D&D General Nostalgia : Thief Percentages

[Dioltach]

I played a Thief in 2nd Ed up to 7th level, when I got sick of being outclassed by the party's Kender. I multiclassed into Fighter and spent the next few years sucking at everything.

I do see that no one I play with (3.5 now) besides myself plays a Rogue. And quite honestly I generally take one level of Rogue to get the massive skill points at first level, and then multiclass into something a little more combat-effective.

 

[Dormammu]

Same could be said for the Thief's original design, re: breeding many unfortunate outcomes. The problem is that you don't see those outcomes as unfortunate, because they're subjectively so (in both cases), and you're in the extreme minority (to judge from sales of RPGs that at least TRY to balance this stuff against those that intentionally do not). I get that that may be annoying, but it doesn't make that any less subjective. Just sayin'...

I agree that the OG Thief has tons of issues. The best example was just how poorly defined backstab was. I skipped over 2E (it never seemed different enough from 1E to warrant the new rules purchase and I was exploring other games by them), but in those original incarnations it was essentially left up to the DM to determine when you could backstab, as I recall.

I liked Thieves, but I was nearly exclusively a DM in those days so I wouldn’t say I played them a bunch. What I don’t prefer about combat balance being the core tenet of class design is that it teaches people coming into the hobby with a modern edition that combat is the core of role-playing. My experience is that too much combat (whether too often or too long) is the #1 cause of players checking out and getting bored during play.

Elements like the Thief or weird old MU spells could steer players in the direction of problem solving over fighting. I would never go back to those systems though; they’re just too rough. I would choose 5E over 1E for any game I played or ran, but that element is one of the things I like about early D&D.

 

[Voadam]

Backstab was a pretty juicy ability, though it was often limited by DM interpretation. Looking at the rules, though, it's pretty clear - the thief just has to attack from behind with a sword, dagger, or club. Honestly, I know past Ralif has been more strict about it than that.

It looked like that from the 1e PH:

Back stabbing is the striking of a blow from behind, be it with club, dagger, or sword.

But it was actually a little more restrictive than that.

Back Stabbing: Opponents aware of the thief will be able to negate the attack form.

and

SPECIAL ABILITIES: When striking unnoticed from behind, a thief gains a bonus of +4 on "to hit" rolls and inflicts twice the normal amount of damage.

and

By striking silently from behind, the thief gains two advantages

 

[The Crimson Binome]

Yes. I miss percentage skills the way they were in 2E. I would often go with a high F/R traps and open locks at first level. I find the current method disingenuous as progression is an illusion because they just raise the DC on the trap or lock so that you really aren’t getting better. Sometimes the opposed roll is a joke that means you don’t really get better.

It's easy to run a game poorly, regardless of mechanics, but the presentation of those mechanics is also important. When you're using a d20 system, it's trivial to represent a higher or lower difficulty by adjusting the DC, so the DM is much more likely to do so (which can seem like they're moving the goal posts, if it happens to line up with level progression).

With a percentile system, the base value has a much more intuitive meaning, and it requires somewhat more work to adjust for difficulty. Even if it's theoretically the same math, the DM is less likely to apply modifiers, compared to a system where they have to set the DC.

 

[Voadam]

It's easy to run a game poorly, regardless of mechanics, but the presentation of those mechanics is also important. When you're using a d20 system, it's trivial to represent a higher or lower difficulty by adjusting the DC, so the DM is much more likely to do so (which can seem like they're moving the goal posts, if it happens to line up with level progression).

With a percentile system, the base value has a much more intuitive meaning, and it requires somewhat more work to adjust for difficulty. Even if it's theoretically the same math, the DM is less likely to apply modifiers, compared to a system where they have to set the DC.

Locks had a 90% (+30% to -60%) variability range from the core 2e DMG.

Lock Quality

The quality of a lock can increase, decrease, or leave unchanged a thief's chance of picking that lock. The higher the quality of the lock, the harder it is for the thief to pick. Table 24 lists the different lock qualities and the amount they add or subtract from a thief's percentage chance to open it. Unless otherwise noted, assume that all locks are of good quality.

