Everyone is a Thief (a different approach to skills)

Well, let's consider this: everyone is a thief.
Or better yet, every character is an adventurer, a tomb robber; they are out for silver and gold and nothing will stop them. So why not make everyone a thief?
If we accept that a lot in the OSR is about adventure and discovery - about playing low-levels characters, about exploring the space provided by your modules from the prospective of adventurers, who are basically tomb robbers or similar... well, then yes, every one of them is a thief.

We saw in the previous posts about skills (one and two) that a skill system could be beneficial... Without it, it's hard to have consistency and to judge your chances, as a player, and a lot is left to the GM's fiat.
But if we introduce a skills system, we need to solve a few problems:
1. How to avoid making characters other than the thief unable to attempt certain tasks?
2. How to avoid skills replacing role-playing?

My answer to the above (and it's only one of the many possible answers) - at least in the context of this article - is the following:
1. Make everyone a thief
2(a). Make a meaningful skills list

To the above, I would add:
2(b). Have solid principles: as the GM you should be open to say "yes", and description, imagination, and planning, skilled and creative (role)play, and GM judgement, prevail of the need to roll.


Make everyone a thief
This house-rule would be simple. Take your thief skills list and every character is of one of the other available classes (i.e. a fighter, a magic-user, or a cleric) but is also a level one thief (with the thief skills).
When leveling up, every character has its own progression as by its class (fighter as fighter, etc.) but also increases the skills as a thief of the same level.
It's not multi-classing; everyone is a thief of a level equal to its own level, with the related skills.

Pro: no "you cannot do that - it's a thief's thing" any more. No better progression in skills for a single class (the thief). No unbalanced probabilities with mixed thief's skills and roll-under-attribute or other skill's rules. No more getting stuck in front of a door, of a wall to climb, of a trap to disarm. Actually, anyone is has the potential to resolve those situations - and everyone is a potential victim of those situations!

Con: there is no more a "pure" version of the thief (or specialist). But in any case those who play the thief very often are frustrated that they are not also playing an assassin, in some way, or some top-class adventurer... I bet those playing the thief would be the most eager to be able to multi-class (as a fighter, I guess, and close second as a magic-user to have even more tricks at their disposal).


Make a meaningful skills list
But what are the correct skills for this list? If you use a standard game such as S&W or LotFP, you can start with their respective original lists. But most likely, you want to tune that list.
What your new list should present, though, is not the list of skills for a thief or specialist in the traditional sense. What the list should provide, is a list of skills for an adventurer.

A good list, in fact, has a clear focus on the most important actions of the game - keeping in mind both a certain consistency (i.e. if you disarm traps well, you probably can pick a lock or pickpocket someone decently enough, while disarming traps has no connection to climbing or swimming) and perhaps also the frequency/importance of the skills (i.e. picking locks will always be favored vs. language skills unless your game focus heavily on negotiation... therefore you may want to give language points more easily at level-up...).
Your list will determine not just the chances of success, but also will inspire your players to do certain actions (or avoid those with low chances or those not listed, for fear of failure or uncertain odds).
So, making this list, should be like painting the portrait of an adventurer - with what it does better and what worse.

Most important, though, is to make sure that your skills list does not replace role-playing. Having a skill for "medicine" for example could make sense, especially if you want to minimize or replace the cleric's healing spells. "Medicine" makes sense because you cannot role-play the action of trying to heal another character (you may still improvise a good plan for example if you lack resources, but that's not the same as the consistent use of a "medicine" skill).
On the contrary, a skill like "negotiation" or "persuasion" is very risky because it takes away a crucial part of role-playing.


The original skills lists
The skills' lists below are taken from BXE, S&W Core, LL and LotFP for reference, and sorted so that the same (or similar) skills are on the same line.

BXE	S&W Core	LL	LotFP
Climb Sheer Surfaces	Climbing Walls or Cliffs	Climb Walls	Climb
Hide in Shadows	Hiding in Shadows	Hide in Shadows	Stealth
Move Silently	Move Silently	Move Silently	Stealth
Backstab	Backstab	x	Sneak Attack
Pick Locks	Opening Locks	Pick Locks	Tinker
Find or Remove Traps	x	Find and Remove Traps	Search/Tinker
Pick Pockets	Delicate Tasks	Pick Pockets	Sleight of Hand
Hear Noise	Hearing Sounds	Hear Noise	x
x	Read Normal Languages	Read Languages	Languages
Scroll Use	Read Magical Writings	read and cast from scrolls	x
x	x	x	Architecture
x	x	x	Bushcraf

Possible Issues in LotFP
So, what are the possible issues that may arise from applying this house-rule to your game?
The main problem that I can see is only for LotFP (or other games where players decide where to assign the skill points):
- it becomes "Everyone is an assassin" for fighters
- the party (as a whole) becomes overpowered

With players free to assign points as they wish, level by level, it's likely that fighters with go first of all for Sneak Attack and Stealth to combine the benefits of backstabbing with the to-hit bonus of the fighter.
And in general, a group of smart players would differentiate the points-investments (for example, one player maximizing points the investment in Search and Tinker, another in Climb and Swim, another in Languages, etc.). This way, the group would be composed practically by very specialized characters, each one of them getting quite fast to 5 or 6 skill points in the chosen skill(s).
A "normal" Specialist (intended as the chosen class when not everyone is a thief) would probably have to differentiate more the points investment to be able to cover more areas. With "everyone is a thief", on the other hand, you effectively have as many Specialists as characters, thus more points available every level-up.
This is slightly mitigated by the fact that for example Magic-Users take longer to level-up, and that some characters with high specialization would die (this is LotFP, in the end), forcing the players to re-do the points investment.

