Tuesday, March 11, 2014

Review: Zzarchov Kowalski's Scenic Dunnsmouth (PDF)

Scenic Dunnsmouth is great. Get it.

A bit more detail:

Dunnsmouth is a delightfully creepy swamp hamlet, which is randomly generated using some dice, a deck of cards, and a piece of paper. Therefore, you're effectively getting thousands of possible towns for your buck, most of which revolve around a mutant Spider Cult, and all of which will include: the reality-warping Time Cube, a Roma witch, a cannibal serial killer with a thing for bear traps and rock falls, and backwoods incest, rivalries, murder, and sin. It has an amazingly pulpy, '40s Radio Horror Adventure feel, and a lot of possibilities. The scenario is geared for a 3rd or 4th level party, or maybe a bit higher.

Organization

The module starts off with a description of Dunsmouth, three different ways you might draw a party into the swamp, the two major elements of the module, and a description of how to generate the town. It's all easy reading, and town generation is very straightforward - I generated three towns in 50 minutes, and if I was about to run it it would probably take about that long to generate a single town, print out all the NPCs that appeared, and be ready to go. The whole thing really boils down to three or four major steps - roll the dice, draw the cards, figure out which NPCs are present and where, and tidy up. But Zzarchov splits it up into fourteen crystal clear mini-steps, which is very helpful, and prevents you from having to flip back and forth across scores of pages to figure everything out.

The next part discusses the most important locations in town. Some of these are guaranteed to appear, though their status (and sanity) depends on how the dice came out. There are also two "kickers," each of which have a 50% chance of being a special location - so Dunnsmouth could play host to one or two elves (one of which is, to my mind at least, one of the top 3 most horrifying NPC's in the book), or perhaps an old fortification, or possibly some light industry.

After this comes the largest part of the book. Dunnsmouth is inhabited by up to four families, one for each suit in the deck of cards, with each card representing one household. So this book nets you 52 households, with multiple NPC's in each one, only 10-12 of which will actually appear in the game. The presence or absence of certain individuals and the relative strength of each family can provide a lot of fuel for blood feuds and drama, even if the Spider Cult doesn't make an appearance.

Finally, we have an appendix and index of sorts, with references for some of the items and spider-mutants that might be found, and a helpful step-by-step example of town generation.

Strengths

I think my favorite part of the module is how the way each location is changed by the NPCs inhabiting it, or how each household's behavior & characteristics are altered by the spider mutation, gives you much better insights into their psychology than a simple description does. Ivanovik and Magda are potentially located at any one of the ten special locations, and in addition to that, have several possible homesteads they might build if not. Reading about how they (and the Original Spider) set up and fortify each location, reacting to its unique characteristics, more than makes up for the scant direct information we're given  about who they are. The infection does the same for the rest of the NPCs - simmering resentments and hidden attractions are brought to the fore by the spider gene's quest to reproduce and dominate. It's showing, not telling.

I also liked (as I mentioned before) the replayability and variety built into the module. You could keep running it, with a different setup each time, or you could mine out its stockpile of NPCs and use them to spice up other, less distinctive villages. There are a few "joke" NPCs in the book (used if you fail to completely clean the deck), which you could always throw in by choice if you wanted to move from horror to Bizarro. (Although the Black Joker, Jesse McLaud, didn't strike me as that bizarre, and I left him in the deck). I think this aspect, makes this the perfect module for a new DM to run - it's like training wheels, but still puts you through the paces of personalizing a scenario and forces you to think through it beforehand, rather than blindly assuming it's all ready to go.

Finally, the art by Jez Gordon is amazing. It evokes a delightful, pulpy-horror feel and really helps conjure the darkness and fear that's found in Dunnsmouth. The character portraits in particular are very well-done, in most cases conveying just as much information as the character descriptions themselves. They also make great icons for player maps (as I've done below) so you can show them who lives in town without immediately giving away any information about family allegiances or ranks.

