Turning it up to 21 Sided

Dramatic Magic in Dungeon World

2012-02-01T10:20:00.000-08:00

As often, a blog post that came out of a twitter discussion. And as often, started by Quinn Murphy:

[The important thing] for me is in making casting dramatically relevant. Making choices, taking risks, raising stakes.

In order to keep magic feasible in combat or other timed situations where characters take turns, yet still have some sort of drama associated with the casting of the spell rather than just a couple of die rolls, a spellcaster would need to be taking multiple actions to complete their casting. The problem can be managed by splitting the casting into phases, in which different things can go wrong.

Summon Energy
Shape Energy

Dungeon World already has some drama associated with partial success in casting spells: the player chooses between attracting unwanted attention, losing the spell (Vancian style) or taking an ongoing penalty to cast the spell in the future. But the drama is so compressed in such a system as to not be significant... or at least not as significant as Quinn was intending.

Assuming that each spell (or supernatural effect in general) has some energy cost that must be paid,

Summon Energy (CON/INT/WIS/CHA)

10+ You gain 3 points of energy towards the spell
7-9: You gain 1 point of energy towards the spell, and choose two:

You don't attract unwanted attention from the energy source
You don't take -1 forward
It's not obvious what you're doing

Why the multiple stats?

CON: You're using your own life force to power the spell.
INT: Arcane magic, you're summoning the raw magic stuff of the universe.
WIS: Divine magic, you're channeling the power of the Gods.
CHA: Spirit magic, you're coercing another entity to give you the energy.

Powerful spells might require multiple types of energy to cast.

Shape Energy (INT/WIS)

10+ You shape the energy into the desired magical elements
7-9: You shape the energy into only one of the desired magical elements, and choose two:

You don't suffer a backlash from the energy you're shaping
You don't lose any energy (may be recovered with further Summon Energy)
You can continue to Shape Energy

Magical Elements? Up to the world, but a spell might require multiple schools or shapes for the raw energy to generate the desired effect. For example a Haste style spell would require manipulating both Life and Time. In a more Elemental system, you might need both Air and Fire for a spell. Or in a Verb/Noun system, telepathy or scrying might be Move/Mind. If you don't have all of the magical elements required, then the GM/MC will adjudicate the results of the spell.Once all of the elements have been shaped, the spell goes off! There isn't a further roll for this, it happens after the final Shape Energy.

Dungeon World XP Hack

2012-01-21T13:53:00.000-08:00

First, read Ryan and Hamish's posts about the XP hack for Dungeon World, listed in my previous post. Okay ...

... and now on to my take on it.

Ryan drops Aid/Hinder from the initial six. I agree that there's a problem with it, as the Aid or Interfere move is weak by itself and that's the only real way to roll for this type. So do we keep it, or drop it? Before we cast it to the winds, let's see what can be done...

Move Healing into Aid from Defend. This helps Elven Wizards, Bards, Clerics, Paladins, but Rangers, Thieves and Fighters not at all. Maybe that's okay, as Fighters don't get much from Discover, Wizards not much from Defend, and so forth.
We can fix the Aid or Interfere move to not be worthless! Given that you're giving up an attack or other useful move, to have a chance at giving +1 and only a small chance of not attracting unwanted attention for that +1, it's hard to see when the move as written is ever a good choice. How about:
Aid or Interfere (Bond)
When you help or hinder someone you have a Bond with, roll+Bondwith them. On a 10+ they take +(1 plus your modifier for the action you're helping with) or -2, your choice. On a 7-9 you also expose yourself to danger, retribution, or cost.
Allow indirect assistance to count, so long as it doesn't already count towards marking XP. So a Defy Danger roll to get in someone's face before they make it to the tasty tasty wizard could count as Aiding the wizard (so long as it doesn't count towards Stunt or Defend). But this gets really messy, in an otherwise quite clean set of actions.

You still don't like it? Truth be told, I don't either. Okay let's look at the highlight types in more detail, and I'll highlight (ha!) what I would call them, plus get to why I think Aid/Hinder is janktastic. As in Ryan's first post, it's about saying "I want to see you do X", and originally, "so I highlight Y." So we get rid of the middle man, Y, and just tell it like it is:

Attack. I want to see you Destroying things (notably monsters).
Defend. I want to see you Protecting things (notably your fellow adventurers).
Discover. I want to see you Learning things (notably about the adventure).
Converse. I want to see you Interacting with things (notably the NPCs).
Stunt. I want to see you ... I don't know, jumping over things? Defying Dangerous things? The time that I saw Stunt used it was a style of play, rather than a particular action. I would rework Stunt to: I want to see you Risking things (notably yourself).
Aid. I want to see you Helping [with] things? Lame. And it's lame because it's not about doing awesome things that you can later tell people about. "My character totally destroyed this evil water demon!" YEAH! "My character totally gave a little bit of advice about how to do things to other people!" Uhhh... not so much.

And there we have it. In a roleplaying game I would never want to tell a player: I want to see you stand on the side lines and offer advice, and as a player I would never want to be told that.

Ryan's solution is on the right track, I feel, but not quite there yet:

When you Aid or Hinder another PC, and the move you’re affecting is covered by one of your highlights, mark experience.

If it was covered by one of your highlights, why don't you just do it yourself? Hopefully you're pretty good at it. And on top of this, in a reasonably balanced set of characters and highlights, it's quite possible that no one else is really doing your highlights and instead leaving it to you, especially with Protecting and Interacting.

So along with the above revision for Aid or Interfere, should you choose to do it, instead:

When you Aid or Hinder another PC, and the move you’re affecting is covered by one of their highlights, mark experience.

Then you're actually helping people to do what they should be doing, not what you should be doing.

[Edits below, covering points that Ryan brought up on Twitter, paraphrased]

If the highlight is one of theirs, it's impossible to use on NPCs as they don't have highlights!
True enough, but along with the principle that every monster has a name, perhaps just arbitrarily say that the player can mark XP or not as if the monster had a highlight. Or just say that it only works with PCs.
If you can interfere, then you can interfere with another PC using parley on you to stop them... and get XP if they have Converse/Interacting highlighted. That's fucked up.
Yes, perhaps just Aid and not Interfere. Very much an edge case.

21sided+INT; Dungeon World Analysis

2012-01-20T11:51:00.000-08:00

This is not a post about #DndNext, but I wish it was! As per one comment on Twitter, Wizards should just license Dungeon World for 5th Edition and call it done. Or as my players said, after last night's introductory game:

It has the perfect amount of rules, not too many to get lost or to get in the way, but enough structure to keep everyone roleplaying and creative for the entire session.

