that you are speaking more from reading AD&D1e than playing it.

In my experience anyone mad enough about a true point brought up factually is done contributing usefully to any conversation, especially once they start calling people fedoras and other made up internet quasislurs.

Anyway, back on track- your point about 18/00 being a vanishingly small percent of things if rolled naturally is not as good of a point as it should be. My actual experience involved a lot of tables that mysteriously had one of these supposedly vanishingly-rare people. Did they roll it? Get it as a reward?  I don't know, but there were a lot of character sheets from people in games that weren't mine with 18/51 or up, something that continued well through second edition. 
Also note that there were no equivalent rules for transcendent-but-not-divine wisdom, intelligence, or dexterity.  These special strength scores were there to be played, and played they were, in numbers far more than any of the generative abilities in 1e or 2e ever implied.

Quote

QuoteFighters get multiple attacks, and that +6 to damage applies to all of them

All fighters get level attacks per round against monsters of fewer than 1 hit die.

This is not actually what I'm talking about- fighters get more attacks per round as they level up, and I showed a fighter of reasonable level.  I didn't bring up the multiplicative effect to imply AT ALL that this is when it BECOMES distortive.  It's OBVIOUSLY distortive to add damage to attacks when enemies have less than 10 hit points.  My point is, that as fighters go up to 2/round, these extra attacks apply.  Also there are other ways to get extra attacks besides just leveling, and all of them stack multiplicatively.

AD&D 1e's huge strength bonus totally changes all combat at every level when it's available, compared to when it is not.  The triple damage I pointed out is true even at a level where the attack roll bonus from fighter levels is substantial. That is the point of the math- to show that that +3 to hit is a big deal, and the +6 to strength is a huge deal, and a bonus-less fighter compared to the full one is an incredibly huge difference, at every single level, even at high level when the extra chance to hit might at a glance seem to outweigh the strength bonus.  It never does.

Quote
But let's assume for a moment your concerns about high Strength are valid. If that is so, then it is better to have random roll than point-buy - because the problem will be far less common. And random roll vs point-buy is after all the point of the thread,

You're quite confused.  Under point buy, stats have caps you can't buy above.  I don't know where you'd want to put that cap, but it's below 18/00.

Quotethough apparently you have other more important concerns, like the idea of fictional wimminz in a fictional game world with fictional physics being stronger than your puny self.

Seethe about bell curves more.  Feminism delenda est.

Quote

You're quite confused.  Under point buy, stats have caps you can't buy above.  I don't know where you'd want to put that cap, but it's below 18/00.

Point buy probably is the best solution to the mess that is percentile strength. With one point being 1 step. So you basically take Dark Sun Revised and map it to the percentile table.

There are some systems that let you roll percentile strength if you get an 18, which is probably not a great idea, and others that make you pay 1 point for every 10%, which is also pretty terrible.

Quote from: migo on December 18, 2023, 12:47:29 PM

Quote

You're quite confused.  Under point buy, stats have caps you can't buy above.  I don't know where you'd want to put that cap, but it's below 18/00.

Point buy probably is the best solution to the mess that is percentile strength. With one point being 1 step. So you basically take Dark Sun Revised and map it to the percentile table.

There are some systems that let you roll percentile strength if you get an 18, which is probably not a great idea, and others that make you pay 1 point for every 10%, which is also pretty terrible.

Regardless of how it is generated or purchased, exceptional STR ruins the math assumptions and severely weakens many combat encounters that should prove more difficult for characters of a given level.

Quote from: Old Aegidius on December 18, 2023, 08:56:46 AM

Quote from: jhkim on December 18, 2023, 02:57:33 AM
because the GM and the players reined in power-gamers by saying "no" rather than allowing rules hacking.