Weaknesses

Since this is my first review, I should state that I'm not always looking for the same thing out of every module, so of course everything's going to be pretty subjective. My main critique is that, for all the characters this module has, I felt like only a handful of them actually had character. Almost all of them are pretty cartoonish backwoods stereotypes of one type or another. I don't think this will come up in the course of running the actual module, because you'll only have a dozen households and it's clearly intended to be a bunch of cartoonish backwoods stereotypes - but after the thirtieth straight page, it got exhausting. Few of them are actually compelling enough to draw a party into their struggles and feuds, Spider Cult or no. Because of this, I don't think Dunnsmouth lives up to its promise of "moral peril." Maybe it's just my Utilitarian ethics, but there's really no character sympathetic enough, or interesting enough, to give me much pause. The module wants you to be forced to decide, what am I willing to do to save these people? But I'm stuck on whether I want to save any of them at all, aside from a generic "well these people have children!" element.

Suggestions/Recommendations

The full text of the actual Time Cube Treatise is available online. I just love that this turned up in an RPG somewhere. Needless to say, my Dunnsmouth will have 96-hour days.

Always offer your players the chance to purchase Dunnsmouth's debt at the start of the module. It's my favorite, by far, of the three "leads" - it gets your party right up in the villager's faces, and I'm pretty sure most versions won't actually be able to scrape up the cash.

Though there's a certain romance to the idea of rolling the dice and rolling with the result, I think it's worth generating two towns, and picking whichever one seems more interesting at a glance. That way, you can take the second town, remove duplicate characters and generally align it with the "Expanding Dunnsmouth" section at the back, and add a bit of heft to the module if you need to.

Duncasters? A secret shame? A defunct cult with a mountain headquarters? Zzarchov's practically yelling at you to put Death Frost Doom in the mountains next to the swamp.

Maybe you don't tell players about the Time Cube. Maybe you just draw concentric circles around it on the player map, mark down the slowdown ratios, and let them figure out the distortion themselves.

Town Sampler

So I whipped up player and referee maps for each of the towns I rolled, just to show a bit of the variety you can get from the module.


Scenic Dunnsmouths

Friday, March 7, 2014

The Tools of Others

Like most people, I use a ton of other people's ideas and tools to make my games run better, or just to get a bit of inspiration. So, here are a few tools and compilations I've made recently to facilitate my use of other people's tools.

First up, since it's currently trending, is +Arnold K.'s Career Paths character generation. If you haven't seen it, look at it - it's a cool hack of regular 3d6-based attribute generation that creates a bit of vague history, and it's a lot more elegant than some of the crazy Excel-formula-based versions I've seen, and a lot less involved than Beyond the Wall's analogous approach.


https://drive.google.com/file/d/0B4LyPp6ko4hJZjBKMTh3THRmVE0/edit?usp=sharing
Click me!
What I did was make a bunch of cards to make the Statistically Anal method of generating your character's adolescence significantly less anal. Instead of rolling a d15, you just cut this out, place the cards face down, and everyone picks one. I used a bunch of Telecanter's silhouettes on the reverse side to add a bit of Rorschach-ish psychology to card selection. A few questions are edited a bit here and there so they could all come out the same size.

I didn't do the Careers section mostly to save time and to leave starting PC's with a bit more blank space. So, instead of rolling 1d6 for each stat during Adolescence, I'll have players roll 2d6.


https://drive.google.com/file/d/0B4LyPp6ko4hJTWNVUENSc0ViZm8/edit?usp=sharing
Right here.


Next is the Peddlers of the Deep Dark table from Aeons and Auguries. Basically, it's a pretty cool (if die-roll-intensive) underground trade caravan generator. Unfortunately, bits and pieces of it are scattered all over JD Jarvis' blog, so I put it all in one place.

I included all of the Peddlers info, the Dungeoneer's Cache generator, and information on all of the unique items in both tables - that includes the Magic Ropes, Candles, and Footwear posts. Plus a bit of organizational tweaking.


Finally, Justin Alexander's Node-based Design has been pretty much the guide for running my campaigns, and the Three Clue Rule a great tool for coming up with plots on the fly, in a way that the players can meaningfully interact with. But, as I've mentioned before, I'm a visual thinker, and outline's just don't do it for me - so I set up two on-the-fly node sheets, for the Layer Cake and Loop models. Each node has its clues listed, with a bit of space to describe the node, and each of the clues leading to the other nodes.

https://drive.google.com/file/d/0B4LyPp6ko4hJS3dwZllyNEJpRWs/edit?usp=sharing
You know the drill. The Layer Cake one is here.

