... now with 35% more arrogance!

Wednesday, March 21, 2018

Thief Skills As Surprise: Stealth

Beginning a long time ago, I decided to treat all (or almost all) thief abilities as being based on surprise. Here's an elaboration on that for two skills: hide in shadows and move silently.

If an ordinary adventurer wants to hide, they must make that decision before a surprise roll is made. For example, if they know that a guard will be walking down a corridor at some point, they can hide around a corner. There must be some kind of cover. There's no "hide skill", but the cover used may give +1 (partial cover) or +2 (full cover) to the surprise roll. Similarly, if an adventurer takes steps to reduce noise, they can move quietly, getting +1 to surprise, or +2 for extreme measures. These modifiers do not stack.

(I use "surprise on 5+ on 1d6", which makes adding bonuses directly to the roll easy. If using the more standard "surprise on 1-2 on 1d6" mechanic, add the bonus to the target, making it "surprise on 1-3" or "surprise on 1-4".)

Cleverness may matter. If it seems reasonable that the adventurer may leave subtle clues to their presence, even when not visible or trying to be quiet, the adventurer only gets the bonus if either their Int or Wis is higher than the target's Int or Wis.

Speed may matter. If there's a chance that the adventurer could notice the opponent first, they can dart for cover if their Move is higher than the opponent's Int or Wis, whichever is better. This only works if the adventurer is not surprised, of course.

A thief does not need cover, only shadow, nor do thieves need special equipment to reduce noise, other than avoiding metal armor. The thief adds half their level (round up) to surprise when hiding in shadows or moving silently, plus any bonus for cover or sound dampening, if they chose to do so. If the thief decides to hide in shadows on the spur of the moment, they only need to dart for shadow, not for full cover, and they can do so if either their Dex or their level is higher than the opponent's Int or Wis.

Friday, March 16, 2018

Multiple Attacks for Fighters

I really need to get back into the flow of posting here again. So here's a topic to start with: how many attacks per round should a fighter get in OD&D? Not talking about attacks against enemies of 1 hit die or less, but house-rules similar to AD&D's "3 attacks every 2 rounds" for 7th level fighters.

Attacks per round is an abstraction, of course, when using one-minute rounds. There are actually many attacks in one minute -- but we only roll once. If high-level fighters get multiple rolls, it's not because they make more attacks per round, but because they have a chance for more significant attacks per round.

I don't like the AD&D system, but I'm thinking of two other options:

  1. Keep it as just one roll, but increase the damage: 1+1 dice of damage at level 4, 2 dice at level 8, plus another die of damage every 4 levels.
  2. Fighters facing opponents of 2 dice or more get an extra attack roll if their level is twice that of their opponent, or two extra attack rolls if it's three times the opponent's level or hit dice.

Saturday, November 18, 2017

Treating Saving Throws As Attack Rolls

I've realized for some time now that you could combine the saving throw table with the combat table by setting each saving throw category as the equivalent of an armor type and just using one table. But I never got around to working out the equivalents -- until now.





(plus bonus vs. spells = Level/4, round down)



Thieves save as either clerics or magic-users, depending on whether you are using OD&D, Holmes Basic, or one of the other Basic D&D versions, or you can construct a unique saving throw scheme by comparing the other classes (perhaps save as clerics vs. death magic and wands, but as fighters vs. everything else.


  1. I'm using "Death [Magic]" as the name of the first category rather than "Poison" to emphasize my opinion that the saving throw categories should be reserved for supernatural situations. I don't believe in using save vs. wands to dodge javelins, for example. Poison is such an extreme situation that it gets the benefit of being treated like death magic.
  2. Magic-Users get the bonus to save vs. spells because in OD&D, they improve in this category faster than their other saving throws. A bonus equal to 25% of their level is a good approximation.

Thursday, September 7, 2017

Deciphering Maps and Scrolls

Continuing with adapting the 2d6 reaction roll table to thief skills, I thought I’d move on to deciphering treasure maps and magic scrolls. Unlike removing traps and picking locks, this is definitely not a mundane skill and shouldn’t be trainable. Thieves do not normally get the ability to decipher treasure maps until 3rd level, according to Greyhawk, although I’ve considered delaying it until 4th level. They can’t read magic scrolls until the 10th level.

