Menu
News
All News
Dungeons & Dragons
Level Up: Advanced 5th Edition
Pathfinder
Starfinder
Warhammer
2d20 System
Year Zero Engine
Industry News
Reviews
Dragon Reflections
White Dwarf Reflections
Columns
Weekly Digests
Weekly News Digest
Freebies, Sales & Bundles
RPG Print News
RPG Crowdfunding News
Game Content
ENterplanetary DimENsions
Mythological Figures
Opinion
Worlds of Design
Peregrine's Nest
RPG Evolution
Other Columns
From the Freelancing Frontline
Monster ENcyclopedia
WotC/TSR Alumni Look Back
4 Hours w/RSD (Ryan Dancey)
The Road to 3E (Jonathan Tweet)
Greenwood's Realms (Ed Greenwood)
Drawmij's TSR (Jim Ward)
Community
Forums & Topics
Forum List
Latest Posts
Forum list
*Dungeons & Dragons
Level Up: Advanced 5th Edition
D&D Older Editions
*TTRPGs General
*Pathfinder & Starfinder
EN Publishing
*Geek Talk & Media
Search forums
Chat/Discord
Resources
Wiki
Pages
Latest activity
Media
New media
New comments
Search media
Downloads
Latest reviews
Search resources
EN Publishing
Store
EN5ider
Adventures in ZEITGEIST
Awfully Cheerful Engine
What's OLD is NEW
Judge Dredd & The Worlds Of 2000AD
War of the Burning Sky
Level Up: Advanced 5E
Events & Releases
Upcoming Events
Private Events
Featured Events
Socials!
EN Publishing
Twitter
BlueSky
Facebook
Instagram
EN World
BlueSky
YouTube
Facebook
Twitter
Twitch
Podcast
Features
Top 5 RPGs Compiled Charts 2004-Present
Adventure Game Industry Market Research Summary (RPGs) V1.0
Ryan Dancey: Acquiring TSR
Q&A With Gary Gygax
D&D Rules FAQs
TSR, WotC, & Paizo: A Comparative History
D&D Pronunciation Guide
Million Dollar TTRPG Kickstarters
Tabletop RPG Podcast Hall of Fame
Eric Noah's Unofficial D&D 3rd Edition News
D&D in the Mainstream
D&D & RPG History
About Morrus
Log in
Register
What's new
Search
Search
Search titles only
By:
Forums & Topics
Forum List
Latest Posts
Forum list
*Dungeons & Dragons
Level Up: Advanced 5th Edition
D&D Older Editions
*TTRPGs General
*Pathfinder & Starfinder
EN Publishing
*Geek Talk & Media
Search forums
Chat/Discord
Menu
Log in
Register
Install the app
Install
Community
General Tabletop Discussion
*Pathfinder & Starfinder
Monte Carlo versus "The Math"
JavaScript is disabled. For a better experience, please enable JavaScript in your browser before proceeding.
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an
alternative browser
.
Reply to thread
Message
<blockquote data-quote="Truename" data-source="post: 5025170" data-attributes="member: 78255"><p><strong>About the AI</strong></p><p></p><p>Okay, about the AI, and why it's so stupid. Short version: Computers are stupid. <img src="https://cdn.jsdelivr.net/joypixels/assets/8.0/png/unicode/64/1f600.png" class="smilie smilie--emoji" loading="lazy" width="64" height="64" alt=":D" title="Big grin :D" data-smilie="8"data-shortname=":D" /></p><p></p><p>(Slightly longer version: ...and I'm not an AI programmer.)</p><p></p><p>Cheekiness aside, the big challenge here is creating an AI that doesn't have to be rewritten for every character and every level. Remember, our goal isn't to create a dwarf that fights <em>intelligently</em>; it's to create a dwarf that fights <em>consistently</em> across 30 levels. Consistently low performance is just as good as consistently high performance as long as all powers, magic items, and so forth get equal love.</p><p></p><p>In practice, though, that means that the 'bash it in the head' powers are used to their potential more easily than the 'set up a big combo' powers right now, so they may be weighted more favorably. So we do need <em>some</em> AI. But I want to remind you that we don't need <em>great</em> or even good AI. Just consistent AI.</p><p></p><p>What I have the ability to do, easily, right now, is this:</p><p></p><p></p><ul> <li data-xf-list-type="ul">Have powers be turned off (not chosen) until certain conditions are true. For example, Second Wind is set to :do_not_use until the dwarf has received at least a healing surge worth of damage.</li> <li data-xf-list-type="ul"> Have powers set to :low priority, which is the same as :do_not_use, except that the power isn't turned off entirely. :Low priority powers will be used if no :normal powers are available.</li> <li data-xf-list-type="ul">Make decisions based on perfect information. For example, I can give Brute Strike :low priority if the enemy has fewer than 4 hit points remaining.</li> <li data-xf-list-type="ul">Change the overall, non-power-specific rules about power selection. For example, I can have all dailies fire before encounters, or all standard actions before minor actions.</li> </ul><p></p><p>What I want to avoid is anything that increases maintenance costs. In particular, I don't want to have to reprogram the AI in order to test different builds, power combinations, or levels. This means avoiding:</p><p></p><p></p><ul> <li data-xf-list-type="ul"> Complex logic with a lot of special case rules. ("If this power is active, but that one isn't, and the monster is such-and-such, don't use this power.") This is hard to test thoroughly and will lead to hard-to-find bugs. A few special cases are easy to test, and potentially okay (for example, I've programmed ":do_not_use Boundless Endurance if Second Wind is still available"), but I don't want to program in every combination of every power.