Thursday, February 6, 2014

Reaction-Roll Stealth

So far, the skill-based stealth system I've been using has been difficult to run without a battlemat, while the default "X on 1d6 to surprise" system still strikes me as pithy and undramatic (except in the simple case of encounters). But the reaction roll, which I've been using extensively through On the Non-Player Character, fits the bill perfectly - instead of binary success/failure, you get piling bonuses which slowly push you towards success or failure. Additionally, this is consistent with the current systemic division I'm contemplating - using d20s for the combat system, 2d6 for noncombat rolls, and dice pools for anything to do with magic. The following stealth system is mostly inspired by On the Non-Player character, and a bit from this Google+ discussion.

Stealth actions are resolved in segments of one minute, rolling 2d6 (plus Dex mod) each time. Bonuses are persistent, and cumulative. Consult the following table:

SneakBackstab
2Found or -4You're surprised!
3 - 5-2No surprise
6 - 80You surprise!
9 - 11+2Ranged backstab!
12+4 or HiddenMelee backstab!

Sneaking is used for any actions you take while undetected. The following bonuses apply:

No movement or other actions (hiding): +4
Combining two actions (moving twice, moving while tracking, etc.) -4 per action
Trickster/thief/etc. +level
Distance: +1 per 100' from nearest sentry/enemy, -1 per 10' closer than 100'
Terrain/weather/etc.: +2 or -2 depending on conditions

Found means that stealth ends (for you) and you're detected. Start encounter as normal - stealth can no longer be used. You can, of course, try to hide again - you can roll on the Sneak table instead of taking another action each round, with cumulative bonuses. Stealth starts again once you get Hidden.

Backstabbing is what you roll when you're trying to end stealth and initiate combat. On a 2, you totally bungle it, and your opponent gets to act against you in the surprise round, while on a 6 or higher, you surprise them, getting a free round. On a 9 or higher, you've spent enough time observing and positioning yourself to get a sweet sniper shot on an opponent, gaining a +4 to hit. On a 12, you're close enough to do it with a melee weapon, and can precisely target weak points or chinks in your opponent's armor - you get a +4 bonus to hit, and deal double damage.

Backstab bonuses:
Multiple actions (opening a door and then backstabbing, for example): -4 per action
Trickster/thief/etc.: +level
Distance: -1 per 100' from target, +1 per 10' closer than 100'
Terrain/weather/etc.: +2 or -2 depending on conditions

Wednesday, January 15, 2014

Lineup 2014

It's a new year, and I've resolved to get some regularly scheduled programming up on here. There are a few small side projects I've been working on, and my Qelong campaign is starting to pick up steam, but 2014 is really going to be held up by a few tentpole projects. In order:

The province of Acacia has been the only safe domain on the New Continent for over a decade, enriching its citizens and its rulers through a steady trade in wondrous furs, exotic herbs, and tough lumber. Now, however, the mercenaries of the Union Charter have set up competing forts, and inciting the natives to make war on the Acacians and on each other. As the balance of power whirls out of control, once-human things stir in the darkness, and the prehistoric forests begin to reveal exactly how they can stand the test of time...

For centuries, the Dark Continent has lain in state. Exoticized, exploited, or ignored by oceangoing imperiums, the Dark Continent is known as exactly that - a broad curlicue of unknown land, a rudimentary coast, a mockery of both cartographic authority and royal power. But there will soon be bands of human adventurers spilling across the Dark Continent's baking hills and steaming jungles, erecting coastal forts and way stations, penetrating deep into the ancient Elven homelands and the tropical cradle of Dwarven civilization. Will your band be swallowed up by its trackless hinterlands, or will they stumble to the coast, heavy with looted gold?

It's 1931 in the Tennessee Valley. Herbert Hoover is President, lending his name to shantytowns all across America, and Henry Hollis Horton is Governor, embroiled in scandals following the bankruptcy of his political allies, and the loss of $6,000,000 in corrupt endeavors. The law is largely in its own hands, and nobody important is interested in investigating those "haunted" woods up the river, or the "cursed" treasure hidden deep within a local cave...

Forgottenland is what happened in my head when the Old School Renaissance got run over by the Drive-By Truckers. Outside the Valley, society grinds on much as we know it. Inside the Valley, crooked cops hunt backwoods moonshiners, lynch mobs ensure white rule with institutionalized terrorism, and mythical beasts guard buried treasure in shadowed forests.

Veil

The Faith rules the world, and God rules the Faith. But who can ever know how God wishes to be served?

