Paper vs Electronic Notes

 OK, for real: paper is extremely hard to beat. Paper has specific qualities: it has a higher energy cost to commit things to paper or rewrite them and you are always constrained by physical space. These might seem like disadvantages but the effects are positive: I am a firm believer in the effort principle (it's shown that students that write things down retain information better than those that write electronic notes), and you are naturally encouraged to be concise and clear.

Physical material can be physically organised - less-used stuff can be moved completely out of your field of view, you can quickly learn a specific and natural sequence of movements to access the information you need. The workflow of annotating paper notes is also just.... nicer. A mouse can never be as dextrous as the pencil. Going keyboard-only helps but maybe we don't want to learn Vim just to run elfgames.

If you're going to go electronic it needs to be because paper either isn't an option at all (impractical to physically bring the stuff with you) or you need electronic features like hypertext, word-search, indexing, file-linking, or versioning.

So personally: I'm leaning into the electronic stuff but with the mindset that you need to invest heavily in the underlying skillset+tools to make it good.

Levers, not Buttons: non-combat resolution doctrine and how to prep for that

Pressing buttons on the character sheet

 There's a standing complaint against the mindset of seeing the contents of a character sheet as a set of UI buttons with hard-coded effects and specific appropriate contexts for use for a player to press to create effects ingame.

The mindset is artificially confining, turning an infinitely open-ended activity into an analog computer game that also forces you to be in a room for 4 hours with other people arguing semantics. Sometimes fun if the argument is stupid enough but we should aim higher.

The thing that pains me the most is the the TTRPG genre flows from following the golden rule: "that anything may be attempted". Of course, the sticky bit is being able to actually adjudicate what happens generally. I judge systems on whether they provide efficient and flexible tools for resolving the most common gameplay situations satisfactorily and how well they equip you to step beyond them.

People run/play games adhering strictly to RAW for multiple reasons:

• the game is functionally a contest, which must be fair
• players must be able to move between different tables and get consistent rulings
• the GM, the players, and the system designers aren't operating on the same fundamental assumptions on how the game works, and possibly are all aiming to play different games
• sometimes the different game that they're trying to play is "win at system mastery so I will be the Alpha Nerd"
• the system breaks catastrophically or leaves you with no useful tools if you don't keep to the safe processes it provides
• habitual thought; GMs and players are trained to expect that things will just go badly and awkwardly if they step outside the safety of the clear printed rule. Happens a lot when people enter the hobby from reading game theorycrafting online.

Attempting to overturn this framework is one of the big banners that the OSR rallies under. But you can't just "remove" a habit (including a way of habitual way of thinking); you must replace it.

Alternatives to the button-based mindset (which can be combined in varying amounts):

• roll for skills as they are attempted - repeated successes "reveals" (permanently) that you "have" the skill and don't need to keep rolling. Repeated failures proves the opposite. (Tim Kask described this in one of his videos but I'm sure I muddled it.)
• rule of cool (I hate it)
• total loosey-goosey; whatever makes sense for the character, which is defined only in relation to several (hopefully shared) cultural touchstones. Requires a firm grip on the ideas in question and a good memory to resolve things quickly and consistently. My memory sucks, I usually play with really creative people who are trying to push the limits of the system, and I like to chew on problems, so this doesn't work for me.
• The tag is the definition: if you have the word on your sheet you Have The Thing (to a useful, consistent degree). A short step from loosey-goosey. Glog does this kind of thing.
• No- or few-keyword definitions: phrase abilities in natural language with abstract phrases that may be fitted to the game fiction and as few keywords as is necessary to make it meaningful. Glog does this for most class features, and it's my current go-to.
• (Special) Take a -2: Player wants to do something clearly related (but different) to their printed ability. DM tells them to use their normal rules with a -2 penalty. I remember seeing this frequently playing PF1, it seems practically universal. Works fine to extend rules-heavy high-granularity systems while preserving meaningful character building choices.

My current Glog-descended WIP hack, Strange Moons, is based on using "tag as definition" for skills, and classes provide "no- or few-keyword definitions". The class bit is fine, but actually using open-ended skills in a game needs a bit of structure; GMs need clear intuitions and processes that can accept them, players need to be taught how they can use them.

Firstly, do not name skills as a verb (e.g. "fishing") but from origin ("swamp fisherman"). The term is effectively a pin connecting backstory (or played history) to a sphere of implied skills; you know your way around a boat in calm conditions, you know how to maintain and use normal fishing gear, you know how to navigate a swamp. We can assume that a character will simply succeed at those things under normal and unhurried conditions (that framing is important).

If you are what you repeatedly do, you can repeatedly do what you are!

My mental model is to not see the elements of your sheet as "buttons" but as "levers". You're still ultimately looking at your sheet for things to use but now you're thinking about how to creatively and appropriately apply them to the situation - approach must be specified and gives the GM information for resolving things. "I want to do X thing, using my skills from Y."

Like a real physical lever, it lets you apply your energy more efficiently and effectively to the situation than if you lacked it (often making the difference as to whether you can do the task at all). It can be more or less appropriate for the job at hand. You can combine multiple levers (multiple different semi-appropriate skills) on the same job, but there are limits - you can't lift a truck with a crowbar, and having 20 crowbars doesn't help much.