Although most people consider this to be a Read Languages ability, I specifically limit it to deciphering things like treasure maps. Thieves don’t learn any extra languages without study. They pick up important words like “gold” and “pit trap” in multiple languages, as well as the rudiments of multiple scripts, enough to figure out important details in maps, inscriptions, and possibly other writings, without necessarily knowing everything that’s said.

Other characters, in contrast, either know the language (and can read the entire document) or don’t know it, in which case they get no chance to decipher the text. If you feel it’s reasonable, a character who knows a related language can roll on the table as well, probably shifting all results one step worse. They will not be able to decipher magic scrolls, in any case.

2d6 Result Description
2 Very Bad magic scroll backfires, important info on treasure map missed/misread
3-5 Bad unable to decipher map or scroll
6-8 Average treasure map read successfully (but not magic scrolls)
9+ Good treasure map or magic scroll read successfully

Using the table’s results for magic scrolls should be pretty straightforward. Treasure maps, inscriptions, and the like might be a little trickier, depending on how you create treasure maps. My assumption is that a treasure map tells you how to find a start of a route to a treasure, how to stay on that route, and what traps are along the way, as well as any known guardians and a clue, at the very least, as to what the treasure contains. A Very Bad result, then, would still allow you to find the route and follow it, but would not mention a trap, or misrepresent a guardian (reading the word “wraith” as “goblin”, for example.

Deciphering maps and scrolls is harder than other thief skills. Thieves only add one-fifth their level to the roll (round down.) Again, a natural result of 2 is always treated as a failure. “Low-level maps” is not really a thing, although if you felt like defining a language’s difficulty as higher than normal, you could adapt the concept to this table as well.

A scribe class (something I’ve toyed with before) would add half their scribe level to the roll, instead of one-fifth. It’s more in keeping with a scribe’s primary talents.

Creative Commons license

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0) license.

Written with StackEdit.

Wednesday, September 6, 2017

Lockpicking Reaction Roll

Here’s another 2d6 Results table for a thief skill, in this case for lockpicking. Thieves add half their level to the roll, but a natural roll of 2 always fails. You might want to be generous and say that a thief only jams a lock on a modified roll of 2 when picking a low-level lock, but still breaks the lockpick. Low-level means that the lock’s complexity or difficulty level is lower than the thief’s level. I define a lock’s level as 0 for a typical door lock in a town, or equal to the dungeon level, unless otherwise specified. A merchant, for example, will probably have a higher-level lock to protect valuables.

2d6 Result Description
2 Very Bad lockpick breaks, jamming lock
3-5 Bad lockpick breaks, but may try again, if you have another pick
6-8 Average unlock low-level lock
9+ Good unlock any lock
(12) Very Good (non-thief unlocks lock)

Treat unjamming a jammed lock as a separate lockpick attempt. In other words, with the right tools, the thief can remove the broken lockpick, clearing the lock, and allowing further lockpicking attempts.

If the thief has no professional tools, they can improvise, shifting all results one step worse (unlocking low-level locks on 9+, other locks on a 12.) Thieves do not ignore “jam” results on a natural 2 when using improvised tools.

Non-thieves, if you allow them to pick locks, also shift all results one step worse. Non-thieves cannot quickly improvise tools. Instead, they must try to make a tool, then use that as an improvised tool. This will slow things down quite a bit, compared to thieves.

You can follow the same rules described for non-thieves removing traps: treat effective level as zero unless trained, in which case use years of experience, and the maximum roll possible equals the character’s Dexterity. Non-thieves never get a level bonus to the roll, even if they have mundane training.

Creative Commons license

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0) license.

Written with StackEdit.

Tuesday, September 5, 2017

Traps Reaction Roll

There was a discussion about thieves on the OD&D forums recently, during which someone brought up the topic of using the 2d6 Reaction Results Table for resolving thief skills. I don’t think that’s a great idea when you are only looking for a binary, yes-no result. But a couple of the thief skills could benefit from having 3-5 possible outcomes, instead of just yes-no. And I whipped up some quick tables, which I will preserve here.

First up: a Remove Traps table with four (and a half) possible results.