Veil is both my main campaign setting and what I'll call the gaming system I'm currently using. This is a grittier adventure fantasy, where rewards are swift, but death is swifter, clinging to your heels like an ancient cobweb.

Influences

I get vibes from a lot of different things, and I like to write about them. I'll review anything that has something I'd like imprinted into my projects, and put it here.

Monday, December 16, 2013

Horses, Dogs, and Falcons


Inspired by Tower of the Archmage's post on dog breeds, and designed to run with my basic system for combat and overland movement. I wanted to set up a system for animal ownership that reflected the same ideas of simplified accuracy, and gave each animal breed a bit of personality.

Animals

A variety of domesticated animals are available to assist the intrepid adventurer, merchant, or soldier. Mainstays are equines, dogs, and falcons. Animal prices are expressed in silvers per HP – healthier and tougher animals, of course, being worth more. Most animals consume animal rations, which weigh 1 stone and cost 1 silver piece, but dogs and falcons can both eat human iron rations.
All animals gain a +2 morale bonus when their master is present. A new master must be designated by spending a full month with the animal. Additionally, animals have their own reaction rolls toward NPCs, and must pass a Morale check to avoid attacking or fleeing on a 2, or in order to attack on a 12. They gain XP as normal and level up with the noted effects. Leveling also doubles their value. Animals must pass a Morale check to do anything other than fight or move with their master, with a -1 penalty applied for each word after the first.
Animal base saves are all 16 by default, though their base Might save increases by 1 for every 10 hit points they have. They cannot normally improve their saves.
Most animals have marching speeds equal to humans, but begin Forced Marches on their second march, rather than their third. Horses have marching speeds of 24/18/8 miles per Watch. Animals can be pastured for a full Watch, which in favorable environments fills 1/3 of their upkeep needs.

Equines
Horses are usually sold at 4 years, and begin aging at 20 years, penalized every two years. Mules and donkeys begin aging at 30 years, penalized every three years. Failed saves reduce HP by 2, encumbrance by 1, and running/sprinting speed by 2’. Horses can jump 4’ obstacles with a Will save, -1 per 6” of extra height. Failure means a 50/50 chance of shying away (with the rider passing a STR check or falling off) or crashing into the obstacle, inflicting 1d10 damage on both horse and rider. 
DRAFT HORSE
Phlegmatic, thickset horses for plowing or milling.

Price (per HP): 20sp
HP: 5d6
Encumbrance: 8/16/32 stone
Attack: 1@1d8
Defense: 8, Large Size
Morale: 6
March Speed: 4/8/12
Run Speed: 60’
Special: -
Upkeep: 3 animal rations, 15 water rations
ROUNCEY
A basic light riding horse or warhorse.

Price (per HP): 20sp
HP: 5d6
Encumbrance: 7/13/26 stone
Attack: 1@1d8
Defense: 10, Large Size
Morale: 7
March Speed: 8/16/24
Run Speed: 70’
Special: -
Upkeep: 2 animal rations, 15 water rations
PALFREY
Comfortable, easy-to-control riding horses.

Price (per HP): 20sp
HP: 5d6
Encumbrance: 7/13/26 stone
Attack: 1@1d6
Defense: 8, Large Size
Morale: 4
March Speed: 7/13/26
Run Speed: 60’
Special: All mastering & training only takes 1 month
Upkeep: 2 animal rations, 15 water rations
CHARGER
Fast cavalry horses, the most common type in war.

Price (per HP): 30sp
HP: 5d6
Encumbrance: 8/16/32 stone
Attack: 1@1d10
Defense: 14, Large Size
Morale: 8
March Speed: 8/16/24
Run Speed: 70’
Special: -
Upkeep: 3 animal rations, 20 water rations
JENNET
A small, quick desert horse, with low upkeep.

Price (per HP): 30sp
HP: 5d6
Encumbrance: 7/13/26 stone
Attack: 1@1d6
Defense: 14, Large Size
Morale: 8
March Speed: 8/16/32
Run Speed: 80’
Special: -
Upkeep: 2 animal rations, 12 water rations
DESTRIER
Heavy, highly prized battle horses.

Price (per HP): 90sp
HP: 6d6
Encumbrance: 8/16/32 stone
Attack: 1@1d10, 1@1d6
Defense: 14, Large Size
Morale: 8
March Speed: 4/8/12
Run Speed: 70’
Special: -
Upkeep: 3 animal rations, 20 water rations
MULE
A mix between a horse and a donkey.

