Oldschool D&D attack bonus
First, a little context. I've been playing Dungeons & Dragons since the Red Box—being one of that generation who grew up watching Venger chase a bunch of isekaied teenagers across "the realm of Dungeons & Dragons"—but I never really got to play much (or properly!) when I was a kid. I got all the BECMI sets, spent my paper route money on loads of AD&D hardcovers, and enthusiastically made the transition to 2nd Edition and its profusion of cool settings. Still, I did way more reading and pointless character-making than playing. I think I didn't really get D&D's rules back then, and I got way into other RPGs that offered more interesting options and ran on systems that seemed to make more sense.
Then, in the early 21st century—sometime around the release of v.3.5—I got dragged back to D&D for a forum-based game. I don't think I was thrilled to learn that the Magic: The Gathering people had bought TSR, but I had to admit that WotC had created a version of D&D that actually seemed pretty cool—flexible, logical, comprehensible, and fun!
Anyway, I stuck with WotC-era D&D, my enthusiasm occasionally waning, but then being won back when the company would release another new edition or take things in a new direction. I thought 4e was a whole lot of fun, and that the vast majority of complaints against it were born of misunderstandings, bad marketing, and basic resistance to change. And 5e almost managed to deliver on its ludicrous promise of being the D&D for all of the game's disparate communities, while also bringing in loads of new people!
…But at this point I'm kinda burnt out on 5e. A system that seemed really fast and lightweight during the playtests is starting to feel clunky and fiddly. And the 2024 update—clearly v.5.5 in all but name—feels like progress in the wrong direction, embracing a combat-centric gameyness that feels like 4e without the tactical depth.
Meanwhile, I've been enjoying the OSR/post-OSR scene's work for years! They haven't managed to lure me back to TSR-era D&D—I still have trouble seeing the virtues in a lot of those mechanics—but the blogs and new games those folks have produced are frequently loads of fun.
The actual point of this post
So, I recently read through all of Chris P Wolf's very cool "Thursdays in Thracia" posts. Chris' thoughts on B/X D&D lead me to look up Old-School Essentials and contemplate those rules a little. Also, Chris mentioned using ascending armor class and attack bonuses in B/X, which sounded pretty rad. So I decided to go back and look at the attack mechanics of some 20th-century D&D editions and see how easy they'd be to convert into that modern paradigm, and what that would look like. Mostly because I knew that it would involve spreadsheets, and that kind of thing is somehow fun for me.
OD&D
I actually tried looking at Chainmail first, but (perhaps unsurprisingly) D&D's wargame predecessor didn't prove to be a great candidate for this exercise. So I dug up the hit roll chart from the 1976 "White Box" edition of Dungeons & Dragons. It looked something like this.
There are some immediately interesting things here.
- It seems kind of crazy that the table only really shows the numbers for Fighting Men, and just offers a footnote on how the reader can figure out Magic-Users and Clerics for themself! (This is, of course, the pre-Thief era.)
- "Normal men" is such a weird term for "people without class levels". I'd kinda forgotten how that was used.
- The armor-to-armor-class connection here is really straightforward. Each tier of armor is worth another 2 points of AC, with shields being worth 1 point and thus filling out the rest of the AC scale. Satisfying!
- I can't see why AC is descending, or why it should start at 9, given that it's effectively disconnected from the actual rolls required to hit. I feel like I heard that it goes back to some other company's naval combat game, but I ain't going down that rabbithole today.