Quick summary:Three players, playing a Dwarven Cleric, Human Paladin, Elven Wizard. The Cleric and Paladin were of different deities. They started out having off-camera found out where an evil merchant lived, who is supplying an evil temple with slaves and weapons that they have vowed to destroy. (Which is essentially what we played in #BigGamingWeek). They came in through a tunnel from the river, beat up some guards (just), found and killed the Merchant and looted his secret study, before returning triumphant to the tavern.

What we love:

Simple and fast. DW plays super quickly with minimum fuss and mathematics necessary.
Bonds. Using bonds has the game start out with players and characters having investment in *each other*, not just the story or advancement/loot. By making bonds part of the system mechanically (via highlighting, XP, Aid), not just fluff sitting on top, there is motivation for everyone to be cohesive and tell interesting stories that engage the other players. This is total genius.
GM Moves rather than GM Rolls. This speeds the game up SO MUCH it is just insane.
Choice as balancing agent. By this I mean the way DW uses the "Fast, Reliable, Features: Pick 2" strategy to balance actions that are partially successful. I think DW would benefit from further expansion in this area, pulling from its parent Apocalypse World.
The LA Crew's XP hack*: The baseline XP system (highlight something, mark XP when you roll it) is great, and encourages and rewards doing things. But it's so broken it's not funny when you highlight a stat instead of a type of activity. The XP hack (read it now!) fixes this in one quick change.

*: Ryan's first post, plus Hamish's extension, and Ryan's Follow up

What we found confusing, after the first game:

GM Moves. If the players don't roll anything, then the first appearance is that the GM can't make a move. I didn't find it in the rules, but our assumption is that the GM can make a move in turn, using one of his NPCs.
The directionality of Bonds. This could be explained a bit better in the rules, with clear examples of all the uses of Bond. Also the Character sheet should have a slot for bond values per character rather than having to count the individual ones.

What we would have liked to see:

Guidance on initiative. IMO, rules are there to determine the outcome of a conflict. The conflict in initiative is who acts first, and it shouldn't be whoever shouts the loudest at the table. Our quick solution was to go around the table, including the GM.
Healing. It's freaking dull. Between the paladin just moving the damage around, and the cleric failing to cast spells, extra-combat healing is a total chore. Our solution, after two combats, was to have characters heal to max automatically given a quick rest. Cos, you know, fuck it. Also, it breaks the XP system into lots of little bits, as the Paladin gets XP for moving the damage around, and the cleric gets XP for then healing (or trying to heal) it.
Streetwise. Neither Discern Reality or Spout Lore really cover what should be a roll+CHA move to find out the word on the street.
Symmetrical Damage. The Paladin with 3 armor sat there defending all night, and sucking up the static damage as he knew exactly what was coming. Yet the players had to roll, doing sometimes fantastic and sometimes terrible damage. The asymmetry isn't great, either all damage should be somehow random, or all should be static.

What we added, but are personal preferences:

Reroll tokens: I love throwing tokens at people for outstanding roleplaying. And given that stats don't contribute that much, having those tokens used for rerolls works outstandingly.
Character backgrounds: Yeah, conversation and questions, but having a well thought out background actually written down (and extended during the game) helps both the player to keep in character, and the GM to have time to think between sessions about how to include it into the story. The reward for a background is +1 to a modifier that isn't already +2.

10 Thoughts on Deck Building Game Design

2011-08-16T10:36:00.000-07:00

What have I been doing for 6 months? Well ... running a Cyberpunk game instead of D&D, and thinking about deck building games (like Dominion, Ascension, Nightfall, Thunderstone, etc.).

Given you probably don't want to hear about my own campaign using my own system, here are my thoughts on DBG Design. I'll use the (probably) three most popular as examples of what's good and what's not in the current generation of game, and compare to the much more established and richer CCG market.

Current DBGs lack "factions", which are important for longer term engagement.
For example, colours in Magic, clans in Legend of the Five Rings (L5R) and Vampire: the Eternal Struggle (VtES), families in Deadlands, Secret societies in Illuminati, etc. Players will associate themselves with one particular faction, which may not be the most mechanically advantageous. The same way that sports fans will support their team which hasn't won in decades.
Setting/Story is important.
Other than Magic, most CCGs have a very strong setting and/or storyline. In L5R it is key, where player tournaments drive the world's story and which cards come into sets. DBGs to date have had minimal actual setting. Nightfall is a mishmash world reminiscent of the World of Darkness, Ascension has some repeated fantasy-esque names but nothing really coherent or engaging.
Player Interaction is important, but does not have to be "attacks"
Current DBGs have minimal interaction, compared to CCGs. The attack cards in Dominion aren't a good strategy, the attacking in Nightfall is important but very shallow, and there's essentially none in Ascension. Compared to the rich interaction in VtES with bleeding and votes, the multiple economy interaction in L5R (honor, combat, rings) and the current DBG fare is sorely lacking.
Winning is better than not losing.
To "win" Nightfall, you have the least number of wounds in your deck. You essentially lost the least, which isn't a substitute for proactive winning by your own actions. Ascension's honor counter is a bit weak, as it's just a timer, compared to Dominion where you actively choose to buy VP cards. Nightfall is chaos Magic, where everyone can attack anyone... except in Nightfall it's required every turn to attack with every minion (creature) you have out.
Economy is more than just card combos.
Most DBGs tend to be combo construction games rather than interactive economies. In Ascension card draw and deck thinning cards rule the roost, and Nightfall's chaining is unashamedly all about the combo combo combo. Even Dominion revolves around additional actions in a turn (Village!) and avoiding the VP cards to recycle your deck to get back to the start of the combo. Compare to the interplay between Blood Pool and Vampire's life in VTeS or the necessary mix of attributes in Spycraft, and DBGs come up short.
(Thanks to @gamefiend for this point)
Personal card stocks are good.
Lest I appear to be dissing Nightfall, I really like the personal sets of cards that only the player can buy. This means that the optimal strategy for each player is different, as opposed to Dominion where you could end up with 4 players each with essentially the same deck, or Ascension where the available cards are (for all practical purposes) random. See also point 1 about factions.
Cards should be personal, not rules substitutes.
Spycraft's mission deck was a very bad idea, and contributed to the game's demise. The rules should be solid enough to swap in or out any cards without compromising the integrity of the game. Anything where the cards are somehow a substitute for something that could be generalized to a simple rule is poor design.
Starting deck mechanics are crucial.
The starting deck in Nightfall is set, and the cards remove themselves from the game after use. This is awesome as you don't have the problem in Ascension/Dominion where you can get stuck with a hand full of junk half way through the game because you couldn't get rid of the cards or they're no longer useful.
Start up time should be minimized
I love playing Ascension on the iPad as it shuffles for me, lays everything out and generally takes out the chores of DBGs. Compared to Nightfall, where the cards are separated only by dividers and you need to fish out 2*number of players plus 10 more ... I get tired just thinking about it. If you can't quickly play a second game, then people won't. And that's bad.
Randomness from cards is a cop out.
Random Roll in Blood Wars, void score in L5R and similar, show a degree of poor design if used for random rolls rather than a balancing attribute. Illuminati is the worst with actual dice. If you want a board game with dice, make a board game. Admittedly this hasn't come up in a DBG yet, and I hope it never does!