Prepped materials and game processes that are compatible with that mindset

You make playable material with reusable components and maintain a library of those components that is optimised for quick access - and preferably committed to memory.

Locked Door toy example (untested).

• The "locked door" idea is decomposed into two different tables; the door and the lock, since you can have many ways of securing a door that aren't locks, and you can have locks on many things and different ways of putting locks on them. This makes each table more reusable.
• The door doesn't have a row for "open unlocked door" or "knock on door politely" or "cast disintegrate"; the lock doesn't have a row for "use key". The tables are for both preparing and running material when the component presents an obstacle and how it responds to obvious approaches. If the door is unlocked or the players have the key, it's not an obstacle and you don't need the table!
  • Obstacle == risk or resource cost (usually time) and some kind of meaningful choice between options. Free options shouldn't be listed unless you're likely to forget they are possible, and there should never be an option that is 100% better than another.
• to assist in prep (including sneaky-during-the-game prep), add flavour text, links to inspiring material, other tools, perchance rollers, hp-vs-break-time-vs-material-vs-size calculators, whatever
• to assist in running the game ongoing, they need to accept game notes to evolve with you over time. Include blank rows. In-game annotations are simulated with italics here. Include links to critical rules. Ensure that what you're writing is compatible with those. If you foresee that you'll be referring to multiple different pages regularly, try to combine them into a single quick lookup / flowchart document.
• The reusable component table suggests defaults, but specific examples always override these. You could say that the goblin hideout is filled with 10-second doors, secured with random rotting crossbars that contain biting grubs.

Door

Doors! They promise passage through the wall and block you from simply doing it. They may be locked or not, barred, barricaded, stuck, secret, big or small, trapped, squeaky or quiet, and perhaps talkative.
Listed values represent a strong door intended to slow down an attacker.

Approach	During	Then	Notes
Schmuck (bash, targeting lock)	6DR, 24HP / 5 minutes Noisy!	Flesh vs. fatigue	Groups don't reduce time but do eliminate fatigue  breaking down the whole door (no identifiable weakpoint) takes twice the effort
Massive damage: Breach, battering ram/explosion	as action	Noisy!
Bypass security / reach the unlocked side	5 minutes to make tool, 1 minute per blind attempt, Hard		wrong guesses about the other side make this Fruitless
normal door	3DR, 12HP, 1 minute	no fatigue	HEEERE'S JOHNNY!

In trying to make this I realised that I had made a silent assumption (that the door is strong) and that there was a lot of variation there. My personal solution was to add that fact to the top clearly, and then force an additional door type as its own row since it almost the entire rest of the table was reusable as-is. You could make multiple tables, multi-level tables, pull that information into a header block.... this one seemed good enough.
Now here's another problem I've created for myself: I've asserted that with these concrete (high-granularity) DR and HP values it will take the average adventurer to break through (good for quick resolution), but I have no actual rules to substantiate that.
The DR and HP is the part that I don't know, and the expected time is the part I do know; the times have been left in place, and the DR/HP model for them can be added back in when I'm able to do the math.
In fact, it's possible to draw a direct link between expected time needed, a default weapon, and a specific DR and HP. That can be its own general-purpose chart!

Lock

Every lock was made with its keys. The gods have thus far kept the secrets of perfect security all to themselves, via self-explanatory means.

Approach	During	Then	Notes
Rogue, pick lock	3x (finesse vs. DC 10), 1 turn per attempt	-
Strike off (padlock)	1 Action, flesh, DC 10
Break out of object	object's normal destroy time		This is the normal way

Writing this one forced me to re-examine how doors break and being distinct about whether an attacker is chopping the lock out of a door, or if they're breaking the whole thing down (e.g. in the case that it has a couple of bars securing it).

Finally, seeing this all in action

Player, faced with obstacle, suggests some plan for getting through it.

GM asks themselves:
"What's the expected skillset for solving this problem in this way?" (the prepared material might tell me)
"How much of the appropriate skillset is overlapped by what is being used?"
Finally, decide:

• pick a row on the table and use it directly
• pick a row and adjust it to be better or worse (consider annotating that line in the notes)
• create a new row and write it in
• interpolate between two existing rows  

Then you just do that. You resolved the situation consistently and flexibly and retained information to keep doing that consistently. Mission complete. 

 

Q: Where's the post introducing this obstacle-table thing properly?
A: It doesn't exist. I invented it and then refined it on the fly while writing this post. This is where it it is introduced. The top-level doctrine, the process for how it is carried out, and a material implementation of that process are too connected - they must be explained together and in relation to each other to be meaningful and useful.

Q: What's your stats and resolution system?
A: I'll tell you when I know :)

Q: Are you meant to do this with electronic or paper notes?
A: You actually run games?