Other games (with a % skills progression table) do not suffer from this problem. Everyone is a thief but with the same chances at performing the same task at the same level.
So if you want to bypass this potential issue, you could adopt in LotFP a % table for the skills, forcing players to a classic slow progression.
I don't like this a lot, though, because I think that specializing and differentiating skills is part of what makes each character unique, and I would rather risk to have characters a bit more powerful than what they would usually be if they were just one class, and not class+thief.

Skills (part two)

See part one here.
We've seen already that Skills are a difficult topic to when it comes to OSR.
On one hand, having skills system helps players mostly to judge chances and ensure a somehow consistent behavior across sessions.
On the other hand, a list is always somehow limited. Not having a skill in the default list might influence players in thinking that they cannot do something out of the list, while a very long list is impractical, confusing, and leads players to distribute skill points in skills which will rarely come up, thus frustrating them. Also, a low value in a default skill would discourage players to attempt anything which has a basic low chance of success.


The first article in the post of today contains a sample list of skills, which is well suited for your classic fantasy game, and presents a different approach to skills. The author calls them "Genius", but the interesting part is not the name per se, but its premise ("time, the most important resource of an adventurer") how it works (roll for success and time).
The system presented uses a classic d6 with also attribute modifiers, if the GM allows it - which makes for a decent chance of success. But its most important part is the roll for the unit of time: the skill check will determine how many units of time (rounds, turns, hours, days, etc...) it will take for the attempt.

Genius: Yet Another Approach to Skills in OSR Play

by E.T. Smith
[...] Skills are a thorny subject in the context of old school gaming. Learning to embrace the freedom that comes from forgoing codified action resolution is one of the major experiences of old school play, and yet it can't be denied that delineated skills show up early in role-playing's history. And it can't be ignored that most people, when role-playing, expect to have defined skills on their character sheets.
Skills are definitely utilitous, from a purely procedural perspective. They offer quick clear means to resolve events and to define the capacities of characters. Unfortunately they tend to take the narrative away from discussion and negotiation, turning it over to the dice instead. And they curtail player initiative by discouraging any action that doesn't have a clear numerical advantage behind it. [...]
Regardless of origin, the approach immediately appealed to me because it put the focus of resolution not on pass/fail, but on time, the most important resource of an adventurer, the passage of which is the danger intensifier of any adventure.  It doesn't really matter if you can unlock the door, so much as if you can unlock it before being discovered by somebody with reason to stop you.

Genius, a System for Character Excellence
Every character has aptitude in a non-combat, non-magical field of expertise. All characters start with 1 point of Genius to define as they choose. [...]
Some potential types of Genius:
Wilderness Travel
Ancient Lore
Masonry & Construction
Religious Ceremony
Weaponcraft
Politics & Statecraft
Seamanship
Trade & Barter
Burglary
Taxonomy of Monsters
Forgery & Counterfeiting
Brewing & Cooking
Music, Dance & Theater
Alchemy [...]

Resolving Challenging Use of Genius 

To check genius in a challenging situation, throw 1d6 twice. The first throw determines how many units of time the effort takes. This can be days, hours, turns or rounds depending on what makes sense for the situation (maybe even years!) but usually it’ll be turns.

The second throw determines positive or neutral results. Add the points of a character’s relevant Genius to the throw, and the relevant attribute modifier if the referee allows it. If the total is 6 or higher, the effort succeeds. Characters may subtract 1 from the initial time result for every point a successful throw exceeds 6. If after modification the time throw is zero or less, the effort requires only one unit of the next lower time increment (hours down to a turn, turns down to a round, and so on). [...]

http://trollbones.blogspot.com/2018/09/genius-yet-another-approach-to-skills.html

This system provides a "slow" progression for characters (thieves, or Journeymen as called in the article, or specialists as in LotFP) will receive only one point every two levels, and other characters one point every four levels. Either you level-up very fast, or the system will still cause some trouble (most likely, not attempting to use skills with low scores).
I believe it could be inserted, just for the 1d6 for the calculation of the units of times it takes, in a game like LotFP with no particular effort.
As GM, I would also consider the option of spending units of time to increase the chance of success. The system offers a rule focused on levels of success: the better you succeed, the faster you are.
But what about a fail-forward mechanic which actually allows you to "buy" a +1, a +2, etc., by spending more time on the task at hand? Instead of keeping that door locked, grant your players a bonus but roll an additional encounter or two.

The second link also provides a fresh approach to the issue of skills and what characters can do. Or, as the title says, what characters can't do.
The author recognizes all the issues that we mentioned for skills, and tries a different approach: assume that you are decently capable except for a few things that you are really bad at.
While I like the premise of the article, I am not so sure on how the author developed it (it looks to me too close to a list of character Traits, rather than a focused rule for Skills), but it is an interesting take nevertheless, and I am sure it can inspire you to come up with your own house-rules.
If indeed you do, I'd love to link them here.

