CMUd ideas

[Peijen]

One possibility I don't think I've seen mentioned is not to differentiate between skills and attributes. Treat the attributes as very general (and presumeably very costly) skills. For example, one point more of offensive dexterity or one point more of sword skill each increase your chance to hit with a sword by the same amount. However, the sword skill does nothing when you wield an axe, but if your willing to limit your weapon choice you'll be able to increase your sword skill faster.

Sort of. Most feats are affected by both attribute and skill. Here is the design reason, to balance between casual player and hard-core player. In the beginning attribute are more costly then skills. However the cost of attribute is constant (not really, but for now let's say it is) while the cost of skills increase with skill level. This way casual player can increase their skill to a point where it's more costly then attribute, and continue to increase attribute after that point. With high attributes and numerious average skills player can still be effective (generalist) compare to high attributes with small sets of highly developed skills (specialist). While spcialist still has an advantage, it's not overwhelming

You could also add an intermediate level of skills for weapon groups (like slashing or bludgeoning).

I like this idea. In fact I think this solves the problem I was having with how to handle unarmed combat.

Also, do you have more notes on how Seal of Evil's system works?

[Jonathan]

Analysis in the lab reveals that my PS is non-functional. I'll get the svn server up sometime after I get a new one.

Say what you will about lab techs, you must respect someone who can put together a Pentium III system in less than 60 seconds after being asked, "Uh, do we have any P6 systems sitting around?"

[George]

Seal of Evil's webpage
http://eng.objectgames.com/

It turns out that only the witch's skills are based on using a particular weapon element with a particular skill. The other characters are pretty much standard diablo. The assassin (who hasn't assassinated anyone) uses combos. The paladin has a bunch of special strikes. The wizard summons stuff. I never bother with the beastman's skills since his normal attack is just fine. I do still like the idea of a skill having greater effectiveness if used with the right type of weapon and the right spell active.

Boots unlock armor, necklaces unlock helms, rings unlock weapons. Metal unlocks water unlocks wood unlocks fire unlocks earth unlocks metal. To equip equipment of a particular element, your score in that element must be above some level, so most of my characters wear armor, helms and weapons of a single element and necklaces, rings, and boots of the unlocking element. I accidentally discovered an alternate way to unlock powers. If the helm, boots, and armor are made of the same material, then they unlock the powers and add hidden powers. I've only managed to do this on two characters, and I haven't figured out how necklaces, rings, or weapons fit in. The weird part is that the same materials have the same elements so the boots don't unlock the armor on an elemental level but you still get the same powers and more.

Item creation is damn complicated. What little I understand of it is still too complicated to explain here. As bad as creation is, "refining" (upgrading an existing item) is even worse. And there are so few stores that you are required to create most of your equipment.

[Peijen]

Ok, I have worked out the combat skills. Based on George's suggestion.

Combat Style
1H Melee Weapon & Shield
2H Melee Weapon
2 Melee Weapons
1H Range Weapon & Shield
2H Range Weapon
2 Range Weapons
Unarmed*

Weapon Mastry
Dagger
Sword
Axe
Mace
Polearm
Spear
Bow
Firearm
Unarmed*

Defensive
Parry

*I decided to make unarmed both combat style and weapon mastry, which might be over powering...

[Jonathan]

Is shield not a defensive skill? Or is it part of parry?

Also, how do these skills fit in your equation?

(50% + attack_skill / 2000) * (1 - block%) * (1 + (dex - agi) / 100) * (1 + bonus%)

[Peijen]

Combat Math

I haven't factor in item bonus, but I haven't figure out how the customization is going to work UI wise so.

Attributes ranges from 10 to 150, and skills go from 0 to 1000.

Multipliers

Code: Select all

	                       damage_multiplier    attacks / round
1H Melee Weapon & Shield        4                      4
2H Melee Weapon                 6                      3
2 Melee Weapons                 3                      6
Unarmed                         3                     10
1H Range Weapon & Shield        3                      4
2H Range Weapon                 5                      3
2 Range Weapons                 2                      7

Code: Select all

                 block_multiplier
                 Melee        Range        Magic
Shield            0.05         0.08         0.01
Melee weapon      0.025        0.04         0
Range weapon      0            0            0

Only highest block multiplier applies

Code: Select all

                armor_multiplier
        Piercing   Bludgeoning   Slashing
Light      4           5            3
Medium     3           4            5
Heavy      5           3            4

To Hit = (50% + (attacker_combat_style - defender_combat_style + attacker_weapon_mastry) / 2000) * (1 - (parry * block_multiplier)) * ((dexterity - agility) / 100)
Damage = ((strength * damage_multiplier) - (defense * armor_multiplier) / 20) * (weapon_mastry / 4000) * rand(weapon_damage_range)

[Peijen]

Combat rounds

Turn-base for now. Each round has 12 phases, a character can't have more than 12 actions. So increase attack rate is probably not useful for unarmed attacker, but works well for 2 handed weapon. Range attack always goes first (to simulate ranged attack).