Do you agree with my points? What else would you want to see in your perfect DBG?

Combat Speed, Revisited

2011-02-21T17:19:00.000-08:00

As always, a twitter "nerdfight" has prompted the thinking and writing of this blog post. And it's the same fight as previously, where people complain that combats take too long, others say that's the heat of the D&D kitchen and to GTFO if you don't like it, and the more reasonable among us try to come up with solutions.

Most of the suggestions to date try to patch the system without really attempting to understand the cause of the slow down. The posited reasons include player choice of actions, extra dice to add and so forth. Compared to previous editions and other games, these are insignificant as they happen everywhere. Vancian magic did not dramatically slow the game, nor does the power choice in White Wolf, Weapons of the Gods, GURPS, or any other game I know make such a difference. Similarly, almost every game has random damage in the same order as 4th ed D&D.

The problem that I see is that the system inherently does not scale properly without total character optimization, and even then it's dubious. Monster and PC HP scale linearly with level due to the math of the system. Monsters start at around 20 + (8*level) and PCs at about 24 + (5*level). To hit and defences also scale relatively linearly, assuming the PC takes the mathematical system fix feats. As damage is entirely random, with a minimum of 0 (on a miss) and the only relation between to hit roll and damage is the 1 in 20 critical (modulo rare powers and items), the difference at low levels is not recognizable but becomes more apparent as experience increases.

On average, the damage per level for monsters using the DMG's "Damage per Level" table increases by 2 per 3 levels. This assumes that the monster hits 50% of the time, and then does average damage per the dice. The damage used is the average of medium and high normal damage expression, as the typical amount an average monster would do per round.

Given this, a single first level monster can take out an equivalent single first level PC in 3.2 rounds. At 6th level, this has gone up to 4.9, by 11th 5.6, after which the rate of increase levels out to a maximum of 7.2 at 28th level. It's clear to see that the monsters will take longer to get rid of the PCs as levels increase. The extra damage from monster encounter or recharge powers with Limited damage expressions may reduce the number of rounds, but the effect is linear so the overall progression remains the same.

The same is even more pronounced for the PCs. As PCs can have at most 4 Encounter powers, which one assumes do more damage than At-Wills, and 4 Dailies at 20th level, the gap between PC damage output and monster HP will increase significantly. As level does not add to damage and an at-will does the same damage at 20th level that it does at 1st, it should be no surprise that combats take longer.

Added to this are the prevalence of conditions, and in particular action depletion conditions. With fewer actions (Dominated, Stunned, Dazed ... even Knocked Prone gets rid of a Move action), there is less opportunity to deal damage over time. This is compounded by the additional time spent tracking the conditions, saving against them and calculating to hit and damage rolls which are out of the ordinary.

And after writing this, I come to the undesirable conclusion that the necessary variation of powers in the level/role matrix makes this practically impossible to track, which in turn makes it practically impossible to properly fix with house rules more complex than reducing monster HP by 3*level and limiting the number of conditions in play at any given time.

Some suggestions include:

Add 1/2 level to all at-wills and basic attacks. This makes these attacks more desirable, reducing time spent on action choice and increasing the damage done.
Increase PC damage to follow the monster damage per level charts, if less. Remember that the monsters gain 8HP per level, whereas the PCs gain only 5 on average, so these should be the absolute minimum! Role and power type would determine the table and column to use.
Make all daily/encounter attacks do half damage on a miss, and all at-wills do 1/2 level on a miss. This pushes the average damage per round back up without overly affecting the low levels.

Post interrupted by earthquake! More later :)

Enjoying Combat: Speed and Awesomeness

2011-01-14T12:26:00.000-08:00

@gamefiend, in response to criticisms on twitter about the glacial pace of higher level 4th ed combats, recently posted a great blog entry about possible causes and the use of environmental accelerators as a way to speed things up. I'd like to give some further ideas about the way in which my group has attempted to make combat both more enjoyable and faster.

The group consists of 5 players plus myself and I would say that the most tactically focused player is about 60/40 for tactics/roleplaying, and the least is closer to 20/80. I start with this rough guide so that you understand that combat in our game MUST drive the story, rather than being an end in itself.

As many have noted, it is very tempting in combat to simply announce the name of the power that is being used, roll the dice and cheer. GameFiend might even call this a procedural trick, along with tricks like strictly limiting time per player's turn. Gamefiend also notes that reducing the HP of monsters simply destabilizes the game, with which I agree. The math behind the game design has been well scrutinized, leading to corrections through feats and more feats, better monster design and so forth. Our experience goes very much in line with GameFiend's analysis: procedural tricks and math adjustment do not help in any signficant way, and the most effective route is when the players all know the rules and their character's abilities. I'll add to this that the DM must also know their monsters' abilities -- time spent considering which attack to use must be reduced on both sides of the screen.

This is not always possible even for experienced players. Especially when bringing in a new character at paragon or higher when the interactions of feats, abilities, powers and items make it more and more complex to participate effectively and quickly.

Secondly, the gap between PC damage output and monster defences/hp is a significant factor that has not yet been discussed. Clerics in particular provide a large chunk of healing power that drags out combats, or makes the party feel immune to the monsters until the Cleric is either dropped or has run out of magic. Neither of these are acceptable targets for the DM -- picking on a single PC sucks, and the game is already a chore by the time a 8+ level cleric runs out of healing. Thankfully the player for the cleric was good-natured enough to bring in a new character, and the party's Bard should be able to prevent random PC death. I have previously made this point on twitter that Striker should not be a class. Every character should be a Striker, with a sub-role of Controller, Leader or Defender.

Before going on to the solutions that we have employed, it is also important to consider the philosophy of combat encounters. If you think, as I do, that the heart of any roleplaying game is a social contract between the participants, then what are the social norms for combat?

1. Every participant should be able to show off their strengths: Be that awesome control, mega-damage, tricks and cunning to astound, or any other aspect. And this applies to both DM and player. A frustrated DM that is having their monsters locked down and slowly whittled away is just as bad a situation as a frustrated player, if not worse as the DM controls more than one combatant.

2. Every participant should ensure that the combat is enjoyable for the others. The most obvious aspect of this is being quick with your turn, but also all players that I have ever gamed with appreciate interesting descriptions and innovative play more than effective tactics. Get on with doing your cool shit, even if it's not the best possible choice.