You Are What You Can't

by ktrey parker

I've been musing a little bit about one of my main issues with Skill Lists that have been inserted into D&D and the various clones over the years. I understand that defining the chances of success at a given task was an important driving factor behind this, but I chafe under the limitations it tends to impose. If a Skill List is provided, players have a tendency to use it as a guide what their character is capable of and this is a little anti-ethical to a certain style of Old School Play. Conversely, extensive skill lists also do tend to inform play by reminding a player of options that might not be readily apparent. [...] 
I'm wrestling with a very different approach.
What if characters were actually defined or differentiated by what they were “bad at” or “incapable of doing” rather than what they are “good at” or “capable of doing.” In a way, flaws are essential for interesting choices and characters, but other than the occasional low ability score (increasingly rare in more modern systems with standard arrays, cracked and subverted bell curves, and point buys) limitations above something broad (“No Edged Weapons”) or vague (“I'm Lawful, I wouldn't do that”) are still conspicuously absent from most D&D games.
They may develop through play or be informed a little by Alignment choice, and the classics always crop up: “My Dwarf is Greedy,” “My Fighter cannot tell a lie” “My Thief is a Kleptomaniac” but there's nothing up front to encourage or direct players to behave in this fashion, other than this wonderful spontaneous development during play or occasionally a seed of an idea on the Player's part (informed by fiction or the prior imagining of something that “seems like it would be fun or interesting to try to play,”) or maybe even as a rationalization of low Ability Scores (I've messed with this a bit before with my previous post on Ability Score Tags).
So instead of a comprehensive skill list, or waiting for skills to develop through play and negotiation, what if part of character generation spelled out some of the character's flaws and deficiencies with a specific eye toward “You can't do this?” [...]

Roll Or Choose

1 - I Always Cheat or Favor Short Cuts

2 - I Always Donate 10% Of My Wealth to The Church or Charity

3 - I Always Overestimate My Capabilities

4 - I Always Manage To Disappoint My Family and Friends

5 - I Am A Buzz-Kill

6 - I Am A Compulsive Gambler

7 - I Am A Coward

8 - I Am A Glutton

9 - I Am A Vegetarian

10 - I Am A Very Heavy Sleeper [...]

The object here initially was to highlight “anti-skills” or “incapability” more so than “flaws” or GURPS-style “Disadvantages” but that's what it morphed into as I started trying to brainstorm and fill out d100 table.
Entries like “I Can't Swim” are much more in tune with the original intention, this limits a Player's options very succinctly and concretely and could lead to interesting situations, back-story expansion, and tough choices. Binary characteristics like this have a more obvious impact, but naturally it's difficult to come up with a negation for every potential skill.
I can't shake the feeling that beginning play with a few examples of INCOMPETENCE or INEPTITUDE would immediately and directly inform play in a pretty different way. Instead of “I Can't Sneak Around because I'm Not A Thief” this option remains open or on the table (unless you choose or roll “I Cannot Be Stealthy, or am Very Noisy” of course). [...]
http://blog.d4caltrops.com/2015/02/you-are-what-you-cant.html


The last link provided in this post is very long, but definitely worth checking out.
It includes first a list of interesting guidelines (i.e. how do you decide if instead to just say yes and go on playing, or how do you decide if to roll dice as a resolution system), and a nice summary on the best attributes of a good skills system (and number 3 is indeed the most interesting part to read of the introduction to the post).
Consider, to summarize, things like this:
- Determine what basic characters can do without the need to roll (just say "yes")
- Determine what requires a description, or imagination, or planning, and what requires a dice roll
- Resolution mechanics (and dice rolls) should not replace skilled and creative play or GM judgement
- A list of skills is not the ideal solution, and a 1 in 6 chance is probably too low
- Skills are not necessarily just a function of the character level: they don't need to start very low or to improve with levels (levels are mostly just for combat/spells skills and related stuff); this changes the way Thieves work, but they are not one of the original classes...

OSR Skill System Fight Club
by Lloyd Neill
The first rule of OSR Skill System is there is no OSR Skill System…
One of the defining tenets of old school play (as much as there actually is such a thing as a unified “old school”, or defining tents of it) is player skill over character skill, specifically that description, role play, creativity, and problem solving should be looked to before game mechanics and character sheets when resolving in-game challenges. It's also often taken to mean rules-light, however it’s possible to have quite complex situational rules within an old-school game. I think a more helpful distinction is around the attitude a game takes to rules. The commonly used maxim rulings not rules, suggests that flexibility, judgement and at-the-table decision making should take precedence over codified rules, or at the least not be excluded by the existence of rules covering a situation. I find personally, (and for many other folks in the OSR if the amount of blog and G+ posts are any indication) that one of the murkiest areas of this in practice is around resolving non-combat situations and the use of “skills”. [...]

1) Does the task being attempted even warrant more than hand waving it and getting on with the game?

Consider:

Is the task actually difficult / complex / technical and are there significant consequences attached to failing at it?

Should someone with the character’s class, level, ability scores and background / shit they just made up to justify being able to do it, be able to just do it?

(It’s  worth considering at the start of play what the baseline of PC competence in your game will be: what can anyone do?; what can members of a specific class do?)

What will be the impact on play of spending time resolving the task? (e.g. how long will it take, will it make the game more fun?)

As a GM what is stopping me from simply saying "yes" or "no"? ...and if "no" why not "yes"? [...]

2) If the task needs to be resolved in some way, should you be using dice? [...]

