Game design has "moved on"

[Umbran]

3rd, d20 universal is simply a floating 20 number variable result span that modifiers actually shift up and down the natural number line. A line where an arbitrary DC was set, which forced the DM to become a player. This design actually removes any hope of success for lower level PCs and failure for higher ones. It needed "always succeeds or fails" funky results to artificially keep characters playable, but only served to slow down the game. Not to mention the d20 was no longer just used for game elements benefiting from large variables like attacks and saves, but "skills" and plenty of other rolls which should never have used a 20 point spread. Initiative anyone? Boy, did that get tedious quick.

I don't know if it was intentional, but that comes off with enough spin on that that we can hear it whir!

If we are comparing apples-to-apples: for to-hit, all the systems set a DC, via an armor class. It was not generally arbitrarily assigned by the GM.

If we are comparing apples-to-apples: for skills, the DC for a 3e sill check was arbitrary, but guided - the rules do give flat numbers for many tasks. But in prior editions, with less-well-developed skill systems, the entire *mechanic* for resolution of a tasks was arbitrarily assigned, making the GM no less a player.

And... as if a 1 in 20 chance of success keeps the character playable? I don't buy it, sorry.

 

[billd91]

And... as if a 1 in 20 chance of success keeps the character playable? I don't buy it, sorry.

Considering the auto-succeed rule and auto-fail rules on saves and attacks stems right from 1e, it's clear that they weren't injected just to enable any playability in 3e.

That said, the bounded ranges in 1e/2e, for saves in particular, was removed in 3e in favor of the open end. I consider the 1e/2e method generally better. That said, aside from that boundary, the attacking systems are so very close in effect that the BAB system is very much a user-friendlier version of the THAC0 system. And the THAC0 system is really just a minor variation on the 1e combat tables. These are not radically different.

 

[Janx]

Considering the auto-succeed rule and auto-fail rules on saves and attacks stems right from 1e, it's clear that they weren't injected just to enable any playability in 3e.

That said, the bounded ranges in 1e/2e, for saves in particular, was removed in 3e in favor of the open end. I consider the 1e/2e method generally better. That said, aside from that boundary, the attacking systems are so very close in effect that the BAB system is very much a user-friendlier version of the THAC0 system. And the THAC0 system is really just a minor variation on the 1e combat tables. These are not radically different.

It may also be that the more open-ended nature of 3e's attack roll resolution is an artifact of re-arranging where the 3 values go.

In 1e/2e the PC had an Attack Value (descending in nature). A fighter had 21 - level as his attack value. In 3e, to make big numbers mean better, it was simply Level.

In 1e/2e the PC had an AC (5 for chainmail if I recall). In 3e, as part of the inversion to make big numbers be good, it was 10 +5, so AC 15.

The last variable was the die roll.

So in 1e/2e it was AV - AC < d20 means success

In 3e, it was AV + d20 >= AC means success

Since 1e/2e started with 2 values having fixed maximums (THAC0 20 and AC 10), that kind of locked in things.

Once those variables were inverted, the upper bounds became unlimited, though the mathematical behavior remains intrinsically the same when using the basic game stats for armor and THAC0/BAB, the expression was clearer for most folks (a subject claim, I know).

So in my view, D&D did and did not change it's mechanical design for combat. the basic stats remain the same (BAB progression runs at the same rates per class as THAC0 progression did as did AC).

The inversion can be initially viewed as a minor tweak to make the math easier to absorb and make consistent (big is good). But it clearly had side effects like removing an upper bounds for really good ACs or really good Attack Values.

 

[howandwhy99]

Per my read and understanding of my 1e, 2e, 3e combat rules, the AC system remains equivalent per armor type (allowing that 3e the number goes up instead of down).

Honestly, I wasn't looking to raise another topic. Needless to say I really just disagree with you. For simplicity sake, perhaps we can at least agree on an easily discerned difference between all three? An average unarmed adult human actively defending in 1e has an AC of 9, in 2e an AC of 10, and in 3e an AC of 11 (inverted to 10).

 

[Janx]

Honestly, I wasn't looking to raise another topic. Needless to say I really just disagree with you. For simplicity sake, perhaps we can at least agree on an easily discerned difference between all three? An average unarmed adult human actively defending in 1e has an AC of 9, in 2e an AC of 10, and in 3e an AC of 11 (inverted to 10).

This sounds awfully quibbly. Per my understanding (and only played 1e for a very very short time as we were really starting 2e with a 1e PH) they all got AC10 (+0 bonus). Padded was +1, Leather was +2, Studded Leather was +3.

My assumption is that an unarmed human adult is expected to be actively defending in all cases barring an impairment. So in both 1e, 2e and 3e he has AC10. I could be wrong that there's no bonus, but I'd note then that both 1e and 3e would then give a +1 advantage, making them equal in treatment (1e AC9 = 3e AC11)

Heck, I remember in 2e there was an actual penalty for being unarmed (or bonus to the attacker, as really it don't matter that much what side of the equation you apply it).

So they all got the same ACs for what armor and Dex bonus they had in all editions. Situational modiers appear to be different (I call BS that Gary would give any +1 bonus to AC for an unarmed person defending themselves, but I don't know 1e well enough to swear that in court).

 

[Balesir]

Honestly, I wasn't looking to raise another topic. Needless to say I really just disagree with you. For simplicity sake, perhaps we can at least agree on an easily discerned difference between all three? An average unarmed adult human actively defending in 1e has an AC of 9, in 2e an AC of 10, and in 3e an AC of 11 (inverted to 10).

I think you are remembering OD&D, not 1e AD&D. With the coming of AD&D the AC of an unarmoured human went to 10. In OD&D it was 9.

 

[howandwhy99]

I'm confused a bit about what you wrote, but I'll try and work it out. It's early.