Two solutions work for our group:

1. Awesomeness

Essentially the same as GameFiend's accelerators, if a player has a good idea that fits in with the environment or otherwise shows off the awesomeness of the character, then unless a skill check is failed very badly, they can pull it off with a good description. The DM does not necessarily have to think off possibilities beforehand -- if the idea is cool, then run with it. This fits the points above, as the PC can show off their strengths and it makes the combat more interesting for everyone.

2. Roleplaying Tokens.

I have two (and only two) tokens that can be gained at any point in a session for good roleplaying. These tokens can be spent at any time to reroll any roll. Further, if the token is gained during combat, it can be used in that combat to regain a used Encounter power. The tokens do not persist between sessions, you cannot have two at once, and only two can be held at any point in time. You may wish to compare this set of rules with the recent fortune cards announcement.

The mechanical choice of a reroll is important. It normalizes the probability of awesomeness, both in and out of combat. It can be used on skill checks, damage, to hit rolls, ability checks or anything else. In practice, it is almost always used to reroll a miss, which follows points 1 and 2 above, and speeds up the combat.

The ability to regain an encounter power is also a well-considered one. Remember that the only way to get a token is to roleplay. Refreshing an encounter power is enough of an incentive to remind players that they should be descriptive and interesting in combat as well as out. It's always a roleplaying game, not a tactical boardgame half the time and a roleplaying game the other half. It is not, however, enough to seriously unbalance the game as drastic changes in the mathematics are bound to do.

There is another aspect to tokens that I am going to put into effect, after a quick trial based on ideas from @DMSamuel. If, in combat, you have a token then you may expend a minor action and describe part of the environment that has not yet been described. It cannot provide any permanent effect (eg no treasure etc), nor any direct damage. It can be used by either the DM or players to gain a mechanical advantage. This is GameFiend's accelerators, but put in the hands of the players.
Let the player describe the world as they see it and they will be more immersed, rather than divorced due to mechanical tactical considerations.

If you try these solutions, please let us know how they work for you!

You've been Adventuring in a Dream, Neo...

2011-01-12T14:00:00.000-08:00

The following are rules that I used to successfully run a dream world adventure. In the adventure, one of the characters doesn't wake up after an extended rest, and the party figure out that he has been taken captive in to a Dreamwalker's dream prison. They find an old Vistani Seeress who can transport them into the dream, in order to enact a rescue. In the dream world, they find that they cannot escape without either waking the Dreamwalker, or finding an exit built into the dream.

"Do you believe that my being stronger or faster has anything to do with my muscles, in this place?"
In the dream, any character may use Will to replace Fortitude.

"The body cannot live without the mind."
If you are killed in the dream, you are still dead.

"When you go into a dream within a dream, the effect is compounded"
You cannot take an extended rest while in the dream. This dream-within-a-dream would allow the dreamwalker to whisk the character down a level of unreality from which escape would become impossible. Or at least overly complex for the DM to cope with :)

"A glitch ... It happens when they change something."
Both the Dreamwalker and the PCs can attempt to bend the dream world to their will.
As a standard action, a PC can make a Charisma + 1/2 Level check against the Will of the Dreamwalker. If they succeed, they can change one aspect of the dream (unless it would break the adventure). If the PC fails, the Dreamwalker may make an attack against the PC. If the PC rolls a 1, then the attack automatically succeeds as a critical hit.

As a standard action, the Dreamwalker may change one aspect of the dream without a roll. In this case, any PC may make a Charisma or Intelligence + 1/2 Level check against the Will of the Dreamwalker. If this succeeds, the PC may either regain 3 expended healing surges or escape from the dream back to reality.

As the originally trapped character is a Monk from an oriental part of the game world, I used the Oni Souleater from Open Grave as the Dreamwalker's aspect in the dream, giving the PCs a target of 26 Will, and healing surge related powers to dissuade them from making things even more complicated.
Following a Rokugan theme, the characters had to recover 5 seals (fire, earth, air, water, void) to open a door to get to the projection of the Dreamwalker. They could also have demonstrated to guards on an otherwise infinitely long bridge that they followed the tenets of Bushido, and hence should be allowed to leave. Given the ... unaligned ... nature of the party, the requirements to act with honor, duty and even compassion made them decide against this route.

The Fine Art of Creating Pre-Gen Characters

2010-10-04T09:28:00.000-07:00

I was challenged by @gamefiend to come up with some better pre-generated characters for the Home Encounters adventures than the quick ones he had thrown together for the playtests. That got me to thinking about how pre-gens should differ from creating a character for myself.

I came up with the following set of design constraints:

It should be easy to connect with as a character. Games are much more enjoyable (and much easier to run) when the players get into their characters and roleplay. This is only possible when they have some connection with the character, which is much harder when they're presented with a character sheet to play at a moment's notice.
The characters should have evocative portraits in order to help with the connection issue, however copyright prevents us from just grabbing images off the web. A great resource would be an artist willing to create these, or a set of creative commons licensed images.
They should be easy to play, mechanically. Roleplaying always takes a back seat to the system at the beginning, while the players are figuring out how the character works mechanically. The quicker it is to understand the mechanics, the quicker the roleplaying can get going, and the better the game will be.
They should be built from primarily (but not necessarily exclusively) PHB material. Everyone has it, and they might not have access to individual Power books. The characters should definitely not use powers or feats from Dragon or other less well known and less accessible content. This is because the players may need to check the fine print on something which isn't on the character sheet (due to bugs in the Character Builder, or just out of curiousity), and hence they'll need the actual physical book. It also helps with being easy to play, as the PHB is the most well known set of rules.
It should make it easy to connect the character with the story, and the other characters. If the characters' background pulls them into the story, the suspension of disbelief as to why they're out killing these monsters and exploring the areas in the encounters goes away much more quickly, improving the experience. As we have complete control over the world and the character builds, it makes sense to have them integrated.
They should thus fit into the world, potentially at the expense of the rules or guidelines. If the area is 99% human, then it makes no sense to have a group of 5 characters with 5 different races, none of them human. Of course if the world is 10% human, 10% dwarf, 10% halfling, 10% deva, etc then that's fine! :)
It's easiest to connect with the story when you're telling the story. If there is any opportunity for the players to take the role of DM without subverting the encounters, then it should be encouraged. In this regard, the use of character backgrounds goes a long way -- instead of having a faceless NPC ranger tell the group that he has found tracks, have a PC Ranger discover them before the beginning of the game and give the player all of the information to relay to the group.

What do you all think? Have I missed an important design consideration?