I like Ben Milton’s simple rubric for deciding when the dice should be used:

Whenever possible, players should overcome challenges by simply describing what their characters do. [Dice rolls] are only used to resolve risky situations that would be too time-consuming to describe, or involve immediate danger. 

 e.g. Disarming a trap doesn't involve immediate danger, so as long as it is fairly simple, you have to describe how you do it. 

  e.g. Picking a lock doesn't involve immediate danger, but describing the process would be tedious and hard to visualise, so you [roll]. 

 e.g. Dodging dragon's breath is easy to describe, but it involves immediate danger, so a [roll] is required. [...]

3) If the final decision is that rollin’ dem bones is the best way to proceed, what’s the best approach? [...]

Ideally a “skill system” will provide guidance for the other two points above, not just a dice mechanic

Anyone should be able to attempt anything

For tasks that are "hard" there should be a distinction between whether the task is technical or difficult (i.e. is the limiting factor for a PC a lack of training or of talent - my personal preference is that a character with experience and training relevant to a technical task should be more likely to succeed than a novice with high ability scores).

The existence of a resolution mechanic shouldn't replace or override skilled and creative play or GM judgement (e.g. no use of "search" or "disarm" checks to circumvent description and problem solving 

The situation at the table should determine the task/skill to be checked, not skills written on character sheets

No lists of skills (this is probably implicit in the last point, but anyways). Whilst I think there are some really well implemented systems based on skill lists (such as LotFP) the moment you start codifying discrete skills you open a door to the need for more (Rolemaster, I'm looking at you... *)

If you are going to rely on dice to resolve a situation, characters should have an average chance of succeeding at tasks.  I'd much prefer a 50% base chance of success for an average task than the 1 in 6 for many activities in B/X (secret doors I'm looking at you:  pixel bitching + 17% chance of success ≠ fun)

Ability in non-combat skills should not directly tie to character level. D&D levels represent combat power, and shouldn't necessarily increase capability in all things (a number of retroclones have basic skill systems tied to character level, either directly, or indirectly through saving throw and I'm not a fan of this approach)

Should be equally valid as a system for detailing a character before play or emergent during play

Advancement should be possible through a number of possible expenditures of resources - character level only being one of them [...]

http://deathanddismemberment.blogspot.com/2016/03/osr-skill-system-fight-club.html

The article provides an analysis of various skills systems, but I think its introduction is very useful if you want to design your own system.


Design notes:
- Skills provide a way for players to judge their chances of success, besides what's described by the GM and which is subject to interpretation
- Skills usually ensure a consistent behavior and consistent consequences across sessions and adventures
- A list of skills might influence players, both if too short (narrowing the scope of what they feel they can accomplish), and if too long (impractical and confusing)
- Low chances of success in a default skill, moreover, would discourage players from attempting those actions
- Time is a crucial factor in a lot of games: connect the skills system also to the units of time it takes to attempt an action
- Consider not only levels of success as a factor to reduce time, but also an additional effort which takes a a longer time as a way for players to get more bonuses
- Present perhaps not a list of skills, but a list of things your character is not good at (somehow related to Traits, perhaps?)
- Determine what basic characters can do without the need to roll (just say "yes")
- Determine what requires a description, or imagination, or planning, and what requires a dice roll
- Resolution mechanics (and dice rolls) should not replace skilled and creative play or GM judgement
- A list of skills is not the ideal solution, and a 1 in 6 chance is probably too low
- Skills are not necessarily just a function of the character level: they don't need to start very low or to improve with levels (levels are mostly just for combat/spells skills and related stuff); this changes the way Thieves work, but they are not one of the original classes...

Next time, we can try to put together a few ideas I had. The post will be called Everyone is a Thief.

Skills (part one)

Well, I guess this post should start with a quote of the first line of the first post I'm linking below:
Skills. Mention them in relation to Old School games, and some grognards start to cringe. 

Skills are indeed a controversial topic in OSR but I believe they deserve some attention.
First of all, if you think that your game should NOT have Skills, EVER, then it's easy. Just skip this article.
But if you are interested in adding a skill system or in some inspiration for house-rules, this post is for you.
(And honestly, I believe that even those who say no-skills-ever, have anyway used at least the roll-under-your-ability-score ruling... which is indeed one of the proposed solutions)


So, the first post contains a nice overview of some basic ideas for Skills:
- a d6 system with a 1 in 6 default chance (LotFP style), perhaps with bonus by class and race
- a d20 system with a roll-under the related ability score
- a system derived by Saving Throws (a similar matrix, or perhaps even the same matrix - adapted with a grain of salt - if you want to save time)
- a list of skills, a list of Difficulty Classes (DC), factoring in the ability modifiers... which steers away from the traditional OSR but it's something already done and proved to be working fine... Except for the part where skills are used as a substitute of role-playing

Skills in Old School Games
by AHL
Skills.
Mention them in relation to Old School games, and some grognards start to cringe. However, this is not one of those blog posts discussing skills or no skills. This post is about how to handle in game situations in old school games.
Also, I mention just a few systems here. There are many more, using 2d6, 3d6 and so on (just google “OSR skill system”), and they’re more or less complicated. Here, I will talk about the ones that I have used at some point in time and some ideas I’ve been thinking of.
I’m not sure that Old School games is in need of a skill system per se. However, I’m positive that Old School games is in need  of a solid task resolution system. These days, “Bend Bars”, “Open Doors” and “Find Secret Doors” just feels wonky. [...]