This sounds awfully quibbly. Per my understanding (and only played 1e for a very very short time as we were really starting 2e with a 1e PH) they all got AC10 (+0 bonus). Padded was +1, Leather was +2, Studded Leather was +3.

My assumption is that an unarmed human adult is expected to be actively defending in all cases barring an impairment. So in both 1e, 2e and 3e he has AC10. I could be wrong that there's no bonus, but I'd note then that both 1e and 3e would then give a +1 advantage, making them equal in treatment (1e AC9 = 3e AC11)

[MENTION=27160]Balesir[/MENTION] is right about OD&D and AC 9, I just use that system so much I forgot 1e made the change too.

AC 9 is an armor class, human wearing clothes (or not) in OD&D. It's not the armor, but the band of defense 9.999(repeating) to 9.0 the armor rests somewhere within. 2e increased the boundaries of the mechanic beyond 20, beyond the variance in the actual die roll, and used 1e's "humans as above are AC 10" as default. Still, even if that doesn't matter to you results 10 down to 1 span 50% of that d20 roll for a greater chance to hit than in OD&D. 3e's base AC 10 is actually converted to 11 in the old system. So that game begins with 55% odds of hitting on its d20 attack roll.

Heck, I remember in 2e there was an actual penalty for being unarmed (or bonus to the attacker, as really it don't matter that much what side of the equation you apply it).

Actually, much of D&D was built on the valuations of increasing modifiers to the target number or the roll inherent in its constant variation. So it very much mattered what side you applied adjustments to. But that's getting away from the simpler point: that even the base To-Hit number is different among these three.

So they all got the same ACs for what armor and Dex bonus they had in all editions. Situational modifiers appear to be different (I call BS that Gary would give any +1 bonus to AC for an unarmed person defending themselves, but I don't know 1e well enough to swear that in court).

A 2e human with a 1 point penalty for being unarmed would apply the point to the attacker's roll. Their AC doesn't change from 10, but odds have shifted in their opponent's favor. Those odds are not the same as if we worsened their AC to 11. Or what I believe you are saying is 9 so it can be just like OD&D, which would actually be an improvement. I'm not clear on what you were meaning here.

 

[Janx]

I'm confused a bit about what you wrote, but I'll try and work it out. It's early. [MENTION=27160]Balesir[/MENTION] is right about OD&D and AC 9, I just use that system so much I forgot 1e made the change too.

AC 9 is an armor class, human wearing clothes (or not) in OD&D. It's not the armor, but the band of defense 9.999(repeating) to 9.0 the armor rests somewhere within. 2e increased the boundaries of the mechanic beyond 20, beyond the variance in the actual die roll, and used 1e's "humans as above are AC 10" as default. Still, even if that doesn't matter to you results 10 down to 1 span 50% of that d20 roll for a greater chance to hit than in OD&D. 3e's base AC 10 is actually converted to 11 in the old system. So that game begins with 55% odds of hitting on its d20 attack roll.

Actually, much of D&D was built on the valuations of increasing modifiers to the target number or the roll inherent in its constant variation. So it very much mattered what side you applied adjustments to. But that's getting away from the simpler point: that even the base To-Hit number is different among these three.

A 2e human with a 1 point penalty for being unarmed would apply the point to the attacker's roll. Their AC doesn't change from 10, but odds have shifted in their opponent's favor. Those odds are not the same as if we worsened their AC to 11. Or what I believe you are saying is 9 so it can be just like OD&D, which would actually be an improvement. I'm not clear on what you were meaning here.

I am equally confused as to your confusion. A bonus to you is a penalty to me. It's algebra. You add on one side of the equation, subtract on the other.

So if I say there's a -2 penalty for being unarmed, regardless of whether there's a minus sign or not, I can figure out whether it needs to be added or subtracted to the appropriate component.

And the effect is the same mathematically, providing I do it right. So I can make your AC worse, or make the attacker's side better by 2.

 

[howandwhy99]

I am equally confused as to your confusion. A bonus to you is a penalty to me. It's algebra. You add on one side of the equation, subtract on the other.

So if I say there's a -2 penalty for being unarmed, regardless of whether there's a minus sign or not, I can figure out whether it needs to be added or subtracted to the appropriate component.

And the effect is the same mathematically, providing I do it right. So I can make your AC worse, or make the attacker's side better by 2.

I didn't think we would get into valuing modifiers within the variance of a d20 roll. That's more difficult to explain. I was hoping we could just agree on little stuff like base AC is different. Do you at least agree with the 45%, 50%, 55%?

For the rest, basically we are doing trig not algebra. The valuation of modifiers to a roll or to a target number are measured in relation to the starting odds, not percentage points. There's a big difference. I think the D&D Next designers know this is part of why they are using what they call "bounded accuracy".

I haven't looked at that design for awhile, but OD&D puts the whole infinite number of outcomes within the 20-point spread of the randomizing die. It is also bounded in its way, but doesn't shift it from there like D&D Next which has target numbers higher than 20.

Now 3e uses the infinite natural number line as its base and then moves the dice's 20-point set up and down it, as well as the target number. AC or attack mod, either could be infinite. But to be meaningful they needed to be within the 20 point variance of the die roll. Further benefits/penalties don't affect the odds. 1 always hits, 20 always misses. Unfortunately this design coupled with a large number of stacking modifiers allowed optimizers to hit that 5% plateau frequently.

Now OD&D has its faults too. Sometimes you could be change your odds, but not significantly enough to reach the next 5 percentage point plateau. The die was rolled the same with the same odds as if you didn't have that extra bit. But it wasn't a total drawback. You knew, maybe next time you find a way to build on your previous strategy and catch another face of that die to mean success for you.