</li> <li data-xf-list-type="ul"> Decisions based on assumptions about other powers being present. For example, we could :do_not_use Brash Strike when we don't have temp hit points. But what if we don't have any powers that grant temp hp in this build?</li> <li data-xf-list-type="ul">Giving certain powers higher than normal priority. This will require constant tweaking as we change builds and levels. For example, giving Crushing Surge a hypothetical :high priority when the dwarf has no temp hit points might be appropriate now, but is it still appropriate at level 25? Is it appropriate for all fighters of any build?</li> <li data-xf-list-type="ul"> Power combos, such as "use power X, then action point, then power Y." Is that combo still appropriate at all levels? Does it apply to all builds?</li> </ul><p></p><p>The key here is keeping AI programming costs <em>low</em>. Remember, the quality of the AI isn't that important as long as the creatures fight consistently across levels. We're looking at how survivability changes from level to level, not the absolute number. So, how can we best achieve that goal given the above constraints? Creative solutions welcome.</p></blockquote><p></p>
[QUOTE="Truename, post: 5025170, member: 78255"] [b]About the AI[/b] Okay, about the AI, and why it's so stupid. Short version: Computers are stupid. :D (Slightly longer version: ...and I'm not an AI programmer.) Cheekiness aside, the big challenge here is creating an AI that doesn't have to be rewritten for every character and every level. Remember, our goal isn't to create a dwarf that fights [I]intelligently[/I]; it's to create a dwarf that fights [I]consistently[/I] across 30 levels. Consistently low performance is just as good as consistently high performance as long as all powers, magic items, and so forth get equal love. In practice, though, that means that the 'bash it in the head' powers are used to their potential more easily than the 'set up a big combo' powers right now, so they may be weighted more favorably. So we do need [I]some[/I] AI. But I want to remind you that we don't need [I]great[/I] or even good AI. Just consistent AI. What I have the ability to do, easily, right now, is this: [LIST] [*]Have powers be turned off (not chosen) until certain conditions are true. For example, Second Wind is set to :do_not_use until the dwarf has received at least a healing surge worth of damage. [*] Have powers set to :low priority, which is the same as :do_not_use, except that the power isn't turned off entirely. :Low priority powers will be used if no :normal powers are available. [*]Make decisions based on perfect information. For example, I can give Brute Strike :low priority if the enemy has fewer than 4 hit points remaining. [*]Change the overall, non-power-specific rules about power selection. For example, I can have all dailies fire before encounters, or all standard actions before minor actions. [/LIST] What I want to avoid is anything that increases maintenance costs. In particular, I don't want to have to reprogram the AI in order to test different builds, power combinations, or levels. This means avoiding: [LIST] [*] Complex logic with a lot of special case rules. ("If this power is active, but that one isn't, and the monster is such-and-such, don't use this power.") This is hard to test thoroughly and will lead to hard-to-find bugs. A few special cases are easy to test, and potentially okay (for example, I've programmed ":do_not_use Boundless Endurance if Second Wind is still available"), but I don't want to program in every combination of every power. [*] Decisions based on assumptions about other powers being present. For example, we could :do_not_use Brash Strike when we don't have temp hit points. But what if we don't have any powers that grant temp hp in this build? [*]Giving certain powers higher than normal priority. This will require constant tweaking as we change builds and levels. For example, giving Crushing Surge a hypothetical :high priority when the dwarf has no temp hit points might be appropriate now, but is it still appropriate at level 25? Is it appropriate for all fighters of any build? [*] Power combos, such as "use power X, then action point, then power Y." Is that combo still appropriate at all levels? Does it apply to all builds? [/LIST] The key here is keeping AI programming costs [I]low[/I]. Remember, the quality of the AI isn't that important as long as the creatures fight consistently across levels. We're looking at how survivability changes from level to level, not the absolute number. So, how can we best achieve that goal given the above constraints? Creative solutions welcome. [/QUOTE]
Insert quotes…
Verification
Post reply
Community
General Tabletop Discussion
*Pathfinder & Starfinder
Monte Carlo versus "The Math"
Top