Prepping from first principles, OR, I wrote this meandering advice post so that I could get the ideas out of my head and focus on writing a completely different post

 

General Advice

• Own your game. Hack, modify, extend, and experiment. Almost all the published stuff is crap. Most of the blog and youtube stuff is crap too. The best system is the one you will make.
• Time, both ingame and out of game, is the most precious resource of all. Usually, we have better things to do than spend our lives prepping games: we work on it to learn to do it better so that we can do less of it. However, if you enjoy some element of it, you can spend as long as you like; if you want to write an entire setting guide, go for it.
• GMing should not be seen as taking your turn in the barrel. You should like doing it!
• The GM's three game duties are: to keep track of what is true, to resolve what players do, and to communicate the current situation to the players... and doing these things should be an enjoyable social activity.
  • Your processes, material, and skills can be tested by how quickly and how enjoyably you can perform these three things. See prioritisation below.
• Reading blogposts and watching videos on how to run games is useful but you should never spend more time reading and watching things than prepping a game, running it, and thinking about it afterwards. After the first couple sessions, no more than 10% of your system (your entire toolkit) should be new to you at any time.
  • Three overlapping processes (can be used for the same situation) is a problem, not a solution.
• Go take a walk. It will help you think.
• What is impressive to you probably seems that way because it seems difficult. However, forcing yourself to work in difficult and unnatural ways and use content you don't understand does not give good results. Experienced GMs pick tools that work well in their own tool ecosystem: you cannot just transplant them. You cannot satisfactorily use anything not properly integrated with the rest of your kit.
• Simpler processes are better than complex ones. Less notes is better than more. Delete steps. Going faster usually means doing less.
• The Quantum Ogre, the Rule of Cool, and Skill Challenges are bad tools to reach for first, and should only be used in situations where the results really don't matter.
  • They break coherence. They are disconnected from the rest of the work, robbing you of value that you should be getting from that work. They create new information (partial information) without foundation, without examining what is implied by new established fact.
  • I believe that it's fine to decide that some things should just happen, but when a player asks "why did that happen" the answer they give themselves should never be "because the GM wanted it that way" or "because the GM rolled on a chart and that's what came up"
• There's an amazingly large amount of overlap between GMing effectively and being good at lying. Be bold and confident. 

 

The priority chart: doing less when you run games

The way that you use your system and processes, what you prepare and how; when the game is on, you must make judgements, resolve actions, and supply information, and better tools let you do this faster, by involving less steps.

Zero steps: Immediately Knowing
One step: Instant Lookup; Improvise; Lie
Two steps: Fill information (all immediately known) into memorised process, resolve
Three steps: Getting a player to give you the answer
Four steps: Look up right process, look up right information, fill it in, resolve
More: Rolling on charts, nested processes, undisguised generating new content
Failure: Searching your materials for the information that you don't know the location of; searching materials for information that may or may not even exist; doing a google search; asking chatgpt; trying to work out what the text means; trying to work out where an error happened

Special: creating things when the players know that the results will be exciting is still engaging (anticipation). Curate your random tables carefully.

Effective prep is whatever makes you able to operate higher on that chart more often.

Trying to prep everything to instant lookup (effectively, scripted content) is the obvious solution to the novice, and it only works as long as you can still navigate instantly it on the fly; the failure mode of this style is having to search it and work out what it means.
Instead, aim to adopt a toolset that lets you operate across 0-3 steps 95% of the time; material that is instantly usable for generating things as you need it is often just as good or better than the thing itself.

Do not just think about whether your processes will help you solve one step in isolation, but whether they continue to be good after an hour of using them ingame and you have a mountain of results and new information that also needs to be tracked: ongoing informational load needs to stay low.

Surprisingly, writing static material (stuff that is written for one specific sequence of events) can be effective prep, as it forces you to get more specific about the underlying facts of the material. Steal "acting is reacting" and practice how characters and factions will react to different situations.

 

 

Make a copy of this chart and empty it. Try to work out how you will achieve each of the three goals with your processes, material, and skills. This lets you identify where you have gaps in your toolkit and you can look for things that complement the parts you do have.

	Processes	Material	Skills
Keep track of truth (especially new / evolving truth)	(Goal: Simulate a world that naturally puts the players in Situations, while tracking as little player-unknown information as possible)	Game-dated searchable notes. Calendars, weather flowers, random tables, moodboards.	Creating efficiently both before and during the game. Keep information organised.
Resolve what players do	Reusable, non-overlapping, quick, believable.	Clear, concise reference information; current situation is a direct lookup to get required info.	Able to perform (Situation x Process x Material => Resolution) smoothly, seamlessly, without hesitating
Communicate effectively with players	(the one, two, many technique is a good one)	Player-visible maps (tactical mats)	Entire books have been written about this.

The Pit Fiend Story

 The system is Pathfinder 1, and happened maybe around 2018. It's some years after the event now and my memory's crap, so I'm probably misremembering a lot of it, but I found out yesterday that one of the other players in the party doesn't remember how it went down *at all*. So since I can't rely on anyone else to remember for me, I'm going to do the best I can. I think that the reason why we had to go to the mountain is wrong, but a wrong story that makes sense is better than an incomplete right story.

I was playing as a Hunter; a bardrangerdruid that's written to invest heavily in the Animal Companion. Basically, Tick-tock the Deinonychus (named for the way that his talons would clack on the floor) was the *real* character and Raul the Hunter was there to provide teamwork feats and other various buffs or spells.

Scene 1: The Dam
The party is in a dam, which is is powered by draining the life out of a Pit Fiend that's bound to a magic circle in the dam. Mechanically, the fiend's got a ton of *negative levels* - these massively penalise it in many ways. We need to activate the dam, but doing so will immediately the Pit Fiend - those have 20 hit dice, and it currently has 19 negative levels.