Old School Galore I – the d6 system
This is the bona fide old school way. When something is to be resolved, you have a x in 6 chance to succeed. Most of the time, it’s just a 1 in 6 chance, but some classes and races get bonuses to specific tasks. [...]

Old School Galore II – the Ability Check system
When a situation isn’t covered by the rules, old school systems often suggest an ability check. Basically, you try to beat (i.e. roll under) an appropriate ability on a d20. [...]

Akrasia’s Save system 
For S&W players, there’s a fan made simple system out there (Akrasia’s house rules), which is also used in Crypts & Things. The idea is to use the Saving throw number also as a target number for various tasks. The list of tasks is deliberately kept vague, to empower flexibility, but all PCs can do anything. [...]

5e’s skill system
I think that the new D&D really have nailed the task resolution system. Everyone can do anything, but some are better than their peers at doing said things. Also, as a GM, I appreciate to have a defined task (or skill) list to choose from when deciding what the player should roll in a given situation. [...]
5e has a list of 18 tasks, usable by anyone. Sweet. The GM decides a Difficulty Class (DC) for the roll, which is the target number to beat. The player rolls a d20, adding his ability bonus for the specific task (for example Stealth uses Dex). [...]

Bolting a task resolution system to your OSR game

Another problem if the system you like haven’t got a built in task resolution system is how to define what skills/tasks your characters are extra good at. Of course you could write down a list of “fighters are good at this” and “dwarves are good at that” and pre-define such things. Another way is to grant the players a couple of things (maybe 2) they are good at. For example, your fighter might be good at “Sneaking” and “Obtaining information”. [...]

Dawnrazor’s Old School Universal Task Resolution System mk I
Here’s what I’ve decided to use for my adventures (unless they’re for a system like B&T or FH&W).
1. Task format (in adventure text): Task at hand (DIFFICULTY: ABILITY). Example: Spot the hobgoblin (HARD: WIS) 
This format excludes the actual skill, but since I suspect that not everyone will use skills, or if they use systems with different skill sets and names, it will get confusing. [...]
https://nerdomancerofdork.wordpress.com/2014/09/20/skills-in-old-school-games/

This second post notes a couple of potential issues with the lack of a skills system in an OSR game. What if instead of skills you have classes (like the thief, or the ranger)? It's easy for GMs to rule that other characters cannot do anything covered by the thief or the ranger skills.
What if you stick with the original cleric-fighter-magic-user but no skills system? Everyone can try everything, but it's hard to adjudicate based only on common sense and GM's rulings.
It introduces a heavy GM-fiat component, and even those GMs with the best intentions would find it hard to be consistent across sessions. And it would always be hard for players to judge what chances do they have to accomplish anything, before actually trying.

So, yes, a skills-system is a good thing to introduce, probably, but it must be easy enough, scalable by level and/or training, and possibly with no DC (no impact on existing old-school adventures which have no DC or ability tests or stuff like that).
The system provided in the link is a bit too complex for my taste, but its guidelines are good and they provide both a system for character vs. character tests, and for which kind of bonus to apply to various kind of other tests.

Skills: The Middle Road

by Benjamin David
Rules for non-combat skills in D&D have spanned a rather broad course over the years. In OD&D, there were no rules for non-combat related skills other than magic until the introduction of the thief--a point that most grognards put in its favor, since it encouraged everyone to try their hand at everything. However, this lack does present problems when trying to determine exactly what a character can do that's better than other characters, and as more classes were created to cover these non-combat niches (first thieves, then rangers) the implication set in that unless your class abilities said you could do a thing, you couldn't except by DM fiat. [...]

The extensive rules for dealing with non-combat skills in other systems, both classic and modern, speaks of the desire of players to be able to know in some quantitative sense what their characters are good at. However, if we are to come up with any such system as a house rule for OD&D, it has to meet several basic parameters:

"Having" a given skill, NWP, or whathaveyou should not, as a rule, be a requirement for attempting any adventure-related action or for having a reasonable chance of success.

The system must be scalable, allowing for characters to improve existing skills by the expenditure of time and wealth or as a reward for a successfully-completed adventure (as described in my previous post),

And yet it must be simple enough that no OD&D product must be significantly altered to employ it (i.e., the referee should not have to go through every adventure and install DCs for every challenge or create a complete set of skills for every goblin) and that any referee can easily ad-hoc it during play. [...]

There are four levels of competancy for any given skill: unskilled, skilled, expert, and master. All characters are assumed to be unskilled at any given task unless it falls under their class (especially in the case of thieves) and/or background. Having a background that encompasses a particular action means that one is skilled only [...]

Of all character classes, only thieves automatically advance in skill levels as they increase in character level, and then only in those areas directly related to thievery [...] In all other cases, advancement or gaining new skills must come as a result of gameplay [...]

Before explaining further, a particular mechanic must be described. The skill die denotes the type of die rolled for a given skill level, to which is added the appropriate ability modifier (using the Moldavy scale) when a skill contest arises between two individuals. For example, a 1st level thief with a 16 dexterity is trying to sneak up on a 1st level cleric with a 15 wisdom. Since the thief is considered skilled at moving silently, he would roll a 1d8+2 (modified for dex) against the cleric's 1d6 (unskilled) +1 (wis) to determine if he succeeds. [...]