I don't mean to pick on you in particular, but there needs to be a name for this general form of argument - that problems can be resolved by the GM or players mitigating a problem via the social contract as opposed to the rules. Even in a dysfunctional group, a sufficiently authoritative GM could simply quash any debate and resolve an issue by fiat to keep play moving. My perspective on that argument is that it doesn't really go anywhere. Pointing out that there are intentional gaps or unforeseen situations in the rules or in the fiction which need GM arbitration is fine. Suggesting that the group can figure out a solution to a problem the game itself introduced is not an acceptable answer for a ruleset, in my view. See PVP Combat rules and the related controversy regarding Candela Obscura.

I don't disagree on the general principle that GM overruling doesn't paper over all rules, but also, games should be designed to be run by an active and intelligent GM. I think we're disagreeing on what are intentional gaps. (BTW, I don't know about the Candela Obscura controversy.)

The question is if closing the gap would make the game more fun for the players and GMs. If a new GURPS version were to be released that had random-roll and/or vastly restricted options to eliminate min-maxing, I don't think it would be hailed as an improvement by GURPS players and GMs. Being able to custom design a wide range of characters is one of the appeals of GURPS. It is a popular feature of the game.

Quote from: Old Aegidius on December 18, 2023, 08:56:46 AM
Modern D&D introduces the attribute problem by making attributes central to a character's competency in everything they do, while limiting the situations where min-maxing is a real liability. It then blows up any established consensus and invites controversy by providing a dozen different attribute generation methods, each with wildly different characteristics. If these at least orbited around some kind of core assumptions that the rest of the design was built atop, we could reason about what identifies a sane generation method.

I don't get this. AD&D1 has five different methods of generating attributes. AD&D2 has six methods. D&D3 has one method (3/4d6 and arrange), while 4E and 5E each have three (4d6, standard array, or point-buy).

At least from 3E onwards, they all have in common that the player can arrange the scores as they like once generated. From experience, I don't find the generation method makes much difference. If a 5E group were to only use 4d6 and disallow standard array, I don't think it would make much difference to how the game plays out.

Quote from: Exploderwizard on December 18, 2023, 12:59:51 PM
Regardless of how it is generated or purchased, exceptional STR ruins the math assumptions and severely weakens many combat encounters that should prove more difficult for characters of a given level.

This exactly. If everyone did the suggested roll method the number of games disrupted with 18/51 and up would be small enough and legitimately interesting to all involved, just because of the sheer novelty of him vastly overperforming. But that's not what happens when it's presented on the same table as the realistic scores. It was always much more common than stats would dictate, and if you had a point buy that made it, or a rolling system that let you cheese it, it would dominate many encounters and twist the game experience away from intended quite a bit. A point buy with sensible stat caps would, of course, prevent it. Honest rolling would minimize it. But tables that allegedly rolled stats would have these way more often than you'd think.

Quote from: jhkim on December 18, 2023, 01:34:26 PM
If a 5E group were to only use 4d6 and disallow standard array, I don't think it would make much difference to how the game plays out.

I think I disagree here- or at least, I have a contrary point to make.
In previous versions, stats were pretty much a bunch of statistical things.  A really high int might give you an unlimited spellbook, but overall you were getting higher chances of learning your spells, or a better chance of surviving a system shock roll.  3.X brought us to the "one modifier for every two stats", which is generally a pretty big deal.  If you look at a game where you could from +0 to +3, that's a pretty big change, but here in 5e that rolling method goes from +0 to +5.  In practice, your 5e character is assumed to have a +3 (16 or 17) in their mainstat at the start of the game, and the actual upgrade range goes from +3 to +5.
But 5e went beyond just this- it also attached feats, an optional rule that almost all tables allow.  If you roll up something really solid at the start, you may be able to just take feats at levels 4, 8, and 12 (and sometimes more depending on your class), instead of having to up your stats.  The point buy is supposed to have you making real choices between "do you take the feat that does cool stuff, or do you go from +3 to +4" (there's a powergamer option in one of the books that can let you start with an 18 and half a feat, but it's optional and everyone knows it is the purest cheese).