Anyway, this actually looks like it would convert to an attack-bonus-vs.-ascending-AC mechanic really easily. So let's try it.
| level | Cleric | Fighting Man | Magic-User |
|---|---|---|---|
| 1 | +0 | +0 | +0 |
| 2 | +0 | +0 | +0 |
| 3 | +0 | +0 | +0 |
| 4 | +0 | +2 | +0 |
| 5 | +2 | +2 | +0 |
| 6 | +2 | +2 | +2 |
| 7 | +2 | +5 | +2 |
| 8 | +2 | +5 | +2 |
| 9 | +5 | +5 | +2 |
| 10 | +5 | +7 | +2 |
| 11 | +5 | +7 | +5 |
| 12 | +5 | +7 | +5 |
| 13 | +7 | +9 | +5 |
| 14 | +7 | +9 | +5 |
| 15 | +7 | +9 | +5 |
| 16 | +7 | +12 | +7 |
| 17 | +9 | +12 | +7 |
| 18 | +9 | +12 | +7 |
| 19 | +9 | +12 | +7 |
| 20 | +9 | +12 | +7 |
| 21 | +12 | +12 | +9 |
| 22 | +12 | +12 | +9 |
| 23 | +12 | +12 | +9 |
| 24 | +12 | +12 | +9 |
| 25 | +12 | +12 | +9 |
| 26+ | +12 | +12 | +12 |
The logic here is that all three classes' attack bonuses advance first by 2 points, then by 3, then 2, then 2 again, and then finally by 3. Fighting Men advance through these "attack classes" every 3 levels, Clerics advance every 4 levels, and Magic-Users every 5. And everybody maxes out at +12 (what we might call "attack class 6"), at which they've got a 95% chance to hit a target in chainmail (with no shield) and a 100% chance to hit anybody in worse armor.
It's very easy to describe, but it seems odd that the progression isn't smoother, right? And why the 2-3-2-2-3 sequence?
So naturally I calculate what the overall bonus-progression-per-level looks like for each class, and I get something like this.
- Fighting Man: +0.8
- Cleric: +0.6
- Magic-User: +0.48
If you're familiar with 3e (or the d20 system in general), these numbers might jump out at you: They're reminiscent of the Base Attack Bonus progressions from that era, in which Fighters get +1 per level (or "full BAB"), Clerics essentially get +0.75 per level ("3/4 BAB"), and Wizards essentially get +0.5 per level ("1/5 BAB"). Admittedly, 0.6 isn't all that close to 0.75, but 3e is notorious for making Clerics overpowered as a way to bribe people to play a class perceived as vital but boring.
Anyway, I suspect the WotC team that created 3e did the same math that I just did, and that's pretty interesting to me.
AD&D 1e
The next major step in D&D's evolution is Advanced Dungeons & Dragons, which starts to come out in 1977. Strangely—I mean straight up fuckin wildly—the new edition's attack roll charts don't emerge until 1979, because they're all printed in the Dungeon Masters Guide. Maybe I'm missing something—I fully cop to not having a great handle on these books I've owned since I was a kid—but the PHB doesn't seem to actually talk about attack rolls. The Character Classes chapter does have charts for hit point progression, spell slot progression, title progression, etc.—there's even a huge one for the Assassin's fees!—but no attack progression! I can only conclude that the results of attack rolls were considered the DM's business, since you have to know the target's AC to adjudicate them. Then again, the DMG also contains the saving throw charts, and I don't see an equivalent rationale for that.
Anyway, AD&D 1e features four separate "attack matrices", basically covering a full page.
And wow, you can really see how much the game has changed in this edition! We've gone from three classes to eleven, there are four progression schedules instead of three, we've replaced "normal men" with the concept of "0 level" characters, we're starting the AC scale at 10 instead of 9, and we're riding the descending scale all the way down to -10! This page is also a classic example of the 1e DMG's famously bad information presentation and weird editing.
One notable thing here is that the roguish types (Assassins and Thieves) get a slightly worse attack table than the priestly types (Clerics, Druids, and Monks). I'm not sure if the Thief was in that same position when it was introduced in the Greyhawk supplement.
Anyway, I can't proceed without bringing up the implied setting details present in this footnote.
*Half-elves use the attack matrix as elves do, while non-player character half-orcs use the attack matrix for monsters. Dwarves, elves and gnomes are never lower than 1st level (unlike halflings and humans, which may be of 0 level).
So "0 level" is only for humans and halflings, while dwarves, elves, half-elves, and gnomes always have class levels! And even though half-orcs can be PCs in this edition (and I will resist the urge to do a whole sidebar on how they're presented in the 1e PHB, but wow, it ain't cool), non-player half-orcs are technically monsters.