Forgetting Powers

2010-09-22T11:59:00.000-07:00

As always, a twitter discussion has prompted this blog post. To summarize the debate, ignoring the Essentials background, some participants (myself included) dislike the rationale that a character becomes too tired after using a Daily power to use it again before an extended rest, even if it's the very first thing they do in the first encounter of the day. In fact, no one seemed to be very keen on the rationale, the difference of opinion was whether or not it should just be ignored in favour of the simplicity that the model brings.

The approach seems to have come from applying the much-maligned Vancian memorization system to every class, as opposed to making spells more like regular abilities. The fighter "forgets" how to make that deadly attack until the next day, and the Rogue "forgets" how to use their daily utility to tumble. From a commonsense standpoint, this is even more ridiculous than wizards "forgetting" how to cast spells.

From a game design perspective, there must be some limiting factor so that the characters do not use their best powers for every single attack. Vancian memorization solved this for magic in earlier editions and for everything in the current edition, and lack of additional powers solved it for non-magic before 4th edition.

Given the amount of design ideas borrowed from online games that have made it through to 4th edition, it is somewhat surprising that one practically ubiquitious idea that did not was magic points, stamina, mana or however you wish to call a pool of points that are spent in order to use abilities. From the original Diablo through to Dragon Age, power points have accompanied hit points as a standard mechanism for adding a limit to the amount of awesome a character can pull off. Otherwise it wouldn't be awesome, right?

The simplicity argument against a stamina system is compelling, however the same argument could easily have been applied to the escalation in Hit Points and corresponding damage capabilities between 3rd and 4th editions. For those of us that like tinkering with the system a little, it would be possible to overlay a Stamina system on the current Daily/Encounter/At-Will framework with only minimal additional complexity. And, of course, I hereby present such an overlay system.

Azaroth's Stamina Overlay System

Stamina Points: 8 + 4 per level.

Stamina Costs:

Daily: 4 + level of power
Encounter: 2 + 1/2 level of power
At-Will: 0

You can spend a Healing Surge to regain 25% of Stamina (eg 2 + level) during a Short Rest. Bonuses to healing surges do not apply to the amount of Stamina regained (although one could easily imagine a parallel track of Stamina boosting powers and feats). Stamina is reset to full upon an extended rest.

Note that the characters do not receive any additional Healing Surges in this system; it is necessary to decide if you want to use the surges to heal or gain access to powers.

Optional Rule: The cost of a Daily power that misses is 1 + level.
Optional Rule: The cost of an Encounter power that misses is 1/2 level.
Optional Rule: Daily powers may only be used once per encounter.

Examples:

At 1st level with 12 points (8+ (1*4)) and without spending any healing surges, a character could use their first level daily (5 Stamina) once and their first level encounter three times (3*2 Stamina). A healing surge would restore 3 points, enough (with the left over 1 Stamina) to use the Encounter power twice more.

At 8th level with 40 points (8 + (8*4)), a character could use their 5th level daily once (9 Stamina), their first level daily twice (5*2 stamina), their 7th level encounter twice (5*2 stamina), their 3rd level encounter three times (3*3 Stamina), and their 1st level encounter power twice (2*2 Stamina), without using Surges to regain Stamina between encounters. A Healing surge would restore 10 points, enough to use the 7th level encounter power twice more, or the 5th level daily once.

Turning D&DI up to 21sided

2010-08-27T16:24:00.000-07:00

Time for a new blog entry. Some may wonder where I've been over the past few weeks and not been posting to the blog. I had thought of writing a post about the awful, outdated, inefficient, insecure and featureless search interface that Wizards charges us for at D&D Insider. Instead, to quote @SarahDarkmagic:

The open-source person in me feels like instead of complaining about certain [magazines], I should work to make them better. Just not sure how.

As an information scientist, I feel exactly the same way about the D&DI database. However, with a little bit of (T&C bending) black magic, some python code and an xml database I developed in a former job, I was pretty sure I did know how their system could easily be improved upon.

Ignoring the black arts required to obtain the data (no kittens, servers or anything else was harmed in the process), the design and requirements for an improved database were straightforward:

Google-style search box that "just does the right thing"
Facets for refining your query (the things in the left bar in most modern search interfaces)
More statistics able to be searched, including from powers
Multiple items able to be displayed at once for printing, in a compact format
Improved screen real estate usage (no huge search section at the top)
Improved reaction times (no delays while downloading long, unseen results)

And indeed, even without annoying my new wife by coding on our honeymoon, it was pretty straightforward given the right tools.

The basic results screen for a search for "goblin" looks like:

And the multiple item display screen, optimised in three columns for printing landscape or two columns for portrait, for the selected creatures:

The system calculates an optimal layout to save space, putting the two minions in one column and the much longer champion in a column by himself.

The search system is intuitive and powerful, enabling searches based on any statistic including parsed from the monsters powers. Want a medium humanoid between levels 10 and 13 that has a power that dazes? No problem: medium humanoid 10..13 dazed

Further improvements would be easy, given user identity:

Saving views per encounter (named from the selection of the three goblins, for example)
Restrict the search to only owned books, currently the 'books' restriction is to the 3 Monster Manuals
Automatically restrict the search to the current levels of the DM's campaign
Suggested monsters or intelligent ranking based on previous searches or campaign information
Tie it into further systems such as combat trackers for DMs, trophy display halls for PCs

So, where can you try it out? Given the "copy for personal use only" restriction on the D&DI data, I can't make it publically available. Given that I can't tell if you're a subscriber or not, I would end up in a very large breach of copyright lawsuit. Now, if the copyright owners were to take this on themselves, that would be a different matter. The good news is that like SarahDarkmagic, I work with open source, and would happily give Wizards everything they need to run it. Perhaps a small consulting fee might be in order for future customizations, though :)

If I were to continue to develop this, what other features would you like to see in a revised D&DI search interface?

Environmental Disasters

2010-07-20T14:05:00.000-07:00

"You know that pool of acid, we'll collect vials of it in our spare potion flasks. It was 10 foot deep?"

"We put a hold on our quest and ferry the iron balls back to town to sell them to the blacksmith as raw materials."

"We take the adamantine off the doors to sell, I calculate it's worth 2 million gold pieces and will take two trips filling our Portable Hole."

While more frustrating gaffs rather than disasters on the proportion of BP's destruction of the Gulf of Mexico, such reckless uses of the environment can provide unintended benefits or disadvantages in D&D just the same. Given PC ingenuity at turning situations to their advantage, it's more likely to be the former. Bonus points available for pointing out which (published!) adventures these environmental disasters are from.

While mixing it up with respect to the terrain can make an otherwise pedestrian encounter into an interesting battle, care must be taken not to either overly punish the PCs for something beyond their control, such as not being able to fly before mid-Paragon, or give them unintended treasure or beneficial effects.