With rolled stats in 5e, a bad character won't just get a -1 or whatever, they'll probably never get the feat they want.  The opposite end isn't just a +1, it's early access to strong feats with no downside.  Basically, the 5e stats have pointy growths that punish rolling tables in 5e more so than 3.5, which in turn punishes rolling tables more than AD&D.

Quote from: Chris24601 on December 18, 2023, 10:49:26 AM
When all Strength does is melee attacks and lifting things with just the Athletics skill associated with it, and then the system also allows you to use Dex for melee attacks for just a net -1 to damage, and also applies to Ranged Attacks, Armor Class, Acrobatics and Thievery and Stealth, and avoiding some of the most common spells... and the lifting things is almost never important because they set the Encumbrance values so that even average Strength can carry 150 pounds of stuff without being slowed down because most players can't be bothered to track equipment...

Why the heck would you ever take Strength in such a system? (yes, this is exactly how 5e handles it which is just one of the reasons 5e sucks).

I'm not defending 5e for something that they are actually guilty of, but I will point out that 5e's design intent appears to be that the highest damage weapons are all melee weapons that can only use strength.  There's no two-handed or versatile finesse weapons, and a weapon of some random type- say, a +3 warhammer will never be able to be wielded with dexterity. But if you want a greatsword with greatweapon master, or a glaive with greatweapon master and polearm master, there's no way to do that except strength.

The issue is that the edge granted by these things is pretty vanishing. If no one in the party can use strength weapons effectively and weapons are rolled randomly, then the party will miss out on some loot, but there's no guarantee that weapons will be random so it's not a universal critique, and dealing a couple more points of damage per round than a ranged build is hardly worth all the restrictions melee comes with.  Also note that once you have your ranged equivalents of those feats, the only thing you miss out on compared to a melee guy is really opportunity attacks- you can shoot people point blank all day long once you have crossbow expert, whether you are using a hand crossbow, a long bow, or eldritch blast.

So it's a totally valid complaint- but that's because of the implementation, not the design.

Quote from: jhkim on December 18, 2023, 01:34:26 PM
I think we're disagreeing on what are intentional gaps. (BTW, I don't know about the Candela Obscura controversy.)

The question is if closing the gap would make the game more fun for the players and GMs. If a new GURPS version were to be released that had random-roll and/or vastly restricted options to eliminate min-maxing, I don't think it would be hailed as an improvement by GURPS players and GMs. Being able to custom design a wide range of characters is one of the appeals of GURPS. It is a popular feature of the game.

I agree regarding GURPS and that we disagree over what a reasonable gap is. GURPS has a design ethos that more clearly articulates what a sane approach is and what is dissonant. Random gen and point buy are not quite diametrically opposed, but they're close.

Some Candela Obscura reviews have been pointing at gaps in the ruleset that are big problems, notably PVP Combat. Candela Obscura knows it doesn't have a solution to the problem, but it just argues the GM should decide how best to handle it. To me, that's an example of an unreasonable and unacceptable gap.

In D&D character generation, random rolling and standard array are both options that make sense in similar contexts (if the standard array approximates the average of the random method). Point buy is the aberration and it enables the min-maxing behavior. The solution to that IMO cannot be "the GM/group should prevent that" because the problem was introduced by this method in the first place and there isn't a clear guiding principle to establish consensus on what's min-maxing with point buy and what's not.

Quote from: jhkim on December 18, 2023, 01:34:26 PM
I don't get this. AD&D1 has five different methods of generating attributes. AD&D2 has six methods. D&D3 has one method (3/4d6 and arrange), while 4E and 5E each have three (4d6, standard array, or point-buy).

At least from 3E onwards, they all have in common that the player can arrange the scores as they like once generated. From experience, I don't find the generation method makes much difference. If a 5E group were to only use 4d6 and disallow standard array, I don't think it would make much difference to how the game plays out.