Price (per HP): 15sp
HP: 4d6
Encumbrance: 8/15/30 stone
Attack: 1@1d6
Defense: 10, Large Size
Morale: 7
March Speed: 4/8/12
Run Speed: 60’
Upkeep: 1 animal ration, 12 water rations
DONKEY
Light, clever pack animals.

Price (per HP): 10sp
HP: 4d6
Encumbrance: 4/8/16 stone
Attack: 1@1d6
Defense: D10
Morale: 5
March Speed: 4/8/12
Run Speed: 70’
Upkeep: ½ animal ration, 8 water rations

Birds

Birds are sold at 2 years, and begin aging at 10 years. Failed saves reduce HP and Attack Bonus by 1. Birds fly and soar instead of sprinting and running, respectively, and must fly before soaring. They all have encumbrance of 1. Leveled birds increase their defenses and attack bonus by 1, and their fly speed by 4’.

FALCON
Price (per HP): 10sp
HP: 1d6
Attack: 1@1d6
Defense: D15, Ranged 18
Morale: 6
Fly Speed: 150’
Upkeep: 1 meat ration, 1 water ration
Special: +4 attack bonus while diving
RAVEN
Price (per HP): 10sp
HP: 1d4
Attack: 1@1d4
Defense: D15, Ranged 22
Morale: 6
Fly Speed: 150’
Upkeep: -
Special: Small size, +2 Will & Wile saves
Dogs
Dogs Run and Sprint at 50’ rather than 40’, and have Swimming at 30’.  All have upkeep of 1 meat and 1 water ration, and encumbrance of 3 stone (though objects weigh triple for them). Dogs with 2d6 HP weigh 8 stone, and with 2d4 HP, 4 stone. Dogs are usually purchased at 2 years old, and must begin making Will saves against the effects of aging at 10 years, penalized each passing year. Failed saves reduce HP by 1, and running/sprinting speed by 2’.

GUARD DOG
Mastiffs, Rottweilers, Dobermans, German Shepards, & Bulldogs.
Big, strong, relatively intelligent.

Price (per HP): 4sp
HP: 2d6
Attack: 1@1d8, +1 bonus
Defense: 12
Morale: 7
Special: +1 encumbrance
Leveling: +1 HP, Might, AB, & Melee Defense

SHEEPDOG
Collies, Shepards, even Corgis and Dalmatians. Smaller, faster, smarter. Independent.

Price (per HP): 3sp
HP: 2d4
Attack: 1@1d6
Defense: 12, Missile 13
Morale: 7
Special: Wile save vs. Surprise
Leveling: +1 HP, Run speed, Wile

BIRD DOG
Setters, pointers, retrievers, even early poodles. Loyal, energetic, but disciplined.

Price (per HP): 4sp
HP: 2d4
Attack: 1@1d6
Defense: 13
Morale: 7
Special: Swimming 60’
Leveling: +30’ Swimming, +1 HP, Morale


HOUND
Hunting and tracking dog, for flushing out game. Best sense of smell.

Price (per HP): 4sp
HP: 2d6
Attack: 1@1d6
Defense: 12
Morale: 7
Special: Examination, per the skill
Leveling: +1 HP, AB, Examination

GREYHOUND
Includes Whippets and other racing/sight hunting dogs.

Price (per HP): 5sp
HP: 2d6
Attack: 1@1d6
Defense: 13
Morale: 7
Special: Run/Sprint 60’
Leveling: +1 HP, AB, & +2’ Run Speed

TERRIER
Small but stubborn. Vermin hunters.

Price (per HP): 3sp
HP: 2d4
Attack: 1@1d6
Defense: 12, Missile 13, Light Armor
Morale: 7
Leveling: +1 HP, Melee Defense, Attack Bonus

LAPDOG
Luxury, toy dogs for rich people.

Price (per HP): 5gp
HP: 1d4
Attack: 1@1
Defense: 12
Morale: 7
Special: Very fancy, weigh 1 stone
Leveling: +1 HP & triple value

Monday, December 2, 2013

Augmenting Age

Character age is at once an actually pretty important attribute and also one that gets pretty half-assed. Who doesn't usually glance at the field and pick a random number between twenty and thirty? 