Some things to keep in mind:

Is it portable? If it can be removed from the dungeon, and it's valuable, then it will be. All of it. No matter how long it takes. Do you really want that acid to be stored up as an effectively endless supply, or sold at even 20% of value to the local Alchemist? And no, putting in an army of trolls just to use it up is not a good compromise!

Can the magical effect be ended? Natural occurrences are more likely to be portable, but magical effects must have come from somewhere too. Think about a way (skill challenge?) that they could be ended. Arcana, History, Thievery, Dungeoneering are all good candidates.

Do the monsters have ranged attacks and a method of movement that the PCs won't be able to keep up with? For example, if the ranged monsters can fly or teleport between otherwise inaccessible areas of the battlefield, a TPK is highly likely. Even kobolds can be deadly in sufficient numbers when equipped with ranged weapons and not easily reachable. Old schoolers may remember the Dungeon adventure "Out of the Ashes" with a split up necklace of fireballs and kobolds on swinging ropes...

Have you just overdone it? Stick to zero, one or at most two tricks per encounter. Always having to deal with crazy terrain is frustrating, as opposed to interesting when used in moderation. This point is borrowed from someone else's recent post that I can't find again (apologies!) Some difficult terrain and a trap, great. Icy spires with wind gusts that randomly move characters around, dropping them down pits or onto spikes, difficult terrain, ongoing cold damage, ongoing grease effect ... maybe you're overdoing it.

Do you have any other disasters to share, or hints to keep DMs from creating a 20 billion gold piece escrow account to take care of damages?

Quantifying Nostalgia: The Magic Missile Debate

2010-07-07T12:26:00.000-07:00

In the most recent set of updates for 4th edition, Wizards of the Coast have made an interesting change to the Wizard at-will power Magic Missile. The interest is partly mechanical, however the more puzzling aspect is the reason given.

Magic Missile
Page 159: Replace the Attack, Hit, and Special entries with the Effect and Special entries in the power below. This update reflects an effort to restore the power to its classical form.

The change to the power is not to reduce its power from being broken, or increase its effects to make it a more valid choice, instead it is to try to capture the hearts of players of former editions by removing the need to roll to hit. One of the well known D&D author twitterers @ChrisSSims said on the matter:

The reasoning behind the change is probably, as others have said, "Red Box" reasoning. I disagree strongly with that sort of reasoning.
Nostalgia is hard to quantify, so you risk much basing any change from an established norm on it.

Which is indeed true, but with twitter and sentiment analysis, there is a measurable pulse to determine the reaction of the internet crowds. The reason that Chris's tweet is true is that sentiment analysis is itself very hard. Especially when the text is limited to only 140 characters!

Twitter Sentiment is a tool that attempts to classify tweets into positive or negative, and if you look at the results for the search for "magic missile" you'll see just how hard it is to get right. Another example is Tweet Feel, or an example of per user instead of per topic: :-) or )-:

Given the very hit or miss nature of automated sentiment analysis, and the relatively small scope of the topic, we can do this exercise by hand. For an active dislike, we'll assign -3 points, and +3 for an active like. For people who would prefer a different change, we'll assign -1 point.

Twitterers who dislike the change:

@cintain, @Neuroglyph, @ve4grm, @Trachalio, @newbiedm, @smitty_mckraken, @the_ravener, @Milambus, @ChrisSSims, @azaroth42

Twitterers who would prefer a different change (and hence dislike the current change):

@jaron95, @plays2much_Dnd, @richgreen01, @AlioTheFool, @Neldar, @RTGoodman, @theweem, @1sowa, @ObsidianCrane, @mayttvw, @offby1, @NecroLounge, @semifamous, @SarahDarkmagic
@loganbonner

Twitterers who actively like the change:

@IanAger, @DaveTheGame, @NMcCoy, @KenOfGhastria, @CaptCalamitous, @robertsongames, @revlazaro, @anthelios, @greywulf, @pdunwin, @Alphastream, @JohnduBois, @carolinacharlie, @WolfSamurai

This gives a total of: (-3 * 10) + (-1 * 15) + (3 * 14) = -30 + -14 + 42 = -3
If instead we take all of the dislikes and prefer different together, we end up with 25 against, to 14 in favour, 64% negative and 36% positive. Interestingly, from first glance it seems that the old timers are all against the change.

This is admittedly a lot less negative than I expected it to be. Feel free to weigh in on the debate in the comments, or if I should move you to a different category!

In Defence of the Human Condition

2010-07-02T08:22:00.000-07:00

"You arrive at the Inn, which is owned by a racist human owner. You are charged 10 additional gold pieces a night for being a group of freaks."

Punishing characters in game for their choices at character generation is not sporting. However the 4e rules encourage "racially diverse" groups constituted from every race other than the traditionally predominant human, a trend that is reinforced in books after the first Player's Handbook. In my game I prefer the group not to be an inexplicable group of races from all over the multiverse, so have instituted the following house rule:

Humans get +2 to any two ability scores, instead of just one.

The instant reaction might be that this pushes them far beyond any of the other races, but let's assess in light of the 3 Players Handbooks and the playable races from Dragon and Monster Manuals.

The only addition is to bring their total ability scores inline with the other races that get +2 to two. Most of the races have two fixed stats that are increased, however in the Players Handbook 3, WotC have introduced races that have one fixed and one choice between two. The Shifter in PHB2 was a start in this direction with two sets of stats for the one "race". So choice in stats is not limited to just humans.

Secondly, with the proliferation of races and choices, the matrix of ability score combinations is already practically full; the only empty cells are CON/DEX and CON/INT, and this without using Dragon magazine, Dark Sun or Monster Manual races (See Table 1 below for reference). This is understandable, as the first class with CON as their primary ability score was only recently introduced in PHB3. So if you want to play that STR/WIS combination, no problem, pick a Longtooth Shifter or a Minotaur. So rather than playing a sub-optimal and boring human, instead there's an option with interesting powers and background that's far more appealing. You might not get the exact encounter power of other features of another race with a different ability score profile, but you wouldn't get an encounter power at all from being Human.

Which brings me to the next point: Humans are dull. As one of my players said at our game last night: "We 'play' a human every day in the real world, why would we want to play one in a fantasy game?" Even if the extra +2 did make them mechanically superior, they still have no intrinsic flavour. Between a bonus to an ability score, versus the power to teleport and not having to sleep, I know which one I find more interesting. Or between a bonus to an ability score, versus an immortal character that can add d6 to a d20 roll due to being in tune with racial memories and is resistant to radiant and necrotic damage. Living constructs, Drow, Minotaurs, even the previously criticised Gnome and Half-elf are both mechanically better and more flavourful than a Human in this edition.