In cases where there is no contest between two statted characters--for example, a thief sneaking up on a goblin whose wisdom score is unknown, a character climbing a cliff, or a character with a jeweler background trying to make a gift to impress a noblewoman--the referee should assign an ad-hoc possibility for success. [...]

Unskilled
Skill Die: d6
No bonuses to die rolls
Skilled
Training: 1 month and 100o gp
Skill Die: d8
+1 bonus when making skill-related rolls with a 1d4, 1d6, or 1d8
+2 bonus when making skill-related rolls with a 1d10 or 1d12
+3 bonus when making skill-related rolls with a 1d20
+15% on making skill-related rolls with percentage dice [...]

http://rodoflordlymight.blogspot.com/2009/03/skills-middle-road.html?m=1

The third link proposes an elegant system with a skill die, which increases as you become more trained at a given skill. To succeed you need to roll 5 or more (Paolo Greco has a similar system in his own OSR system... more on this at the link).
The good things about this system is that it's easy to roll and evaluate, that it has a verbal connotation of your skill level (i.e. rookie or novice, etc.), that it slows down in the improvements as you go ahead in the game, so no character becomes easily overpowered.
What it lacks, is a list of skills.

Skill Systems: Tweaking "The Middle Road"
by ktrey parker
[...] The “Ratings” are expanded a little bit, success is still on a 5 or more (making UNABLE impossible for the player without circumstantial modifiers...which are generally kept low/stingy). The bell-curve comes into play with MASTER which really increases the chance of success, but with a cost.

Skill Rating	Skill Die/Dice
UNABLE	1d4
ROOKIE	1d6
NOVICE	1d8
VERSED	1d10

[...]
I don't tie skill improvement to level in any meaningful way most of the time. This is mostly to mitigate the “20 HD MASTER Blacksmith NPC” and “I gained a level and am suddenly an EXPERT at Tracking” issues, although Thief classes and formal training-as-cash/time-siphon can open up some additional improvement avenues during downtime with some good fictional explanations. Instead I like to tie Skill Rating improvement to actual play.
Any successful Skill Roll (5+) prompts for another skill roll immediately, and on a maximum die/dice result, the first letter of the next Rating is written down. Once you spell it out, you've achieved the next Rating (this is why all Ratings have 6 letters).
This has diminishing returns: It becomes more difficult to improve/master a skill as you get better as the chances of success increase, the chances of rolling the maximum also decrease. I also like that improvement is actually tied to “Doing the thing,” so players are encouraged to attempt it, even if the odds may not be great (Practice makes perfect!). [...]

http://blog.d4caltrops.com/2016/05/skill-systems-tweaking-middle-road.html

I have to admit my preferences for the d6 skills system provided by LotFP, as you've seen already when I proposed to change even Saving Throws to d6-skills, here: d6 Saving Throws for LotFP.
But while I like the default skill system of LotFP, I am less in love with dice pools, and therefore this link is less interesting for me. Still, it is definitely worth mentioning also because James Raggi is doing something similar (but for Saving Throws, not Skills) in the beta document of the next version of the rules (which who knows when and if it will ever be published).

Game Mechanics: Dice, Doors, and Decimal Points
by Dan Domme
So I have a skill system hack I am working on and I want to share it with you.
Start with Lamentations of the Flame Princess [...] One of the nice things about it is that skills are based on d6 rolls. Virtually everything is a 1-in-6 chance, and specialists (read: thieves, but a less pigeonholed concept) get to invest points into expanding these skills. E.g., put two points in Sleight of Hand and your chance to execute such a task goes from 1-in-6 to 3-in-6. [...]
I am hacking that d6 system a bit. Mostly because I love dice pools, but there is also a logic behind it. 
First off, rather than expanding the range of success, I expand the number of dice you roll, while keeping 6 the target number. Note that statistically this is more difficult. [...]
A solution to the increased difficulty is that the GM should be more generous with the bonus dice. We already add dice based on invested Specialist skill points, but let's also add Ability Score bonuses. A +1 to Strength is easily added to your pool if you want to Open Doors, for example. Have a crowbar? Add another die. And so on. [...]
The problem that we have already come across thus far in my Megadungon game is that opening doors is still fairly likely to result in failure. Even with crowbars, people helping, and strength bonuses, you are pretty likely to not open the door.
This is where I got the idea to go straight into story game territory and offer the players a narrative choice. In the first option, the players could choose to let the door be. The door is swollen shut, just like St. Gary said it probably would be. The other option is to note your margin of failure. (E.g., was your highest result a 4? Then your margin of failure is 6-4 = 2.)  I'll let your character(s) persist at opening the door until they succeed, but I get to roll the margin of failure in Wandering Monster checks. (A dice pool of 2 in this example.) Again, each of these is a simple 1-in-6 chance.  The logical basis is that the worse your initial check result, the more noisy your success is going to be and the more likely you are to attract attention to yourself. [...]

http://failedmysave.blogspot.com/2017/07/game-mechanics-dice-doors-and-decimal.html

Design notes
- Without skills but with classes (i.e. the thief, but also the ranger), it's too easy to rule that other characters cannot attempt actions as the thief
- Without skills but with just the original three classes (cleric, fighter, magic-user) it's all related to the GM's fiat
- It's hard to have consistency and to judge your chances, as a player
- Skill systems in the OSR must be simple, scalable by level/training, with no or very simple rules for difficulties (to remain compabile with old modules)
- a d6 system with a 1 in 6 default chance (LotFP style), perhaps with bonus by class and race
- if using a d6 system, consider also to use d6 dice pools and 6 as target number
- a d20 system with a roll-under the related ability score
- a system derived by Saving Throws (a similar matrix, or perhaps even the same matrix - adapted with a grain of salt - if you want to save time)
- a list of skills, a list of Difficulty Classes (DC), factoring in the ability modifiers... which steers away from the traditional OSR but it's something already done and proved to be working fine... Except for the part where skills are used as a substitute of role-playing


Will continue here.