Table 24:
Lock Quality

Quality Modification
Wretched +30%
Poor +15%
Good 0%
Excellent -20%
Superior -40%
Masterful -60%

In 3.5 the DC's for a lock are:

Lock
The DC to open a lock with the Open Lock skill depends on the lock’s quality: simple (DC 20), average (DC 25), good (DC 30), or superior (DC 40).

 

[NotAYakk]

The thief skills where a mess. I remember back when one book at thief skills for low levels; when they came out with the next book, they retroactively made lower level thieves worse, to give room for higher percentages.

---

A method I've played with to make percentages less clunky is to always roll two d10s. Then you get 5 levels of success:

Botch: Doubles, over the target %.
Failure: Reading the dice both ways is over the target %.
Success: Reading the dice one way or the other is under the target %.
Masterful: Reading the dice both ways is under the target %.
Critical Success: Doubles, under the target %. 00 is always a critical (it is a "roll of 0" not 100).

Skill	Botch	Failure	Success	Masterful	Critical
10%	9%	72%	18%	0%	1%
20%	8%	56%	32%	2%	2%
30%	7%	42%	42%	6%	3%
40%	6%	30%	48%	12%	4%
50%	5%	20%	50%	20%	5%
60%	4%	12%	48%	30%	6%
70%	3%	6%	42%	42%	7%
80%	2%	2%	32%	56%	8%
90%	1%	0%	18%	72%	9%

This lets DM's say "just don't botch", or "Regular success" or "Masterful", or "Only on a critical".

Low skill% get Successes pretty quickly. For reliable Masterful rolls, you need a chance well over 50%.

Every +10% skill gives you 1% more Crit, 1% less Botch. If you count a Botch as -1, a Failure as 0, a Success as 1, a Masterful as 2 and a Critical as 3, then each 10% skill "earns" you an average improvement of 0.2 "points" per attempt.

Toss that at the old-school thief check tables, and I strongly suspect you'd end up with much less comical lower level thieves (they wouldn't do things masterfully), and higher level thieves would start being able to pull off masterful things.

(Mathematically, this is really similar to rolling twice, or an advantage/disadvantage system. But with percentiles. Failure means "would fail even with advantage", Success is "would succeed with advantage", Masterful is "would succeed even with disadvantage". But I find it is much faster to glance at 2d10 and see if both/neither/one is under your skill % than it is to do advantage/disadvantage d20 math).

 

[The Crimson Binome]

Locks had a 90% (+30% to -60%) variability range from the core 2e DMG.

As I said, presentation matters. Regardless of the underlying math, a DM is more likely to adjust the DC of a check than they are to modify your percentile roll.

 

[cbwjm]

i have seen that as an optional rule. I don’t think it is in the PHB and have never seen it used. If it was it was rare. And I have every module ever published from that era.

The different quality locks are in the DMG and weren't presented as an optional rule, but it still would have been up to the DM to use them. I should note I'm talking about 2e, I know hardly anything about 1e.

 

[Hoffmand]

Now I think 5E has the best thief class of all editions. Although there are many features in pathfinder 1E I really really like. However, I never had a shortage of peoples wanting to play the thief class in pre 3.x games. I had players lined up to play the thief. They were normally people that really liked to solve things without combat. And I was very generous with backstab with my interpretation of the rules.

 

[Dormammu]

The Thief skills may well have been the direct inspiration for many early percentile-based RPGs, like Basic Role-Playing and Rolemaster.

I randomly came upon a source for my own supposition. I was listening to an interview with Lawrence Whitaker (co-designer of Mythras and other BRP games for many years), and he said RuneQuest came about when Steve Perrin decided to adapt the Thief skill system into a full percentile skill system.

Here it is, if anyone is curious: Lawrence Whitaker interview (YouTube link, queued to the time he mentions it).