Pit Fiends can perform one Wish per year. We could really, *really* use a Wish for a power bump (Wish isn't Unlimited Wish) so we negotiate the following deal.

•  Mutual nonaggression, but the fiend isn't interested in indefinite servitude because that's not an upgrade from its current situation
•  We can't take any direct or indirect harmful actions. 
• We are to let it out and it is free to go immediately, but we meet back up soonish on top of a mountain someplace to get the first Wish.
• I can't remember what we specifically used the wish on but I think it was attribute bonuses, everyone wants more attribute bonuses.
• It gets the next Wish a year later to fix its level drain
• I'm pretty sure that we're damned upon our eventual deaths, either directly or indirectly for the act of letting a fucking pit fiend loose in the prime material plane

All party members and fiend sign said contract and power the dam in the interim with an animal instead. Now, we're currently low-ish level but we actually really don't want a full power fiend running loose in a year's time. We've got a lot of time to do enough adventuring and get strong enough to kill it before then, but it's still something that we'd like to deal with as soon as possible.

Scene 2: Clever Girl

• All PCs signed the contract
•  Tick-tock isn't a PC and therefore didn't sign (not going to get a wish either)
• Tick-tock has Intelligence 3 and therefore counts as a person and is able to independently make his own decisions
• He can't speak, but he does have Cunning Pantomime which lets him communicate just as well very slowly via acting

SO, one evening while we're all in town, Tick-tock takes the initiative and fucks off to a paladin order to try to get some backup on the mountain later. I don't know how that all went down - the interaction was offscreen.

Scene 3: The Mountain
Time passes, and we arrive at the mountain at the allotted time. Disappointingly, no paladins had turned up. We receive the wish and decide that we have to commit to trying to kill it on the spot, because we might not get another chance.

Now, Level Drain ("Negative Levels" to be specific) is a serious debuff.

For each negative level a creature has, it takes a cumulative –1 penalty on all ability checks, attack rolls, combat maneuver checks, Combat Maneuver Defense, saving throws, and skill checks. In addition, the creature reduces its current and total hit points by 5 for each negative level it possesses. The creature is also treated as one level lower for the purpose of level-dependent variables (such as spellcasting) for each negative level possessed. Spellcasters do not lose any prepared spells or slots as a result of negative levels. If a creature’s negative levels equal or exceed its total Hit Dice, it dies.

However, go ahead and look at the Pit Fiend page and apply 19 negative levels to that.
AC doesn't go down. HP drops from 350 to 260. Damage Reduction 15 is still in full effect, immunities and resistances and regeneration are functioning, spellcasting is still up.

Not having access to the statblock, we didn't do the math and seriously overestimated our abilities. We couldn't even scratch it. It just kinda rolls its eyes and teleports away - apparently it had more important business to attend to. Things have gone from bad to extremely fucking bad.

Scene 4: Dread
So we released a powerful and purely evil being into the world, it's completely unbound by contract, we have no significant allies, it knows who we are and the element of surprise is gone. We don't know how long we have before it decides to mop us up or why it hasn't already done so.

I put my foot down. No more half-assed plans. We need to look at what we've got and figure out something that will actually keep us alive. We bring up the entire loot sheet, respective spell lists, abilities. We take at least an hour, maybe two, working through ideas.

It's a devil, those can be summoned with strong enough magic. Magic circles can hold it. They make deals. It's highly level-drained. It's apparently very busy. Any further level drain would kill it, but it's got ridiculous Spell Resistance, so good luck getting that to work.

We have a couple castings of Lesser Planar Ally (summon things), which ignores spell resistance at least, but it's limited to 6 hitdice, and you can't get a specific creature unless you know the name.
    No, wait. With all that level drain, it's within the required range.
    And we *signed a contract with it, so we know its name too.*
 

So we can force another fight to start but that's not a winnable fight, or even stop it from just teleporting away again. Back to the drawing board.

We do know of a circle that held it once, but we don't know if that was due to the existence of a previous contract or due to the circle itself.
    But that doesn't matter.
    Because that circle is, conveniently, hooked up to a machine that can drain levels.

Scene 5: Finale
We fucking leg it over hill and dale back to the Dam as fast as possible; on arrival we are relieved to see that the fiend didn't think of the same plan first.

We only get one shot at this.

One of us stands next to the activation lever. Our divine caster faces the circle and casts for a tense 10 minutes - which is successful. The fiend is barely materialising inside the circle and doesn't even have time to realise what's going on before the activation lever is yanked down and he implodes into dust. Silence.

Did it work? Is it over?

As we understand it, "killed" devils will reform in Hell (like most outsiders). Only one way to know. We pull out another scroll of Lesser Planar Ally and summon it again. This time, a mere 1HD Lemure arrives. No rank, no status, no power, no memory. Confirming that it did indeed have the right name, we dismissed it back to Hell on the spot and ended the session.

The End!

This is probably my favourite game tale. There's no "Nat 20!" moment, just desperation and raw system mastery to take down something that was probably about as dangerous as the campaign's final boss and get a free Wish out of it on the way.

