TechFindings

I propose using 2d6 (+possible ability modifier – possible difficulty) the resolve situations that can not just be role played in OSR games (where ability modifiers are -3 to +3).

General
-4: Failure, with extra negative consequences
-5-6: Failure
7: Partial success or failure with some unexpected twist
8-9: Success
10+: Success, with some extra good outcome

Bluff
-4: Not believed, hostile reaction or plays along
5-6: Not impressed
7: Bluff is not called, but reaction is unexpected (and not in a very good way)
8-9: Bluff is believed
10+: Bluff is believed and target is extra helpful

Breach or Destroy
-4: No success, exhausted some resources (equipment or possibly HP)
5-6: No success
7: Partial success, some malfunction or unexpected side effect
8-9: Success
10+: Success with style or some advantage

Climbing, Swimming
-4: Falling, Drowning
5-6: Did not start, had to go back if previously made progress
7: Some progress
8-9: Reached goal
10+: Reached goal in style or with some advantage

Fish, Hunt & Gather Food
-4: Found nothing eatable, exhausted some resources (possibly eating something bad)
5-6: Found next to nothing
7: Found one days ration
8-9: Found 1d4+1 days rations
10+: Found 2d4 days rations

Jumping
-4: Fall
5-6: Hesitate (on retry, -7 is Fall)
7: Partial success if possible, otherwise hesitate (on retry, -6 is Fall)
8-9: Success
10+: Success with style or with some advantage

Luck (like setting an ambush or a bait)
-4: Things turn out very much the opposite of the desired outcome
5-6: The desired thing does not happen
7: Be careful what you ask, you just might get it
8-9: The desired thing happens
10+: Things turn out remarkably well the way it was supposed to

Make item or mechanism
-4: It fails later, or loss of resources/injury immediately
5-6: No success
7: Not quite fit for purpose, 50% chance of malfunction or requires some support
8-9: Success
10: Unexpectedly good result

Make Shelter
-4: Shelter fails later on, loss of resources
5-6: Inadequate, possible loss of resources if used
7: Decent shelter if something is paid/used/wasted
8-9: Good shelter
10+: Shelter with some benefit

Perform
-4: Failure, making a fool of oneself, possible injury
5-6: No one is impressed
7: Audience is undecided or split
8-9: Good performance
10+: Surprisingly impressive performance, some advantage follows

Recall Knowledge (Lore, History, Geography)
-4: Remembers incorrectly, sure about oneself
5-6: No memory
7: Recall something relevant but not quite useful
8-9: A good general idea about the topic
10+: Knows significant details

Track
-4: The tracked party is aware of being tracked and can choose to escape or ambush
5-6: Lose track
7: Sudden encounter with tracked party
8-9: Localized tracked party at a distance
10+: Undetected, close enough for ambush (or just observing)

Background

Classic OSR games use 2d6 + CHA modifier for Monster Reaction and Retainer Reaction. There are multiple outcomes, not just success and failure.

OSR games don’t really have skills but sometimes things need to be randomly resolved.

Why not using the 2d6 + ability modifier, and comparing to a table of outcomes, not only for reactions? I see some advantages with this. The multiple outcomes drives the story forward in different/random directions, rather than the open/closed gate mechanism of success/fail skill check. Also, the focus is on what the characters want to do in the story, not what skills the characters may have.

A first nice thing about OSR is that things can play out without rolling dice. This is what we call player skill. But occationally I find it unreasonable as a DM to judge the outcome based on the players description on their actions alone. This is where the second nice thing with OSR comes into play: also the DM can be surprised and needs to adapt and improvise.

Difficulty and Ability

The primary purpose of these Reaction-like checks is to produce random reasonable outcomes in significant situations. If the characters are already good (or bad) enough, or the task is easy (or hard) enough to simply decide the outcome, no dice should be rolled at all. Thus, the default is that the DM and the players do not really have any real insight into the probabilities of different outcomes (it would suffice to roll 1d5 with no modifier).

It is not necessary to add an ability modifier. It is not like it is the right of the player/character to add a strong ability score. DM shuld not consider difficulty much and probably most of the time should not add difficulties. After all, the purpose is to take the story in an unknown direction.

Thieves/Rogues

Thieves (or Rogues) have skills of their own. Those are not to be replaced or made redundant by above rules. If a fighter can climb a wall using these rules the thief probably succeeds automatically, and if the thief needs to roll for his special ability no other class need to even attempt.

Custom Outcomes

Obviously nothing stops you from defining your custom outcomes for a specific situation, in advance or when the situation comes up. Something like:

-7: Sentenced to death by hanging next morning
8: Queen approves (back to prison)
9-10: King approves (back to prison)
11+: Both Queen and King approves (released)

Other options

There are options to rolling 2d6 + ability modifier.

• 1d6 + modifier (which is seen in B/X for kicking open doors for example): gives very much significance to the ability score and not too many possible outcomes.
• 1d12 + modifier: I have never seen it but it could work just as well.
• 1d20 + modifier: gives too little significance to modifier in my opinion, and creates longer more arbitrary intervals of outcomes. Also the natural 1 and 20 are very uncommon and I prefer more variation in results more often.
• 1d20 under ability value: feels too much like BRP to me.
• 3d6 under ability score (as for Phantasmal Killer in 1e): Mostly just produces two outcomes and feels overly complicated

So I think 2d6+modifier makes sense, and it is established in old D&D versions (also for Clerics Turn Undead). The BECMI employer reaction table (Rules Cyclopedia, p132) looks like:

2: Resuse, insulted
3-5: Refuse
6-8: Roll again
9-11: Accept
12: Accept, impressed

This scale makes the middle results much more likely and the extreme results less likely than the scale I have proposed above. I think “6-8 roll again” (that will be almost 50% of the cases) is not optimal. The scale I propose leaves more of the options more likely, even when an ability modifier is applied.

If you really prefer 1d20+mod to 2d6+mod, I propose:

-4: Failure, with extra negative consequences
-5-9: Failure
10-11: Partial success or failure with some unexpected twist
12-16: Success
17+: Success, with some extra good outcome