All 6 methods in AD&D 2e are random generation, to my recollection. There is guidance in the text about when/why each method could be used that helps guide consensus. I don't recall AD&D 1e's methods but I'd guess they're similar variations on a theme. 3e is packe with vestigial stuff that made sense as part of a broader design. Point-buy in 4e/5e is the weirdness that definitely played into the direction modern D&D was heading, but to have these options side-by-side is unreasonable because of the design of the game.

In 3e, a Fighter with 18 STR hits as accurately as a 5th-level fighter with average stats and hits roughly as hard as if they had an extra d6 damage (on average). A 3e character with 18 DEX and a finesse or ranged weapon not only hits as accurately as the 18 STR fighter, but gets +4 AC (the difference between light chain shirt and full plate), +4 to initiative (the equivalent value of a Feat), and gets +4 to their Reflex saving throw (Much more than a 4-level jump). By contrast, the difference between a Fighter with 10 STR and 17 STR in AD&D 2e is +1 to hit and damage, so the character fights about as well as a 2nd level character. I dislike percentile STR but from 17 up through 18 or 18/50, you're only getting boosts to damage.

Assuming your campaign makes it to 12th level (which is quite a feat in AD&D 2e and where the game starts totally breaking down in 3e), the 3e character with an 18 is at least 25% better than the baseline (in some instances, much better). The AD&D character with 18 STR by contrast (assuming 50 or less percentile strength) is about 5-10% better than baseline (equivalent Thac0 of 8 vs. 9). And it doesn't really affect saving throws (Wisdom etc. some exceptions).

So in AD&D, the stats are mostly important during character generation to qualify for classes (to establish immersion and worldbuilding expectations) and they play a more minor role during play until they get into really high values (usually due to magic items). In that context, random generation makes sense. Introducing point buy makes some sense to me given modern D&D's apparent goals and how much of a difference an 18 stat makes vs. a 10, but it introduces the min-maxing problem and I don't find "the GM/group should prevent that" to be a satisfying solution for this new problem.

Quote from: Chris24601 on December 18, 2023, 10:01:06 AM

Quote from: Old Aegidius on December 18, 2023, 08:56:46 AM
Modern D&D introduces the attribute problem by making attributes central to a character's competency in everything they do, while limiting the situations where min-maxing is a real liability. It then blows up any established consensus and invites controversy by providing a dozen different attribute generation methods, each with wildly different characteristics. If these at least orbited around some kind of core assumptions that the rest of the design was built atop, we could reason about what identifies a sane generation method.

As time has gone on I've grown rather negative towards attributes that only serve as base modifiers for the actual checks and think a lot of issues could be solved if you just replaced them with more "skill" points and higher level 1 caps for the skills.

Want a strong fighter? Put your skills into Fitness and Melee. Want a smart wizard? Put your skills into Arcana and Lore. Etc.

If you must have Racial attribute-like adjustments, just make them at the skill level; Dwarves get bonuses to Fitness and Engineering, Halflings to Acrobatics and Stealth, Elves to Archery and Lore. etc.

Kind of like how the White Wolf Games somewhat handled it, where you picked a primary, secondary, and third stat, and at the end of character gen got a couple freebie points to spend (though the ratios for freebies were pretty steep)

What do you think of negative XP? a player buys and advantage or a race at level 1, but it has to buy it off with XP before he can progress in level?

Quote from: Socratic-DM on December 18, 2023, 04:16:23 PM

Quote from: Chris24601 on December 18, 2023, 10:01:06 AM

Quote from: Old Aegidius on December 18, 2023, 08:56:46 AM
Modern D&D introduces the attribute problem by making attributes central to a character's competency in everything they do, while limiting the situations where min-maxing is a real liability. It then blows up any established consensus and invites controversy by providing a dozen different attribute generation methods, each with wildly different characteristics. If these at least orbited around some kind of core assumptions that the rest of the design was built atop, we could reason about what identifies a sane generation method.