Obviously, nobody thinks about the choice because the choice isn't made to matter very much, except in the extreme long-term of avoiding the aging penalties you receive at 50. For this reason, there's no mechanical sense in choosing any age other than the minimum the GM lets you get away with. But I don't think it's worth figuring out some sort of life events table that scales with your age and you have to make an excel spreadsheet just to calculate.

So, money. How about one extra copper at character generation for each of your first twenty years? And, one silver piece for each year after, till you're 50, then you get 1 GP per year. Age becomes a little bit of a Faustian bargain - more immediate power right now, during character generation, in return for bringing dementia and incontinence just a little bit closer. 

Is it balanced? Using my 1:10:40 GP:SP:CP standard, a 55 year old character would effectively get 85+3d6x10 silvers at character generation. Probably not enough to balance out the chance of losing five points of attributes. But maybe enough to merit some thought about it.

Sunday, December 1, 2013

Calibrating This Shit

As Charles Taylor says, realism in the game provides a vital frame of reference for the player to be able to predict the effects of their actions, and, more importantly, put the unrealistic elements in their proper context. I'd rather have a few rules, tightly webbed and absolutely realistic, than a whole lot of quasi-fantastical and independent systems.
Finally, I have an hour and a half commute every morning, so I have little else to do but sit on the bus and consider the internal consistency of my system. So I calibrate shit.

Walking

Let's start with walking. In a combat round of six seconds, the average character can take four actions. Walking 5 feet is one action. Walking is, of course, something every player will be familiar with, so it had better work the way we expect it to.

So, you'll walk 20' in six seconds, or 200' per minute. This gets you to 5,280', a full mile, in 26.4 minutes - or about 2.3 miles per hour. This is a tad on the slow end - average human walking speed is 3 mph - but it comports with marching speed, which is a bit over 2 miles per hour over varied terrain. It's a nice, leisurely stroll - or an aware, steady advance. Realistic.

My marching speeds for unencumbered/encumbered/overencumbered people are currently 4/2/1 miles per hour, respectively - putting combat walking in the "encumbered" category. To be fair, 2 miles per hour versus 2.3 miles per hour adds up to a loss of about two and a half miles over the course of a day - but cross-country marching isn't a completely continuous slog. Internally consistent.

Running and Sprinting

Time to take it up a notch. The average human can run at 5-8 miles per hour, and sprint at around twelve to fourteen. Originally, I had running take 1d4 actions for 25' of movement, but that put "running" speed at a measly 4.5 miles per hour. I'll have to balance running speeds (1d4 actions), sprinting speeds (2d4 take lowest), and top speed (which is when every roll comes up 1).

Before we begin, I'd like to mention the rationale for a roll-to-sprint system: The uncertainty it introduces really allows me to run chases and pursuits using the combat system, pretty much unchanged, and I think it does a good job of reflecting the importance of reflexes, momentum, balance, and footing in short-distance running.

The average "run" will use up 2.5 actions, which allows for 16 runs per minute. At 40' per run, we get 640' run per minute, netting you a mile in eight minutes, fifteen seconds. This gets you 7.27 miles per hour - close to the upper average! Ideally I'd want it a bit closer to the 6 MPH mark, but it's not really a big deal.

The average of 2d4, take lowest, is 1.875 (about a 25% action cost reduction). That's 21 and 1/3 "sprints" per minute, or a 33% speed increase, for 9.7 MPH. Admittedly, this isn't a huge leap ahead of running speed, but you'll be getting top speed 80% of the time, rather than 25% of the time, which is a big advantage over short distances - exactly the balance I'd want to have.

Top speed means that you get four moves of 40', or 160' a round, 1600' per minute. That's a 3.3 minute mile! You're running at 18 miles an hour, mate. Of course, to run even a third of a mile at top speed, you'll need to roll 1 on 1d4 at least 40 times. The percentage chance of that happening on a normal run has twenty-five zeroes in it. Even while sprinting, with an 80% chance of reaching top speed each sprint, you only have a 0.0001% of getting any significant distance out of it. Which, of course, is how it should be.

The Fast Runner

You get bonus actions based on your Dexterity modifier in my system, which means a maximum of 7 actions per Round, or 75% faster movement. I'm not so concerned about the effects that a negative DEX modifier will have - crippled characters are going to be crippled, after all - but obviously we don't want a few extra points of DEX turning your character into some kind of freakish superhuman.