Some theory behind Saving Throws

After a month or so discussing Saving Throws, I promise this will be the last entry about this topic, at least for a while. In this post, I want to present some very interesting discussions and theories about the subject.
For those coming late to the party, though, these are the links to the previous posts that discussed of Saving Throws:
- d6 Saving Throws for LotFP
- Saving Throws compared across rulesets
- Saving Throws ideas
- Saving Throw Effort


Let's look at an interesting prospective about Saving Throws: they give better odds against more mortal threats. Let's look at it in the order as in the article:
- vs. Death (or Poison) has the better score of all
- vs. Wands or Paralysis or Petrification it has a still a "decent" score (well, not so good but we're talking about Saving Throws)
- vs. Magic or generic Spells even lower
- vs. Breath or Area Effects, the lowest of all

Now, the progression is not so linear and it is lost, if you check across the various rulesets present here in Saving Throws compared across rulesets, in many OSR retro-clones.
The change of scores messing up the order of severity is something I never noticed before.
I particularly like this analysis because it does not focus on the usual "simulation of reality" but instead presents a game design reason for scores that otherwise would just be arbitrary numbers on a table.

Saves as Severity

by Delta

[...] Here's an observation I haven't seen expressed before: The save categories in OD&D are most easily interpreted as just levels-of-severity. Consider the following:

The first category is "Death Ray or Poison", and receives a +4 bonus [...] which represents instant death. In order to give our characters a fighting chance, a fairly hefty bonus is given.

Second is "All Wands -- Including Polymorph or Paralization". [...] render the victim effectively helpless and subject to a follow-up coup de grace. Hence a relative +3 bonus is given to avoid these effects.

Third is "Stone", i.e., turn-to-stone (petrification). [...] Bonus is +2 here compared to baseline.

Fourth is "Dragon Breath", which is not instant elimination from a failed save, but (obviously) pretty bad, major business. Bonus is effectively +1 in this case.

Fifth and finally you have "Staves & Spells" which is in some sense "everything else", i.e. non immediate death or incapacitation. This is our baseline, hardest to avoid, i.e., +0 bonus in similar terms. [...]

But more generally, you could use these principles for judgement on the fly about the severity of effect: basically you're awarding between +0 and +4 to the save, with more heinous effects given a more generous save (again, just to give the characters a fair, fighting chance). [...]

http://deltasdnd.blogspot.com/2009/12/saves-as-severity.html

Another post about Saving Throws, by Brendan (which has proven already that he has many interesting things to say about this subect): 4 reasons to love them (as GM, I guess...), and why are they so important.
First, they allow to introduce a save vs. death or save vs. consequence which remain threatening even against high HP or other high scores at high levels. No one is invulnerable. No one rolls dice vs. a tremendous damage, which would kill nearly anyone. You roll a Save or die, not grind HP to zero.
Second they progress by level but do not make characters invulnerable. This also implies that even high level characters may die with a single failed Saving Throw roll. Keeping the game deadly at high levels is a powerful thing.
I care less for the fact (third) of giving players something to roll.
But the fourth point speaks aloud to me: what are the Saving Throws categories saying about your game and setting? If you play low-fantasy with no dragons, change the Breath Saving Throw into something else.

Why I Love Saving Throws

by Brendan

1. Saving throws solve all the problems with hit points. [...]
Saving throws (coupled with critical hit tables) are even the best way to model serious injuries (allow a save versus critical hit when HP drops to zero or whenever your system of choice would threaten an injury). [...]
2. Progress without certainty. [...]
The essence of the traditional saving throw is progression with level. What this means is that better saving throws are a reward for surviving a long time. However, even if your saving throw is really good (down in the single digits), there is still a nontrivial chance of failure, and failing a saving throw is often fatal. [...]
3. Proactivity. [...]
Saving throws are proactive in the sense that they are something the player does. They get to do something in order to avoid some potential bad outcome. [...]
4. Atmosphere.
This is a minor point compared to the others (which in my opinion are critical to traditional Dungeons & Dragons), but it is still worth mentioning. The saving throw categories in original D&D are:
Death Ray or Poison
All Wands – Including Polymorph or Paralization [sic]
Stone
Dragon Breath
Staves & Spells
This communicates a tremendous amount of information about the setting and the challenges that are present. If these categories don’t match the challenges that characters are likely to face in your campaign, I would recommend changing them [...]
http://www.necropraxis.com/2012/07/31/why-i-love-saving-throws/


The third post today is by Jeff Rients, and contains a list of categories from various editions, for comparison. Note how Jeff also observes how categories have merged or separate across different versions (sometimes it's Wands / Staves & Spells and other times it's Rod, Staff or Wand / Spell for example).