In our experience of using this rule, there are still non-human PCs, it's just that not every PC is non-human. The Dwarf Fighter, Dragonborn Rogue and Eladrin Wizard continue to shine in their own ways in contrast to the Human Druid, Bard and Monk. Instead of unbalancing the game, it corrects the power-creep of additional books that have turned a believable group into an X-Men like bunch of misfits.

We have found that the house rule re-balances humans rather than over-balances them, and any perception of being too strong is catered for both by lack of flavour and being the predominant race in the world in which we play.

Table 1: Ability Score Combinations

STR CON	Goliath, Minotaur, Warforged
STR DEX	Half-Orc
STR INT	Genasi
STR WIS	Longtooth, Minotaur
STR CHA	Dragonborn
CON DEX	--
CON INT	--
CON WIS	Dwarf, Wilden
CON CHA	Half-Elf
DEX INT	Eladrin
DEX WIS	Elf, Razorclaw, Githzerai, Wilden
DEX CHA	Halfling, Drow, Changeling
INT WIS	Deva, Githzerai, Shardmind
INT CHA	Tiefling, Gnome, Shardmind, Changeling
WIS CHA	Kalashtar

Carnival: Unwinnable Encounters?

2010-06-25T09:16:00.000-07:00

This post is part of a Blog Carnival, started by @ThadeousC. Please see the end for the rules and links to the other posts.

The question posed was whether it was allowable for a DM to put in encounters that forced the party to run away. Essentially this means encounters that the party has no chance of overcoming; expressed in function of monsters that would take an incredibly improbably amount of luck to kill before they killed the party. This is a slight rewording as it is mathematically possible for any party to survive any encounter, as the party could roll critical hits every attack and the monsters could either fail to hit or do so little damage as to be healable during the encounter. Of course, this is much like New Zealand's chances were of winning the World Cup: any encounter over 8 levels higher than the party level is highly likely to be a TPK (total party kill).

There has been a great deal of debate on these points, to which I add my own thoughts. To summarise, I am strongly against putting in unwinnable encounters that are not obviously a death trap, and do not have an obvious way to avoid them. This is for three reasons:

The idea of a social contract at the heart of the game
That the question can be reduced to the issue of character death
And that the question is a symptom of the issue of free will

Social Contract

I strongly believe that the basis of any enjoyable tabletop roleplaying game is an implicit social contract between the players and the DM. The players tacitly agree not to do things to intentionally make the game unenjoyable or act in such a way as to make it impossible to continue to play. For example, players that kill other characters, act in essentially random ways, try to break the flow of the story at every turn, or otherwise are disruptive in or out of character are in breach of the social contract. Most DMs would be extremely displeased with players that behaved in this way.

The DM's side of the contract is the same as the players'. They should not act to intentionally make the game unenjoyable, and should not act in such a way that makes it impossible to continue to play the game. Assuming that it is unenjoyable to have your character killed by an impossible-to-beat foe, then the DM should not do that. Equally, it is impossible or at least very hard to continue after killing every single character in the game, and this should thus be avoided.

If you agree with the premise of the social contract, it seems that the answer is very plain that DMs should never put in encounters that cannot be beaten. However, so long as the encounter is obviously unable to be beaten, and there is a way to avoid the situation while still acting in an in character fashion and advancing rather than destroying the story, then this can be enjoyable and the game can continue. Hence the two are not mutually exclusive, so long as those two conditions hold.

Character Death

The reason that the question has generated so much attention is that is a derivative of the most perplexing issue in roleplaying: character death. The question can be reduced, in essence, to whether or not the DM should kill characters. If the encounter is designed to be unwinnable, then the characters will certainly die... unless the players can see that it is unwinnable and can escape without compromising their in character personae.

My view is that characters should be only killed when the player makes it unavoidable through stupidity, and conversely that characters should not be killed due to bad dice rolls, if possible. This is consistent with the conditions above, as if the players ignore the obviously unwinnable nature, then their own complete stupidity caused the outcome.

However this means that the overpowering difficulty needs to be as obvious as possible, as players can be distracted or simply not be thinking about the plot in the same way as the DM who knows all the variables. Is it bravado on the part of the monster, or is it really going to kill and eat them all? Are those bones everywhere the bones of previous, higher level adventurers or is it just a cool setting? Is this a big young dragon, or a small ancient wyrm lord?

As the party may still not notice the difficulty after all of the clues, there must be a way for them to escape once it becomes obvious that they will otherwise all die. The monster thus cannot move faster than them or have better modes of movement, must be somehow unable to chase them, or have some very good reason to not chase them. Simply letting the party get away is not sufficient, this leaves a very bitter taste and looks like the DM is toying with them, as this is how the DM's monster is behaving.

There must also be a way planned in advance for the party to recover dead or unconcious party members. Character death in this case is bad, especially if you agree with the social contract, but complete character loss because the corpse could not be recovered is much much worse.

If, given all of this information and the possibility to escape without permanently losing characters, the party decides to continue to the inevitably bitter end, then they are either throwing away the social contract out of frustration, or incredibly stupid and deserve to be rolling up new, hopefully smarter, characters.

Unless you belive that the DM should never kill characters, this means that a well designed overpowering encounter is still possible. However the stress must be on the design aspects; it is not sufficient to simply throw in a super monster and expect everything to come out for the best. 3rd ed players may remember a certain Roper encounter, in which the low level party should NOT attack the monster, instead skirt around it ... however the map was drawn incorrectly meaning that there wasn't a way to avoid it, nor any reason for unknowledgable characters or players to do so.

Free Will

Finally, the issue can be thought of in terms of free will.

As a philosophical primer, determinism is the idea that every action is caused by prior actions and conditions. That every action is the effect or some prior cause, and hence there cannot be free will as one action leads inevitably to the next, which leads to the next, ad infinitum. For more information, the Wikipedia and Stanford Encyclopedia of Philosophy pages on the subject are very informative.

The reason that players do not like being railroaded into a particular course of action is that the DM is enforcing the most rigid definition of Determinism, when there are obviously other factors that mean the characters should have more free will to behave differently. No one, in game or in real life, likes to believe that they do not have free will, and railroading is shoving unwarranted determinism in the faces of the players.

This is compounded by the fact that character determinism is demonstrably false. There are always at least two outside factors that affect the choices and behaviour of characters: the player and the dice. The player has knowledge and experience that the character does not have, and will bring this to bear, especially on unwinnable encounters. And the converse is true, even the most immersive player cannot really know how a character would act in any situations. Secondly, failing a skill check or attack roll due to the inherent randomness of the system is something that defeats a deterministic world view. These actions should be predictable, yet are not even if complete knowledge of the state of the world was available.