Fighter: Student of Many Styles

It's a fighter that develops in random directions, steered by the player.
This is not polished, it is minimum viable product. Only the barest thoughts have been given to balance. This might be overtuned. On the other hand, fighters should be good at fighting.
I think that you could take Parry, good with shields, the parry-improving technique, and wear heavy armour and you'll be REALLY hard to kill. I'm not sure how I feel about that.

Or if you got extra reach, cleave, and finished with Kill The Mooks you'd be able to mow down weaklings like anything.

I can see some anti-synergies between some things: if you punish misses by attacking more, and attach a maneuver to each of your attacks, so the enemy tries to hit you and misses EVEN MORE,for more attacks... then is it good or bad for you to improve your maneuver so that it doesn't provoke enemy attacks? Same thing with a crit build vs Kill The Mooks.

I also note that it's quite possible to create a fighter that only ever gets one attack per round. That sorta feels wrong. But maybe that's OK.

 In short: if something in this feels wrong for you, it probably is wrong, and should be adjusted.

Chassis

A: 2 Fundamentals, 1 Technique, Attack +1
B: 1 Technique, Attack +1, Extra skill
C: 1 Fundamental, 1 Technique, Students
D: 1 Secret Art, 1 Swappable Technique

You begin with an interesting scar, two weapons OR a weapon and shield, medium armour, and a tome of esoteric fighting techniques.
You are proficient with all normal weapons and armour.
Starting Skill: 1. Mercenary 2. Foreign Parts 3. Monsters

Fundamentals

Each time you get a Fundamental, pick an item on this list.

• Weapon Notches: Weapons get better at 3, 10, 30, and 100 kills. You can take the same thing more than once.
  
• +1 attack
• +1 damage
• Extra thing
• Keen edge
• Parry: Or you can call it Stance. You have [2, 3, 4, 5] hp of overshields each round.
• Tactic master: When you hit an enemy, you can also make a maneuver against them (does not remove the normal chance of being hit)
  
• Cleave: When you kill an enemy or crit, you can make an extra attack.
• Good with Shields: Extra +1 Armour from shields, can always sunder for maximum
• The Good Power Attack: When you power attack, the penalty is only -2 rather than -4.
  

Technique

Each time you get a new Technique, roll TWICE on the list, and pick ONE to gain. If you get an entry you already have, reroll that one. I kinda want to be able to increase this chart to a D66, so I need 3 more ideas.

1. Head of John: Your head becomes indestructible. It's a good weapon.
2. Pattram Sword Hand: Your unarmed strikes are just like swords. Work out the specifics.
3. Horse Stance: You no longer get tired from walking or running any distance as long as you are not overencumbered.
4. Chan Style: No item you use as a weapon does less than 1d8 hp and you don't take penalties just for trying to use it (though it may only be good for one use if fragile). If you have a strength penalty to attack/damage, that's also ignored.
5. Swat Missiles: You can smack incoming missiles out of the air if you make a successful attack roll.
6. Protecc but also Attacc: If an ally is attacked by an enemy, both within your reach, you may make an attack against that enemy first.
7. First strike: When an enemy moves within your reach, you may make an attack against them immediately.
8. Infinite Tumble: Move past enemies or disengage without provoking attacks
9. Discliple of Guts: You learn how to carry and effectively use ridiculously oversized weaponry.
10. Punish Incompetence: When an enemy misses you, you may make an attack against them.
11. Keen. Your critical range expands by one.
12. Spellsword: Roll a random spell and gain one MD. You're good with spells now.
13. Bloodlust: Gain HP when you kill an enemy, equal to how many HD they had. No "angry turtles in a bag" strategy, please.
14. Ignore Injury: Mere pain, shock, bruising, bleeding, cold, burns and the like no longer hampers your ability to act - only extensive damage like shattered bones or loss of limb.
15. Follow Through: When you kill a creature, damage exceeding their remaining hitpoints gets applied to other targets you could also conceivably attack with the same action.
16. Size up: Can tell most important enemy stats at a glance: hitpoints, armour, and their strategy.
17. Air Blade: Can make ranged attacks with melee weapons.
18. Implacable: You are immune to fear and get a +4 to saves against other mental effects. Allies near you get a +4 to saves against fear.
19. Extreme Parry: Double your parry.
20. Dash Master: Get a 10' Dash as the Skirmisher (that is, a 10' horizontal dash, 1/round, usable at any time including when an enemy is about to attack you)
    
21. Extra Quickdraw Slots: You get another 3, like a Thief
22. Lightfoot: Your footsteps are so light you can tread on water or snow without falling. +2 stealth.
23. Elemental Affinity: Pick an element. You can use it to perform 1d6 attacks and achieve other small-scale effects.
    
24. Elemental Resistance: Pick an element. You gain resistance to that element equal to your HD.
25. Danger Sense: You cannot be surprised.
    
26. Situational Extra Attack: Come up with some particular circumstance that gives you an extra attack.
    
27. Effective medicine: Gain the Medicine skill. Rolls to prevent someone dying from their injuries succeed on a 3 in 6.
28. Equipment trick: You can make better use of a kind of equipment, especially in combat. E.g. drinking things like potions can be done for free. Or you're good at using ropes are weapons.
29. Zoom Punch: Space seems to warp around you, your reach is much longer than usual.
    