The big news here, though, is that the AD&D attack mechanics cannot be cleanly translated into an attack bonus the way the OD&D ones can! Check out those columns: We've got a very normal scale at the lower ends, with an AC improvement of 1 predictably meaning that the attacker needs to roll 1 higher to hit. But then, up in the negative zone of heavily armored targets, we've got all these 20s! Each column goes through a stack of five 20s before returning to 1-to-1 scaling. It's weird stuff.
B/X
Here's where we come to the system that started this whole distraction of mine, the 1981 Moldvay/Cook continuation of the non-Advanced line. I get the impression that B/X (generally in its Old-School Essentials or Labyrinth Lord incarnations, or of course some kind of homebrew) is much beloved by lots of modern players, including some folks who started playing TTRPGs in this millennium. I'm kind of interested in it.
Anyway, here's what the attack table in the Expert Rulebook looks like.
Okay! So it's immediately clear that B/X evolved from OD&D, not AD&D. Even though the axes have been transposed, the AC scale goes into the negatives, and we've now got Thieves and demi-humans included, this looks a lot like the White Box chart. The "normal man" terminology is back, although this time it's clear such regular folks are worse in combat than novice adventurers, just like the 0-level NPCs in AD&D. Thieves have the same attack progression as Clerics, again prefiguring 3e (in which both Clerics and Rogues are 3/4 BAB classes). It's slightly surprising that the Halfling (which is, of course, a class in this edition) is on the Fighter's level instead of the Thief's. Oh, and I'm happy to see we've gone from "Fighting Man" to "Fighter", as in AD&D.
This should be easy to turn into an attack bonus table.
| level | Cleric, Thief | Dwarf, Elf, Fighter, Halfling | Magic-User | ||||
|---|---|---|---|---|---|---|---|
| 1 | +0 | +0 | +0 | ||||
| 2 | +0 | +0 | +0 | ||||
| 3 | +0 | +0 | +0 | ||||
| 4 | +0 | +2 | +0 | ||||
| 5 | +2 | +2 | +0 | ||||
| 6 | +2 | +2 | +2 | ||||
| 7 | +2 | +5 | +2 | ||||
| 8 | +2 | +5 | +2 | ||||
| 9 | +5 | +5 | +2 | ||||
| 10 | +5 | +7 | +2 | ||||
| 11 | +5 | +7 | +5 | ||||
| 12 | +5 | +7 | +5 | ||||
| 13 | +7 | +9 | +5 | ||||
| 14 | +7 | +9 | +5 | ||||
| 15 | +7 | +9 | +5 | ||||
| 16 | +7 | +9 | +7 | ||||
| 17 | +9 | +9 | +7 | ||||
| 18 | +9 | +9 | +7 | ||||
| 19 | +9 | +9 | +7 | ||||
| 20 | +9 | +9 | +7 | ||||
| 21+ | +9 | +9 | +9 |
So, it's almost exactly like the OD&D table, except that it includes more classes and tops out at +9 instead of +12. But there's some else to keep in mind: The B/X chart doesn't include target numbers higher than 20 or lower than 2, so I guess this might be where "20 always hits, 1 always misses" entered the canon. I will not bother to chase that question down today, though.
AD&D 2e
Several years later, in 1989, we hit one of D&D's big, controversial edition changes: the release of Advanced Dungeons & Dragons 2nd Edition. I remember feeling like this one didn't actually change all that much, but of course this is where we got the real precursor to 3e's Basic Attack Bonus: the "To Hit Armor Class 0" number! This is where we exchange looking up our attack results on a chart for doing some slightly awkward math! I think I considered it an improvement at the time. I think I still do.
Check this out.
I'm including two tables here, since they're so simple and concise. I love that 2e explicitly states the progression schedule so that I don't even have to do the math!
Other interesting stuff:
- Class groups! "Priest" includes Clerics and Druids, "Rogue" includes Bards and Thieves, etc.