2d6 Result Description
2 Very Bad trap triggered
3-5 Bad trap not removed, not triggered
6-8 Average low-level trap removed
9+ Good trap removed
(12) Very Good (non-thief removes trap)

So, about a third of the time, any trap can be removed. If the trap is low-level, it can be removed more than half the time. Low-level is defined however you want. For me, a trap’s level is equal to the dungeon level, unless otherwise specified. You could also use trap damage as a guide (4d6 damage = 4th level trap.) If the thief’s level is higher than the trap level, the trap is considered low-level and easier to disarm.

If the trap is not disarmed, most of the time it is still primed. On a Very Bad result, the trap is triggered and the thief takes damage. It’s up to you whether the thief gets a saving throw or not. I think I’d skip a saving throw unless the thief has prepared or otherwise takes action to reduce or prevent damage. Example: Thief suspects a fire trap and pours water over head and clothing before trying to disarm the trap.

Thieves get a bonus to the roll equal to half their level, but a natural result of 2 always triggers the trap. Optionally, non-thieves can try to disable traps as well. Shift all results one step worse, and treat the non-thief’s character level as zero if they have no training. So, no traps are considered “low-level” for untrained characters. Furthermore, rolls are capped by Dexterity.

Every year of mundane training in traps is treated as the character’s effective level when compared to the level of the trap. A character with five years of experience in mundane trap removal can remove 4th level traps or below as if they were low-level traps (9+ on 2d6, for non-thieves.) Mundane training does not give a bonus to the roll, as a thief gets.


  • Fighter, untrained, Dex 11: Can’t remove any traps.
  • Fighter, untrained, Dex 12+: Can remove traps on a roll of 12.
  • Fighter, trained, 2 years experience, Dex 11: Can remove 1st level traps only (on roll of 9+).
  • Fighter, trained, 2 years experience, Dex 12+: Can remove 1st level traps on roll of 9+, other traps on roll of 12.
  • Fighter, trained, 5 years experience, Dex 12+: Can remove 1st to 4th level traps on roll of 9+, 5th level and higher traps on roll of 12.

Creative Commons license

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0) license.

Written with StackEdit.

Monday, July 31, 2017

On the Fly Spell Research

Carrying on from the starting spells post: I’ve repurposed two columns from the Intelligence table in Greyhawk: Minimum # per Level as the base number of starting spells known, and Maximum # per Level as an optional cap on the number of starting spells (but not on total spells known.

That leaves one column unused: % Chance to Know any Given Spell. This column strikes me as unusable as written, for a couple reasons:

  • Too harsh. Players may have their hearts set on getting Fireball, for example, and their hopes can be dashed with a single die roll, no chance to try again.
  • Too complicated. I don’t want to roll for every single spell, even if I put off the roll until the spell is encountered.
  • Too much book-keeping. I’d need to photocopy at least one copy of the spell list to mark off which spells have been checked, and keep those sheets in my records. I’d rather not.

But rather than using this as a harsh limitation, I could use it as an empowerment. Let magic-users have a chance to decipher a spell reasonably quickly, without the need for research.

I already use Read Magic as a research shortcut, allowing M-Us to learn a new spell from a scroll or spellbook immediately after casting Read Magic. Otherwise, they have to use the spell research rules to learn the spell, which is going to take 1 week per spell level, minimum. But I could give M-Us one chance per spell discovered to figure it out just by studying the actual document for a few hours.

I don’t think I’d use percentile dice, myself. I’d rather just use a d20 vs. Int roll. Success means the M-U learns that spell. Failure means either weeks of standard research, or casting Read Magic if they know it.

I haven’t decided exactly how long this should take, but I’m thinking at least 2 hours per spell level, double that for Int 3-4, half that for Int 17-18. That guarantees that it’s still not the thing you’d usually risk in a dungeon, but it’s still useful for skipping a lot of downtime.

Creative Commons license

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0) license.

Written with StackEdit.

Saturday, July 29, 2017

Spells Known

I’ve written before about a technique for assigning starting spells for a new magic-user character, based around the fact that the spell lists in Men & Magic and Greyhawk (and even in the AD&D Player’s Handbook) are numbered.

  1. Roll 1d8 (or 1d12, if using Greyhawk) and look up that number on the list of first level spells.
  2. The basic list of starting spells for that character is that spell and the next five spells, in order, six spells total.
  3. Roll another 3d8 (or 3d12) and consult the list three times. If you roll a spell again, drop it from the base list. If you roll a new spell, add it. The character thus starts with 3 to 9 spells.