Complete free will is also antithetical in the roleplaying context, as it is a breach of the social contract, or detracts from the story if you prefer to think of it in that way. Complete free will would be the ability to make a decision or take an action with no causal factors. Essentially this is acting randomly which will make the game meaningless.

So by playing in the game, the social contract allows for some free will, but not complete free will. By enforcing a lack of free will the DM is in breach, but the player is in breach by taking the opposite extreme.

The Sandbox style of running is, in my opinion, the answer to this question, and can be thought of as Soft Determinism or Compatibilism. Players who are enjoying the game and have sufficient information about the situation will trend towards predictability. This is part of the social contract -- that the players will not actively disrupt the game. If there are reasonable and obvious next steps, they will normally be taken.

To then apply this to the question, we have the same position as the previous approaches. If there is sufficient information for the players to make a good decision, then they likely will, and at the same time will continue to believe that they have as much free will as ever within the context of the game. The same way that the party makes the decision to attack the evil, and killable, monsters, they will make the decision to escape from the unkillable one. They could stay until the end, but they made the choice not to.

The free will issue is the most likely to generate breaches of the social contract, as the players will become quickly frustrated that their actions have no bearing on the game, and they have only one option remaining: mass suicide. A particularly horrific adventure in the Deadlands setting was published in the box set for The City of Lost Angels. To cut a terrible story short, the climax of the adventure had no less than three NPCs who were so powerful as to not have published statistics. The writers obviously believed that the players would recognise them from background to the setting, and run away. Our characters however had no idea who high priest was, or that the demon in the cathedral would be eating our souls for breakfast given half a chance. So on then running into the baddest bad ass in the setting, Stone, we decided that the GM was already in breach of the social contract and that the best resolution was to continue to attack the unkillable, undead, magical gunslinger until every. last. character. had fanned their last hammer.

If you put in unwinnable situations, at least Don't Be That DM.

Carnival Rules and Posts

Your post must be on topic.
If you wish to participate in this blog carnival, please post a comment saying so and you will have 24 hours to write and post your response.
You must add a link to all of the previous authors carnival posts at the end of your post.
Be respectful. No name calling.

Enjoy!

Reputation: Who are you again?

2010-06-19T23:20:00.000-07:00

"Hi, I'm Finn Rockingham, these are my friends and we need to ask you a few questions..."

A typical opening in any mystery based adventure where the characters must find out what is going on by interacting with the rest of the population of the world, rather than just jumping straight into a dungeon filled with monsters that obviously need killing. However, does the woman about to be quizzed to within a HP of her life know the PC, or was the introduction as useful as saying hi to a Rot Grub?

The easy answer is whatever the DM currently wants, but the same could be said of any aspect of the game. Unless it's important that the NPC know or be blissfully ignorant of the PC, a dice roll seems in order ... but what? Diplomacy or Bluff is the next step once the questions start getting fired. Streetwise is to find out about the locality, and a Charisma check doesn't really reflect the issue, as uncharismatic characters can be famous as easily as that Dragonborn Sorceror. The NPC could roll Streetwise, but this keeps the 'power' in the DMs hands as the NPC is unlikely to be fully statted out, and hence the question reduces to how the DM is feeling at the time again; this time regarding the level of Streetwise.

Enter: the Reputation "skill".

By putting the check in the hands of the players, it increases their involvement with the world and the current situation. They will likely want their character to be famous, or conversely an Assassin may wish to actively cover their tracks, as being known is a career ending problem! For players that are interested primarily in increasing their statistics, they are forced to engage with non combat situations or be forever "that guy" who always seems to be following after the more famous Finn. For anti-hero characters, it puts their destiny in their own dice-rolling hands. Rather than it being an arbitrary decision on the DM's behalf that the guard knows the rogue is the wanted master-thief Shagrat, Shagrat's player's roll will make that determination.

Reputation (No Associated Ability)

When first meeting someone, make a Reputation check to determine if they have heard of you. This does not guarantee a positive reaction, just that they know who you are. If you are known to be wanted by the law, and a city guard recognises you, that's likely to cause trouble, however if you have a reputation for helping bring criminals in the city to justice, then they're more likely to be friendly.

This pseudo-skill cannot be trained, and does not receive a bonus from any ability score. The DM will tell you when your reputation increases or decreases, due to actions that you make during the game.

Base Modifiers

Always: +1/2 level. Higher level characters are more likely to be known for their exploits, one way or the other.
Bard: +1. Bards talk, and are talked about. The arrival of a Bard in town is more likely to be the subject of rumours and talk than that of a Fighter or Avenger.
Character Background: +1 or +2. If you have written up a character history, your DM may give you a bonus for what the character has done before entering the game.

In Character Action Modifiers

If you do something particularly note-worthy, then it may be worth a permanent increase to Reputation. Saving a city from a goblin army, crashing the princess's ball and dancing with her all night, helping a rampaging dragon to feast on a village or impressing the crowds with magical tricks are all example ways of becoming better known.

Introduction

Make a Reputation check to be known to another character when you introduce yourself to them.

Introduction: Free Action

DC: Base DC 15.
Modifiers:
If the target is from a different civilization, +5 DC.
If the target is from a different plane, +10 DC.

Success: The target knows you. This may be worth a modifier on a subsequent Diplomacy check, either a bonus or penalty depending on the type of reaction.

Failure: The target does not know you. You cannot try again.

Bland

Fail a reputation check to get a bonus to stealth for blending in with crowds or appearing to be no one special.

Bland: Free Action upon entering an area

DC: Base DC 15.
Modifiers:
    First time in a city: +2 DC
    In a different civilization: +5 DC
    In a different plane: +10 DC

Success: You are, unfortunately, known here and gain no bonus.

Failure: You are not known, and gain a +2 bonus to Stealth for blending in with crowds and shadowing people through the streets.

This also lends itself to new options for Feats, making the reputation score more than just a "roleplaying" statistic to be ignored by the optimizer:

Superstar of Valor
Prerequisites: Bard, Virtue of Valor, Charisma 18+, Reputation 8+

Instead of granting temporary hit points based on your Consitution modifier, you instead grant temporary hit points equal to your Reputation modifier without the 1/2 level bonus.

Further feats are left as an exercise to the reader, feel free to post them in the comments! Perhaps an Assassin can become so little known that they can change form or disappear from plain sight. The Mark from a famous fighter might be harder to ignore.

In summary, by putting the reputation of the character directly in the player's hands, it encourages them to think about their actions, to perform greater acts of good or evil, or at least ones that will make for good stories afterwards, both in and out of game. By making it have mechanical effects encourages the optimizer or the combat-munchkin to engage with the story, as well as the face-man of the group. And by having the player roll the dice, they see that it has a lasting effect on the story rather than being what was going to happen anyway because the plot demanded it.