Which, luckily, it doesn't seem to. Walking speed, with 7 actions, is 4.03 MPH. Running speed is 12.7 MPH. Sprinting is 16.7 MPH. Pretty much all of this is in line with what I'd expect. Walking speed does tend to imply a significantly faster marching speed, but remember that long-distance marching is more an issue of endurance than speed,

Finally, the Fast Runner's top speed clocks at 31.5 MPH - faster than Usain Bolt, though not by much. This is actually a lot less realistic - modern sprinters have access to technology and training techniques that allow them to greatly increase fast-twitch muscle mass, and can thus attain significantly faster speeds than any pre-industrial sprinter. Of course, I'd rather err on the side of letting lucky player characters outstrip Usain Bolt for a few rounds, so I'll accept this quirk of the system.

Endurance

I account for fatigue and exhaustion by folding it into an encumbrance-by-stone system. Both fatigue and exhaustion are treated as a "phantom" stone of carried weight, which you can only rid yourself of by resting for a Turn or sleeping for two hours, respectively. Regular encumbrance is at 5 stones, with the max at 10 stones.

Walking (and even running) have no associated endurance or fatigue penalties, since I want each to be as simple and straightforward as possible. Over long distances, each fits easily into normal march rules, and long-distance running is easily accounted for as forced marching, which I handle by allowing the players to move into a faster march speed level than their encumbrance would normally allow, by quadrupling the rate of exhaustion.

I require sprinters to pass a STR check every sprint, or else lose any defensive bonuses and get reduced to AC6. Fumbling is possible (1 in 20 chance normally, goes up to 4 in 20 when overencumbered), which in my system usually just adds a stone of fatigue. The character can expect 21.33 sprints per minute, and 6 minutes, 11.25 seconds per mile, gaining 1.07 stones of fatigue each minute.

However, by the end of the third minute, she'll have 3 stones of fatigue and will move from "unencumbered" to "encumbered," fumbling 2 out of 10 times. (From here on out I'm going to let my precision slide a bit.) During the fourth minute of sprinting, she'll gain two stones of fatigue, making her "overencumbered" for the fifth minute. At the end of the fifth minute, she'll have gained four more stones of fatigue (fumbling 4 in 20 times!) putting her at 9 stones. This means that, fifteen seconds into the final stretch, she'll need to roll Might saves every Round to even continue sprinting - otherwise she'll be forced to the ground, resting for at least an hour and a half in order to return to zero fatigue, never having even reached the mile mark.

Someone with 18 CON (which puts base encumbrance to 8 stone) would be able to do it - she could sprint for four minutes before becoming encumbered, and would only gain her eighth stone in the sixth minute - finishing the mile on time. with enough wind for two more minutes of continuous sprinting.

I don't know that this is absolutely realistic, but there's definitely a reason why sprint races cap off at 400 meters! Anyway, a good long-distance sprint in this system would be a half-mile in about three minutes, which would require a half hour of resting to get back to tip-top shape. That sounds good enough for me.

Running a Marathon

Marathons are 26 miles, and the world records are all a few minutes under two hours - which comports with the Fast Runner's running speed of 12.7 MPH. Of course, these are extraordinary cases - marathons are better handled with marching speeds.

Most competent marathon runners aim to run a four-hour marathon, and 2% of them breach three hours. Unencumbered marching speed is 4 MPH, which is a six-and-a-half hour marathon - clearly, we have to kick it up a notch. A 6 MPH forced march - at 1 stone of exhaustion per hour - gets you across the finish line in 4 hours, 20 minutes, and an 8 MPH double-time (1 stone per turn) will be 3 hours, 15 minutes, at 7 stones. Good enough - it provides enough of a baseline to estimate exhaustion and times if someone wants to actually race in a marathon, it's easy to remember (+2 MPH when forced-marching), and it's consistent, with 6 MPH being just under running speed and 8 MPH just over.

Fighting
You can fumble in combat! Surprise surprise! You'll be rolling 10 attacks per minute, with a 1 in 20 chance of fumbling - giving you a stone of fatigue every two minutes. At six minutes, on average, you'll cross into "encumbered" territory, averaging one new stone per minute. At eight minutes, you'll be "overencumbered" and gain, on average, a stone of fatigue every 30 seconds. You'll start to collapse from exhaustion after 10 minutes, 30 seconds of continuous fighting - 105 rounds. Not having done fencing or medieval reenactment (only paintball), I couldn't tell you exactly how accurate it is, but it's on a similar scale to sprinting, and is easy to remember, and easily accommodates the effects of excess weight on your combat endurance. The idea that, after six minutes of continuous fighting, you're really starting to push your limits makes intuitive sense to me.