As time has gone on I've grown rather negative towards attributes that only serve as base modifiers for the actual checks and think a lot of issues could be solved if you just replaced them with more "skill" points and higher level 1 caps for the skills.

Want a strong fighter? Put your skills into Fitness and Melee. Want a smart wizard? Put your skills into Arcana and Lore. Etc.

If you must have Racial attribute-like adjustments, just make them at the skill level; Dwarves get bonuses to Fitness and Engineering, Halflings to Acrobatics and Stealth, Elves to Archery and Lore. etc.

Kind of like how the White Wolf Games somewhat handled it, where you picked a primary, secondary, and third stat, and at the end of character gen got a couple freebie points to spend (though the ratios for freebies were pretty steep)

What do you think of negative XP? a player buys and advantage or a race at level 1, but it has to buy it off with XP before he can progress in level?

Not a fan, mainly because it rests on an assumption that specifically tracked XP will even be used for a given campaign (vs. say, Milestone leveling) and that the benefits will be universally advantageous over the entire span of the campaign to be X far behind everyone else (this was one of the problems with ECL Races in 3e).

Far better is to just have a baseline (or an adjustable baseline) that accommodates said advantages/races... i.e. if the GM wants a campaign where Drow are common PCs, then the default human PC (not necessarily all humans, but PC ones) gets roughly equivalent features. If you want weaker starting PCs then put Drow into the NPC only category (or make the default Drow in the setting weaker).

Alternately, just accept the Palladium mantra that fun in RPGs can be completely independent of power level and Vagabond and Glitterboy Pilot can both have adventures together (because the GM will do the work of giving both of them interesting things to do; though probably not at the same time).

Quote from: Chris24601 on December 18, 2023, 05:14:47 PM
Alternately, just accept the Palladium mantra that fun in RPGs can be completely independent of power level and Vagabond and Glitterboy Pilot can both have adventures together (because the GM will do the work of giving both of them interesting things to do; though probably not at the same time).

Just because this is true doesn't mean that some kind of pursuit of balance is fruitless, and putting everyone on the same tactical battle grid and using well defined combat minigame rules is something that a lot of players and DMs really enjoy- though it is definitely not some requirement for ideal gaming or whatever.  It's definitely something I prefer though, so if a system makes some effort in that direction, it's doing work I'd otherwise have to do, and I tend to appreciate that.

Quote from: Old Aegidius on December 18, 2023, 03:48:54 PM
Assuming your campaign makes it to 12th level (which is quite a feat in AD&D 2e and where the game starts totally breaking down in 3e), the 3e character with an 18 is at least 25% better than the baseline (in some instances, much better). The AD&D character with 18 STR by contrast (assuming 50 or less percentile strength) is about 5-10% better than baseline (equivalent Thac0 of 8 vs. 9). And it doesn't really affect saving throws (Wisdom etc. some exceptions).

Comparing to Strength 10 isn't relevant, though. Basically zero PC fighters will have a 10 Strength.

A more relevant question is: compare a 5th percentile fighter vs a 95th percentile fighter. i.e. If twenty characters are rolled, what would be the worst out of those and the best out of those. I'm assuming that the method is best 3 out of 4d6 and then arrange, and that the fighter puts his highest stat in Strength. The 5th percentile fighter has 13.8 Strength, which I'll consider 14. The 95th percentile fighter has 18 Strength, or in AD&D, has 18/43.

In AD&D, the 5th percentile fighter is +0/+0 while the 95th is +1/+3. In D&D3, the 5th percentile fighter is +2/+2 while the 95th is +4/+4.

So in AD&D, the lucky-rolling player gets a bigger boost in damage compared to D&D3, but one less boost in to-hit. That's because in the smoother progression of D&D3, even unlucky rollers still get some bonus.

Also, AD&D stats might not be used for saving throws as much, but they have many other uses, like the 10% experience bonus - or qualifying for advanced classes like the Ranger.