30. Petition DM: Ask for a certain item on the list, or a feature you can find from another class, or something custom, of appropriate power. Try improving a technique you already have, or some set of circumstances that gives you an extra attack?

Other ideas: Extra HP. Reduce impact of Death Dice (make them d4s). Not being able to see doesn't negatively affect your ability to fight.

Attack +1

This is a +1 to your ATTACK ROLLS, not an extra attack.

Extra Skill

Gain an appropriate skill.
[To do: suggest some skills]

Students

A travelling student finds you, wishing to learn from your example. Treat them as a level 1 Student of Many Styles. After a week of learning from you, treat them as a level 2 Student of Many Styles; the technique they gain must be one you have. After a month, they will move on (and are soon replaced by a new student) unless they are furnished with a good reason to stay.
They are willing to EITHER pay you, work for you, or fight for you in exchange for teaching.

Swappable Technique

Pick a technique on the list, gain it immediately. You find this technique difficult; it runs against your deeply ingrained habits, and requires regular practice/meditation for it to remain usable.
You may swap this technique out for another in a week of downtime.

Secret Art

You gain one secret art: some ability of significant power, known to very few.  You might be the only living practicioner. You get a secret art that is most appropriate for your character - according to your needs or historical approach to solving problems. You and your DM can work out the specifics. Some examples are given.

• Ability to cut things that should not be cuttable, such as very hard materials, or space, or truth (pick ONE)
• Perfect Tactical Foresight: declare that the combat round that just happened was only "what you saw might happen". Everything gets put back as it was a round ago and we get a do-over. Nobody will be able to remember the exact hitpoint totals that anything had - close enough is good enough.
• Hadouken. A big one.
• Za Warudo. Two full rounds of action. Once a day.
• Kill the Mooks. Attacks against nameless mooks (2HD or less that the DM hasn't named) always hit and always kill.

Mutating Your Spells

Normal spell mutation incurs one roll on the main 2d20 chart. That chart may instruct you to roll on the subsequent two charts or roll additional times. For the purposes of tallying mutations, Good-Boring mutations are counted with their accompanying Drawback as a single mutation.

Spell Mutation Main Chart (roll 2d20, add them)