Here is a recent modification of that process that makes some use of the Greyhawk Intelligence table (Spells Knowable.) You only need the “Minimum #” column on the table, although you can optionally use the “Maximum #” column, too (and I have some ideas about the “% Chance” column, too, but I’ll save that for later.)

Instead of the first 1d8 roll determining six base spells, use the minimum number indicated by intelligent. So, for M-Us of Int 15 and 16, the steps above remain the same: 1 die for 6 base spells, 3 dice for +/- 3 spells.

For an M-U of average intelligence (10-12,) the base list of starting spells is only 4 spells. For Int 3, only 2 spells. For Int 18, 8 spells.

The number of dice to roll for the spells to drop or add from the base starting spell list is half the minimum number of spells. So, 1d8 (or 1d12) for Int 3-9, 2d8 for Int 10-14, 3d8 for Int 15-17, and 4d8 for Int 18. You can optionally cap the total number of starting spells based on the maximum number column, but I personally would ignore it, and I would definitely not use it for spells found or learned afterwards.

Another option, which I think I will use, is to roll for random spells learned from the 2nd level and higher spell lists, when those spells become available. These would be spells mastered during training and research to level up. Not sure whether to use the exact same random range, or maybe halving the minimum number, so that the character acquires fewer spells later on. I think it’s a wise idea to halve it, so that players are kept hungry and given a reason to keep searching the dungeons for new spells. Under this rule, 3rd level Int 3 M-Us would get 1-2 new spells of 2nd level, Int 10 M-Us would get 1-4 new spells. and Int 18 M-Us would get 1-8 new spells. They would always get at least 1 new spell, the same way that starting characters should always get at least 1 starting spell.

Creative Commons license

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0) license.

Written with StackEdit.

Monday, July 3, 2017

Tenth, Fifth, Half, Whole

I’ve got a vague inkling of an idea I’d like to muse about.

You may have noticed that, in the Highlander Mage post, I wrote that these magic-users gain the ability to cast 1st level spells for free (more or less) when they reach Level 8. Or, to put it another way, that the spell level they can cast for free is 1/8th their character level. But I’m thinking the number should be 1/10th, because it makes it very easy to eliminate dividing by some number and instead just dropping the last digit of the character’s level to get the spell level. 10th level wizards can cast 1st level spells for free, 20th level wizards can cast 2nd level spells for free.

But also, I’m still thinking about very simple class customization schemes. It’s better to use just a handful of easy to remember numbers, for example 1, 2, 5, and 10, and the equivalent fractions.

The Max Spell level of pure spell-casting classes is 1/2 their character level. So, too, is their hit dice. Their high-level abilities like enchanting kick in at level 10.

Compare that to a combat class (Fighter.) Their hit dice are equal to their level, and they get no spells… but if we wanted to broaden spell casting ability, we could say that their Max Spell level is 1/10th their level and must be trained at great expense, so that Fighters couldn’t even cast a single spell until 5th level, if you are rounding to the nearest whole number.

Any extraordinary ability could be generalized to fit that pattern. A Thief class (slightly revamped) would have hit dice equal to half their level, and a bonus to Thief skill rolls equal to half their level. If a non-thief character wanted to learn Thief skills, they would require special training and the bonus would be equal to 1/10th their level. Of course, under this scheme, it might make more sense for non-combat classes to likewise have reduced combat ability equal to 1/10th their level. Not quite as extreme as in Lamentations of the Flame Princess, for example, but not what most old school players are used to.

The 1/5th modifier would kick in for mixed class concepts, where the class has more than one special ability. A magical thief class would have 1/2 hit dice, 1/5th max spell level, and a thief bonus of 1/5th level. The 1/5th modifier would also be used for customization of magical schools, which is an idea I’m still mulling over.

To summarize:
* hit dice are on a 1:1 basis for mundane classes (Fighters)
* non-mundane classes (spell-casters and extraordinary talents) have half hit dice
* magic or talent level is half level
* having two extraordinary talents or spell-casting ability drops modifier to 1/5th class level for each
* any out-of-class ability can be learned with a 1/10th class level modifier
* high-level abilities appear at 10th level and have a modifier of 1/10th class level, where appropriate

Creative Commons license

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0) license.

Written with StackEdit.