Quote from: Old Aegidius on December 18, 2023, 03:48:54 PM
So in AD&D, the stats are mostly important during character generation to qualify for classes (to establish immersion and worldbuilding expectations) and they play a more minor role during play until they get into really high values (usually due to magic items). In that context, random generation makes sense. Introducing point buy makes some sense to me given modern D&D's apparent goals and how much of a difference an 18 stat makes vs. a 10, but it introduces the min-maxing problem and I don't find "the GM/group should prevent that" to be a satisfying solution for this new problem.

Having played with point-buy in 5E for many years, I don't recall ever needing to say "no" to any attribute buy as GM. Under point-buy, an 18 starting stat is very expensive and many players don't go for it, because they want to have more total bonus in other stats.

Quote from: jhkim on December 18, 2023, 05:26:17 PM
Having played with point-buy in 5E for many years, I don't recall ever needing to say "no" to any attribute buy as GM. Under point-buy, an 18 starting stat is very expensive and many players don't go for it, because they want to have more total bonus in other stats.

5ed point buy goes up to 15.  Using PHB races, everyone can bump that to 16 or 17, and nothing goes to 18 ever.  There's one wacky splatbook thing, custom lineage, where you start with a 15, get to add +2, and then for some reason get a feat, which could be say, slasher or heavily armored, half feats that add one more stat.  This is how you start with an 18 in D&D 5e, it's a cheeseball powerpick, and if it's allowed it is often your best pick and recommended in all the build guides (because yes there are build guides, and yes 1d3-1 players at your table have read them for some reason).

And the reason you go for this 18 is because by bounded accuracy standards you aren't expected to have that 18 until level 5 or so, and you don't start missing out on assumed accuracy until around level 12- meaning you can take your two favorite feats at levels 4 and 8, and for a lot of classes and combos (probably all of them these days, given how many feats exist), and be playing the character you wanted with all your mechanics online much earlier than they would be.

Quote from: jhkim on December 18, 2023, 05:26:17 PM
Comparing to Strength 10 isn't relevant, though. Basically zero PC fighters will have a 10 Strength.

Nowadays, and with most game groups, yes. But it needn't be so. I rolled up 4d6 drop lowest, and ended up with Strength 10 (or maybe 12, I can't remember - it wasn't enough to get a to-hit or damage bonus, anyway) but Charisma 17. The DM said I could swap it around, I said no. "He will be Fabio, the Most Beautiful Fighter in the Cosmos." Fabio hired men-at-arms, and between his generous pay and Charisma, they were insanely loyal. They made a very effective first rank going through the dungeon.

That's the statistics of AD&D1e - 3-5 fighters of 0-1st level doing their 3-5 attacks each round will on average do more damage than 1 fighter doing his 1 attack, even if he has exceptional Strength. That's why everyone dreads pissy little monsters like rats, and why MUs can be horribly effective with darts - lots of attacks!

Fabio perished at 7th level after meeting the gaze of a medusa. The other players were intent on bringing him out and recovering him somehow. I said, "No - that Fabio is immortalised in stone, that is the way. A later party of adventurers will find him and admire his beauty."

I'd never have had that story to tell with point-buy.

What I've found over the years is that high attributes can actually be dangerous for a character in combat - it makes the player over-confident with their character, they charge in. After all, thinking of 1st level Fighters, one with 10 hit points who gets hit twice has the same chances of being knocked down as a 5 hit point one who gets hit once. If you have 5 HP and leather armour you're more likely to just stand in the doorway and wait for them to come at you one-by-one, compared to having 10 HP and banded mail.

And so in practice, higher attributes are dangerous. Game design has to take into account human nature, thus for example understanding that point-buy will take longer than random roll, given the same level of complexity in the game system otherwise, which means players will be more pissed off if their character dies, so point-buy systems tend to encourage hero point or other systems reducing lethality, etc. Human nature.