2. Roll on the Drawbacks table.
3. Death Burst: When you die, you may cast this spell immediately with extra MD equal to your amount of wizard-equivalent templates x 4.
4. Inevitable: If you have in your possession something belonging to your target, you can burn it to target them from infinite range without line of sight, the magic will slip between the tectonic plates, crack the barriers between worlds to reach their target. It is impossible to hide from the spell, but it may be stopped by adequate protection.
The target knows who hunts them and where the spell was cast.
5. Gorgon: After a gorgon spell is cast, the invested MD are occupied until a creature sees the caster, at which point they are immediately made the target of the spell, range disregarded.
6. Discovery: When cast, the caster learns one non-obvious thing about the target.
7. Transportation: 1-3: The caster is teleported to the target of the spell, after its initial effects have been resolved. 4-6: Target teleported to caster.
8. Trigger: Whenever you cast this spell, define a condition that also defines a potential target should that condition occur. Lastly, select an inanimate object no larger than an elephant, the spell is bound to that object and will be cast from that object as soon as the condition holds true. The invested MD are occupied, unless you use 25 GP’s worth of magical ingredients per invested MD.
9. Spellborn Homunculus: Your spell becomes a person.  They are neutral towards you. Roll a random personality and goals. They can cast themselves 1/day. They start with 1 MD, and can gain more.
10. Cross-pollination: The next time it is cast, the spell permanently incorporates some aspect of 1. The caster 2. A random spell in/on the caster 3. The target 4. A random spell in/on the target 5. The next thing to touch the spell after the target. 6. The most powerful creature in the area.
11. Peanut Gallery: All players present collectively come up with something new and fun.
    * "The spell is lured by the smell of chocolate"
    * "The spell gives good artistic advice when memorised"
    Go with something that makes the GM or the caster laugh or scream.
12. Phase: Phase out of space and time for [dice] turns after you cast the spell, but before it resolves.
13. Curse: same as 8, but can only target a creature instead of an inanimate object.
14. New Flavor:  The spell changes elements or orientations.  Fireball becomes lightningball, protection from evil becomes protection from good, charm person becomes charm bird (or perhaps infuriate person), web becomes freedom of movement.
15. Counter: If a spell is cast within sight, you can immediately cast this spell at that target as a counter before it resolves.
16. Wild: Whenever you cast the spell, roll on this table and use that mutation for this one cast. (If this introduces too many extra rolls, replace it with Protean: change the mutation at each dawn.)
17. Breeding:  Your spell has just given birth.  This newborn spell is a weaker version of its parent, but after being cast 5 times it will have grown to adulthood.  Roll a d100 to see if it has mutated from it's parent breed (the same as encountering a new spell in the wild). This is not a permament mutation applied to the parent, but more of a singular event.
18. Ritual: If the caster spends 10 minutes casting the spell, the spell can either be cast with an extra MD or with a recovery range improved by 1, your choice.
19. Roll once each on the Good-Boring Table and Drawback Table
20. Roll again twice
21. Roll again twice
22. Roll again twice
23. Roll once each on the Good-Boring Table and Drawback Table
24. Proxy: You may choose to not cast the spell as normal, instead you invest MD as normal and target a creature to allow them to cast the spell instead of you once with all the invested MD. You can reclaim the MD at any point, but until you do, they count as occupied.
25. Cooperation: If other casters close hands with the Spellhost, they may pool their MD together for the Spellhost’s magic. In addition, each participant adds +1 to [sum], even non-casters.
26. Chain: After resolving the spell, if a miscast or doom occurs; cast the spell again with the duplicate MD’s for free on a new creature within sight.
27. Delayed: The spell can be cast and at a later date be unleashed as a free action, while doing so it occupies the invested MD. Multiple instances of the same spell cannot be delayed.
28. Recursive: You do not regain MD when casting this spell, instead any recovered MD are immediately spent on recasting the spell on the same target, repeating this effect until no MD remain.
29. Circumference: As the spell is cast, the caster is protected by a circle of visible energy drawn 5 feet around the caster, any creature that steps through the circle (including the caster) will have the spell re-cast at them, with the amount of previously invested MD added for free, after which the circle dissipates. The circle lasts for [dice] turns.
30. Blood: You can sacrifice 5 of your own HP to add up to one MD to the casting, if this MD resulted in recovery, regain HP equal to the dice’s result.
31. Full Sentience:  Your spell becomes fully sentient.  It talks to you.  Roll a random starting personality and goals.  If it is angry at you, it may refuse to be cast.  If it is especially pleased, it may enhance itself in a way that you request.  Spells enjoy being in your brain and seeing out your eyes.  They do not enjoy being in the spellbook, which is much like a jail.  They may request (or plot) their release, or the release of all your spells.
32. Familiar: Whenever the spell is cast, it manifests instead as a 1HD familiar with a shape embodying the spell’s theme, the spent MD are occupied until the familiar is destroyed, at which point the spell is cast. The familiar can be commanded to destroy itself. You can only have one active familiar per spell at a time.
33. Tribute: The spell can be cast with an extra MD as long as a sentient creature is sacrificed as part of its casting. The sacrifice has to be ritualistic and performed on a helpless target, just sacrificing an active combatant is not possible.
34. Soul: If the spell kills a creature with 1 HD or more, choose to recover or replace one MD. As long as that MD remains unused, you can communicate with their soul.
35. Name: If you can name your target, add an extra MD.
36. Necro: The spell can target corpses with an intact soul (dissipates after 13 minutes) and use it to fuel their spell. Each 2 HD or 1 level of the target creature allows you to add an extra MD, failed recovery rolls on the extra MDs causes a 3 necrotic damage backlash.
37. Cantrip: If you cast this spell investing only one MD, the MD always recovers.
38. Mimic: Before casting, this mutation can assume the effects of any mutation in one of your memorised spells.
39. Powerful: MD’s invested in this spell are rolled as D8’s, their recovery range is 1-4. This is the Philosopher’s Stone of Spellhosts, coveted by all and ruthlessly hunted.
40. Pick any entry on this chart of your choosing. If that entry has a random/indeterminate result, you can probably pick that too.

(2d6) Good-Boring Table

2. Hard:  Targets get -4 to their Save.
3. Subtle:  You can cast this spell quietly and without being noticed.
4. Improved Effect: The spell does more of whatever it does.  5d6 damage becomes 6d6, charm becomes obsession.
5. Improved Reach: Triple the range of the spell. If it was Touch, it becomes 30'.
6. Improved Area of Effect:  The spell affects a wider area, or more targets.
7. Negotiate! GM and players think of something not game-breaking but cool, including fusion. Or you can use this to get rid of a downside to the spell.
8. Improved Finesse.  You have more control over your spell.  You can create gaps in your fireball, or limit the damage, for example.
9. Improved Duration. Triple the duration.
10. Improved Applicability.  The spell affects a broader category of targets.  Charm person becomes charm biped.
11. Reliable: Roll an additional MD, and then exclude any single MD from resolving the effects. This can be used to avert mishaps/dooms.
12. Facile.  You can cast the spell as a free action.
 

(2d8) Drawbacks Table

2. Cancer: Reduces the amount of MD cast with this spell by one. Cancer cannot be removed or re-rolled through Spell Breeding, but it can be swapped to another spell, in which there is a 50% chance the cancer will duplicate and exist in both swapped spells.
3. Enemy of Farmers: Every time this spell is cast the nearest unborn animal rolls on the Physical Mutation chart.
4. Gadabout: Spell has a 50% chance of being present when wanted. Otherwise it's faffing about someplace.
5. Jealous: If the spell is the first spell you cast today, get the benefits of Reliable ONCE for the day. If not, the GM may force a reroll of any single MD invested in each cast for the day.
6. Clingy: The spell will not willingly relocate out of your head. Burn a MD to try to force it out, and Save to succeed.
7. Worsen: Pick a random numerical attribute of the spell and halve it.
8. Calling Card: The spell leaves very obvious and long-lasting evidence that it was cast, usually thematically particular to the associated caster when this mutation was gained.
9. Specificity: Spell's Target becomes more specific. "Charm Monster" becomes "Charm Smelly Monster" for example.
10. Center of Attention: The spell now gains a great deal of pyrotechnics and noise. While not necessarily damaging to be nearby, it is almost impossible to not notice.
11. Touch: The spell can only be cast while touching the target, but if you do, add one additional MD for free.
12. Finale: When this spell is cast, include 2 free MD. However, any available MD that the caster did NOT invest in this spell are rolled and the sum taken as damage by the caster.
13. Aphasia: While memorised, caster cannot verbalise anything except the spell's name or subsets thereof.
14. Parasitic: The spell is less choosy about where it lives, and progressively alters a host more to its liking. This is not healthy for the caster in the long-term, but they may be able to glean some benefit.