Sneaking
Sneaking is the skill used to move without drawing attention to yourself, and I record it in feet per minute - this being how fast you can move before making a ruckus. As a Level 1 Trickster, you get 20' per minute, going up by 20' each time you level up and decide to improve Sneaking.
20' per minute is two feet per round, or four inches per Action - the slowest flat crawl. (Although, since it's notated as a per-minute speed, it also accounts for quick dashes between cover). By Level 10, however, said Trickster can Sneak at 200' per minute, which means their normal walking is completely silent. And at level 20, she'll Sneak at 400' per minute, or about 4.6 MPH - a brisk jog!

Since you can double-time with a DEX check, you can effectively attempt to Sneak at normal speed by level 5, while a 20th-level Trickster can basically Sprint silently, with up to 800' per minute/9.2 MPH Sneaking, which is obviously pretty fantastical. This meshes perfectly with my intended level scaling, where 10th level is intended to comport with the best that heroes could have accomplished in the real world, and all the levels beyond that representing states of increasing mystical perfection. Since double-timing a skill like this adds 1d4 guaranteed stones of fatigue, it remains a poor substitute for normal sprinting when simply looking at speed.

Swimming
Base swim speed is 30' per minute, or 3' per round, increasing in those same increments. This starts you at about 0.3 MPH, whereas the fastest swimmers can reach 5-6 MPH - which, of course, comports exactly with the swim skill in my system. A 10th level Trickster fully trained in Swimming moves at 3 MPH in the water, and at 20th level she'll reach 6 MPH.

Sprinting, of course, doubles these speeds - a 10th level swimmer can sprint at similar speeds to Michael Phelps, while the 20th level will actually be a bit faster in the water than she is on land.

Climbing
I'm the least familiar with climbing techniques, so I'll use some speed climbing records as a reference point. The 2007 record (since broken) on El Capitan, a 2900 foot climb, took about 165 minutes - about 17 feet per minute on average, though the climbers were reported to do most of the climb at 20 feet per minute. Half Dome, at 2,000 feet, has been done in 82 minutes - nearly 25 feet per minute. However, the tree climbing world record is about 50' in 13.65 seconds, or a bit over 250' per minute, and the Climb skill is intended to cover both of those bases.

Of course, tree climbs are clearly sprints, moving at double time, while thousand-foot plus climbs include numerous rest breaks. We can bridge much of the gap by working backwards from my system's "skill sprint" mechanic - cutting a 250' per minute sprint in half to get a "normalized" climb speed of 125' per minute. As for the rock climbing records, we can probably assign CON scores of 18 to the climbers, which gives them the ability to simply 'take' 4 stone of fatigue before things start getting dangerous. Climbing adds 1 stone per minute, and resting removes 1 every 10 minutes. So for every 4 minutes of continuous climbing, you'd need 40 minutes of continuous resting to return to zero.

Let's apply this to the El Capitan climb. 4 minutes 'on,' then 40 off, is 44 minutes. The second stretch puts us at 88 minutes total, 8 minutes active climbing. Third, 12 minutes climbing and 132 total. A half-stretch, 14 minutes climbing and 154 total, and a quarter stretch gives us 165 total minutes, with 15 minutes of continuous climbing, at 193' per minute. As this is a world record climb, we can assume a fair amount of sprinting, as well as simply accepting extra weight, and modern climbing technology to boot. I'll call it a normal climbing speed of 100' per minute, at Level 10 - or 10' per minute, at level 1.

The "catching" mechanic (you get a chance to catch yourself while falling every 10' up to your Climbing skill distance) is definitely less realistic, as even an experienced professional climber would not be able to simply "catch a ledge" after a 50 foot fall, but it does keep things interesting and allow for additional granularity when failing to climb something.

Conclusions
Since I use an action-point system, I realize I have a level of granularity below most people's systems. However, most of this translates well, I think - see this summary of my per-Round movement speeds:
Walking: 20'
Running: 64' (or just simplify to 60')
Sprinting: 85'
Top Speed: 160'
Climbing: 1' per level
Sneaking: 2' per level
Swimming: 3' per level