- No mention of "normal men" or 0-level characters! There are probably 2e rules for 0-level characters somewhere, but not on this table. It's in the PHB, and 0-level characters are usually gonna be NPCs.
- Except for the Priest group (which literally gets a 2-point jump every three levels), we've finally got smooth attack progressions!
- 2e continues 1e's attack roll hierarchy, with Warriors being the best, then Priests, then Rogues, then Wizards. 3e would later return to B/X's three-schedule paradigm, placing Clerics and Rogues on the same BAB progression.
This should be the easiest to convert into an attack bonus progression so far: I'll just subtract every THAC0 value from 20.
| level | Priest | Rogue | Warrior | Wizard |
|---|---|---|---|---|
| 1 | +0 | +0 | +0 | +0 |
| 2 | +0 | +0 | +1 | +0 |
| 3 | +0 | +1 | +2 | +0 |
| 4 | +2 | +1 | +3 | +1 |
| 5 | +2 | +2 | +4 | +1 |
| 6 | +2 | +2 | +5 | +1 |
| 7 | +4 | +3 | +6 | +2 |
| 8 | +4 | +3 | +7 | +2 |
| 9 | +4 | +4 | +8 | +2 |
| 10 | +6 | +4 | +9 | +3 |
| 11 | +6 | +5 | +10 | +3 |
| 12 | +6 | +5 | +11 | +3 |
| 13 | +8 | +6 | +12 | +4 |
| 14 | +8 | +6 | +13 | +4 |
| 15 | +8 | +7 | +14 | +4 |
| 16 | +10 | +7 | +15 | +5 |
| 17 | +10 | +8 | +16 | +5 |
| 18 | +10 | +8 | +17 | +5 |
| 19 | +12 | +9 | +18 | +6 |
| 20 | +12 | +9 | +19 | +6 |
And yeah, this almost looks like a bunch of 3e BAB progressions!
Rules Cyclopedia
Okay, one more. Let's just back over to the non-Advanced line to see how things evolved there. So, the B/X sets were followed by Mentzer's series of five big, beautiful, toy-store-friendly boxed sets, and then in 1991 that stuff was all compiled and revised by Allston into the Rules Cyclopedia. This was after my time in the realm of Dungeons & Dragons, so I haven't got any contemporary impressions of it.
So, let's see what the compiled rules of the whole BECMI edition have for an attack roll table.
I . . . feel like Allston maybe could have done more revision. This one's kind of a wild ride, even when compared to AD&D 1e.
I see that the Thieves of the non-Advanced line continue to share the Cleric/Druid attack progression, unlike those in AD&D. But we've got a really weird thing going on with the demi-human classes, where they progress as Fighters at first, and then switch to a letter-designated "demihuman attack rank". Unfortunately, we've also got those strings of 20s in the negative zone, followed by sequential progression (just like AD&D 1e), which means that this can't really be converted into an attack bonus table.
Of course, the craziest thing here is that this is a descending armor class scale that starts at 9, but the ACs go up to 19. That seems less likely to be the result of catastrophically bad Dexterity than high-level magical fuckery. And, mirroring the aforementioned mass of 20s, we've also got a corresponding band of 2s on the lightly armored end of the chart, beyond which things get really funky: Really high-level characters do bonus damage to targets with really poor ACs! That's pretty interesting.
I kind of want to do some reading on these odd mechanics, but I've definitely gone overboard for this post.
In conclusion
I can see that ascending AC and attack bonuses fit just fine in B/X, if not in every oldschool edition. I think Chris' group's decision to switch to them was probably a real quality-of-life improvement that didn't impact game balance at all (to the extent that "balance" is even a thing in B/X). I'd probably do the same thing if I ran some kind of old-timey D&D. I ain't planning to do that any time soon, though.
Anyway, it was cool to see the logic and patterns underlying these old tables. I might take a similar look at saving throws, some time. Those always struck me as weird and arbitrary, but maybe there are some hidden formulae there, too.