Spell: Mutate Spell

R: Touch (including in head); T: A spell;
Material components: 1gp of bribes and anaesthetic per MD invested
Effect: Target spell is mutated according to the result of a roll on the Mutate Spell chart. You must invest a minimum number of MD equal to the number of mutations the spell already has.

If you spend 100gp per MD invested (minimum 100gp), you may roll twice on the chart and pick the result you want.

Hybrid levelling system: XP for Treasures

What you award XP for has a massive effect. It's one of the best controls you have to change what goes on in your game; players will naturally seek out whatever action awards it, and you can switch out that thing (or selection of things) without having to think too hard about what else will be impacted by that change.

If you award XP mainly for killing monsters, players will begin to prefer plans that include doing that. If you award XP just for getting the gold out of the dungeon, then you're free to design in a space where all methods of getting past any obstacles between the PCs and the gold are equally valid. I've seen discussions about awarding XP for investing gold into settlements; the overall progress of a campaign is then inextricably reflected in the growth of their home base. You can award increasing XP for exploring a sequence of dungeon rooms without rest. And so on and so forth.

Of course, milestone levelling, i.e. just levelling up players when it feels right, is an option. However, that doesn't work so well for games where PCs may have different levels (as mine does), and I want to take advantage of having an obvious universal carrot for my players to chase after.

XP-for-gold is good, but I'm increasingly finding that I want to be free of the burden of math in my games (related: migrate to usage dice or any other method to avoid tracking exact counts of things), as my players are rarely as happy to whip out an excel spreadsheet and keep strict notes as I am - and I certainly have too much to do to want to do that for them. I also have an interesting conflict where I want mundane purchasing decisions to remain a part of the game beyond level 2-3, and for the first big chest of gold to NOT rocket a single PC to level 10.

I think I now have a way out of this problem. I was already familiar with the large treasures of Ultraviolet Grasslands, and this week I read Arnold's post on Popcorn Levelling, which are both very interesting to me. It is these two things that I have blended together into a new system, which works as follows.

• You need Treasures (that's a keyword) to level up
• Treasures are big, fancy, named things
• They usually have history and artistic value; being magical and having a function is more optional
  
• Treasures are immediately recognisable as such! These are an obvious carrot, remember!
  
• A Treasure takes up one* inventory slot, is worth 1XP, and is worth 500 silver* for every WORD in the name
• When you carry a Treasure out of the dungeon (or can be said to have "gotten away with it") you can have the XP
• Keeping a record of all the names of the things you've thus stolen as your XP tracker is highly recommended
• Levelling is a Fibonacci progression: 2XP for level 2, an additional 3XP for 3, an additional 5XP for four, 8XP for five, and so on
• Selling the Treasure, or keeping it, or throwing it into a lake, is up to you. You get the XP either way.
  

Currently, the XP for a single treasure cannot be subdivided, and there is no guidance for how treasures should be assigned to PCs when obtaining them is usually a team effort. At the moment I'm letting my players sort it out amongst themselves, and if that starts giving bad results I'll implement an extra rule or two. I mainly just want to discourage putting all the treasures onto one PC.

It might not be reasonable to know the "true" name of a Treasure when it's first encountered; in that case a purely descriptive name (of the same number of words) might be temporarily assigned, but identifying it properly will improve things somehow. It might be able to be sold for more (or it cannot be sold for the normal amount while unidentified), gain new/improved functions, become a plot hook, be usable as a plot token elsewhere, etc.

For example, you might obtain the AERIAL PHOTOPLATE ATLAS, TURBO ENCABULATOR, and ARTICULATED MODEL DRAGON; carrying all of them at once will take up eight inventory slots, they are worth a respectable 4000 silver, a single level-1 PC that managed to make off with the lot would jump to level 3, and only need to steal two additional WORDS of Treasure to hit level 4.

This whole thing conceptually echoes the GLoG principle that you should gain abilities that reflect your adventures; now your XP tracker will reflect what you've stolen and be a nice little summary history of your successes and a reminder of where you've been. The character sheet is a living document, steeped in adventuring history.

I've now deployed this in my game and I'll get my first set of results next session. Tell me if you try it for yourself, or just say what your preferred XP system is (and why)!

* To help you set expectations with the above numbers: a warhorse, suit of plate armour, or Fireball spell would all be worth about 1000 silver, and a skilled mason can comfortably save 4 silver each week (actual income and cost-of-living are more than that). Conceptualise a silver as being worth $10 in modern money and you won't be far wrong. PCs have their strength score in inventory slots. I'm also working with the Goblin Laws of Gaming, where you stop getting class features at level 4, which is about your life expectancy anyway.