More important, notice two important rulings:

- If a category is quite restrictive such as "Dragon Breath", he doesn't use the Save for different breath attacks

- When assigning the Saving Throw to a category, he goes in order if undecided, so the first one applies

Saving Throws, part 1

by Jeff Rients

[...] Today I'm going to start this analysis with simply listing the categories of save for each edition that I have within reach as I type this.
OD&D
Death Ray or Poison
All Wands -- Including Polymorph or Paralization [sic]
Stone
Dragon Breath
Staves & Spells
AD&D1
Paralyzation, Poison or Death Magic
Petrification or Polymorph
Rod, Staff or Wand
Breath Weapon
Spell
[...]
Observations
One of the interesting things I see here is how the various categories evolve, couple, and de-couple. Poison and Death/Death Magic/Death Ray are always the same category of save, but Wands may or not have anything to do with Staff. [...]
A saving category called Dragon Breath implies that non-draconic breath weapons don't allow a save unless specifically indicated. This is a special case of the broader principal that may or may not inform your own game. I tend to assume that the category names Mean Something, in that if an attack form doesn't obviously fall under one or more category on the chart then the implication is that a save isn't allowed. [...]
In most editions Wands are the second save listed, but AD&D they're third. This matters when I run because of a rule of thumb I use. Some attacks fall under mutliple possible saves. Like a dragon breath attack that is poison gas or a wand of paralyzation. Sometimes the monster/item/whatever description tells you what to save against, but sometimes it doesn't. In the latter case I give precendence to saves listed earlier on the chart. [...]

http://jrients.blogspot.com/2009/05/saving-throws-part-1.html


The last post is very interesting because it suggests a abstract point of view on Saving Throws, which goes against possible house-rules which connect the Save to an Ability/Attribute, but does so for a very precise reason and with an original, yet very interesting intention.
Saving Throws, according to Courtney Campbell, are not a simulationistic tool. They are abstract: in a single roll you determine if you suffer or not the consequences of an impending danger, but if you succeed, a Saving Throws does not tell you how. This is possible if you keep abstract categories such as Death/Poison or Breath/Area, but if you connect Saving Throws to Abilities/Attributes or use the next categories "reflex, will, and fortitude", such abstraction is lost.
To make an example, if you trigger a poisoned dart trap, and make a reflex Save, you should simply narrate how you avoided the dart. If instead you roll vs. Death/Poison, there are plenty of possible descriptions: the trap didn't trigger, the trap didn't hit the character, the character avoided the dart, the dart didn't pierce the armor or the poison had lost its power, etc.

On Abstraction and Saving Throws
by Courtney Campbell

Modern systems seem to assume a baseline representation - i.e. I rolled twice, so each roll represents a swing of my sword or I can possibly move up to 10' a second, so in six seconds I move 60'.

At first blush this seems to make a lot of sense, but if you look at it too closely the abstraction inherent in hit points and saves breaks suspension of disbelief. i.e. Hit points suddenly becomes literal wounds dealt by specific sword blows. There are 3 saves reflex, will, and fortitude, and they literally and in a direct and visceral way represent 'getting out of the way' 'resisting with your mind' and 'enduring with your body'. [...]

But what of old school saves you say? Abstraction, and this indeed is why they are cool. [...]

Why that's the coolest thing about them! Nothing specific at all! All we know is success or failure - the actual means of that is up to you. (and your classes general ability to handle that specific kind of threat is built into the numbers) [...]

The point is, that the game *doesn't* tell you how you make your save - that's part of the discovery of what's happening and the fun. Logistically it's a lot more fun to come up with answers for why things happen then trying to plot out a specific sequence of events that is occurring every six seconds. Also, you've got a lot more room for awesome and rule of cool in your descriptions. [...]

http://hackslashmaster.blogspot.com/2010/12/on-abstraction.html

What I especially liked about this article, is that it pushes the game, the descriptions, to be a little more inventive and original. It resonates with me (although often in my house-rules I used to connect Saving Throws to abilities) and it gives me a new prospective (and the next time, I might go back to the original categories or make my own, but stop connecting Saving Throws with abilities).
In fact, it connects well with my attitude for example to describe combat in original ways. Combat is not just hit and miss. A failed to-hit roll in combat could be many things: it's not just a sword swing which misses the target; it can be described as blocking, as parrying, as hitting the armor or missing just by an inch, as the target dodging cleverly, and so on.

Design notes: 

- Give better Saving Throws scores vs. more deadly threats
- Give lower Saving Throws scores vs. less deadly threats or threats that only affect HP
- This is not to "simulate" something specific but rather to increase a little the characters' resistance, or give them "a fair chance"
- Bypass HP buffers, or in other words: Save or die is still "fair" when it makes sense, because the game has indeed Saving Throws built-in for deadly threats which should bypass HP as a combat/health buffer
- Progress by level but still do not make characters invulnerable (again, Saves are against serious, deadly threats, even at high levels)
- Saving Throws categories say something important about the most deadly threats of your setting
- Make sure that Saving Throws categories make sense for your game, or tune them

- If a category is quite restrictive such as "Dragon Breath", it might not be applicable to similar breath attacks by other creatures (decide if you want a category to be restrictive or inclusive)

- When assigning the Saving Throw to a category, go in order if undecided, so the first one applies

- The original Saving Throws categories are abstract and provide more freedom in their descriptions

- Connecting Saving Throws to abilities or to the new "reflex, will, and fortitude" turns them into a "simulation" tool, which narrows their scope and descriptions