Code: Select all

    Order ===>
Phase 1    2 Range 1    1H Range 1    2H Range 1
Phase 2    2 Range 2                                Unarmed 1    2 Melee 1
Phase 3    2 Range 3    1H Range 2                  Unarmed 2                 1H Melee 1
Phase 4    2 Range 4                  2H Range 2    Unarmed 3    2 Melee 2                  2H Melee 1
Phase 5    2 Range 5    1H Range 3                  Unarmed 4
Phase 6    2 Range 6                                Unarmed 5    2 Melee 3    1H Melee 2
Phase 7    2 Range 7    1H Range 4    2H Range 3    Unarmed 6
Phase 8                                             Unarmed 7    2 Melee 4                  2H Melee 2
Phase 9                                             Unarmed 8                 1H Melee 3
Phase 10                                            Unarmed 9    2 Melee 5
Phase 11                                            Unarmed 10
Phase 12                                                         2 Melee 6    1H Melee 4    2H Melee 3

[Peijen]

Is shield not a defensive skill? Or is it part of parry?

Also, how do these skills fit in your equation?

(50% + attack_skill / 2000) * (1 - block%) * (1 + (dex - agi) / 100) * (1 + bonus%)

Takes a while to format them tables

[Jonathan]

http://jonathan.pearce.name/svn/cmud/

Okay, SVN server is up. You can't commit yet and there aren't any files in the repository. Also, I don't make any guarantees that I'll migrate the data from its current location to its final resting place. It is, however, a proof of concept.

[Jonathan]

The above url redirects you to the actual url which is currently

http://24.20.141.176/svn/cmud/

This is subject to change. Use the j.p.n URL and hopefully it'll work. Success! Works for Dave.

[Jonathan]

Why can't phase 12 and phase 11 be merged?

[Peijen]

Why can't phase 12 and phase 11 be merged?

Because 12 sounds like a good number. Really, there is no good reason.

[George]

The 12 phase to a round system showed up in the Hero System tabletop RPG. It used all 12 because if some guy acting on phase 11 shot and killed a guy who was about to act on 12, then the guy on 12 didn't get to act. It was subdividing phases by dex rather than attack type, so that kind of logic may not apply here.

One dangerous thing about a system like this is that a single extra attack per round for a slow character (I.e., 2H weapon) is much more valuable than a single extra attack for a faster character. I recommend that extra attack speed costs be calculated based on the percentage increase rather than the delta.

[Peijen]

The 12 phase to a round system showed up in the Hero System tabletop RPG.

Hehehe

I was going to use speed originally, but I didn't want to deal with fractional attacks per round for now (when I change it to real time system I intend to use speed attribute for attack rate)

I recommend that extra attack speed costs be calculated based on the percentage increase rather than the delta.

To be sure, you mean like 50% attack rate increase is 3 extra attack for 2 weapons and 1.5 extra attack for 2h weapon? That's what I have in mind, and if that's not what you mean let me know.

[George]

Well, like if you have a 2 attacks per round and buy one more attack per round, it's a 50% boost. If you have 3, it's only a 33% boost. So the 2 attack guy should have to pay 50/33 times the 3 attack guy for his extra attack.

Alternatively, buy extra attack rate by percentages rather than whole attacks. So 2 attack guy has to buy 50% to get his extra attack while 3 attack guy has to buy 33% for his. Both pay the same per percentage point. This would work better once your combat system is real time.

[Jonathan]

Brilliant MUD idea which doesn't matter at all for a GUI game!

When in combat, let people move around or flee. If you flee successfully, you can run to other rooms exactly as one would expect. But your opponent has a chance to prevent your flee attempt, based on number of opponent you face, etc. If you just move around, though, you can always do it automatically, but your opponents get insta-oracle-tracking that lets them follow you perfectly, if they so desire. Also, during movement, you only get one move per round and must use the move instead of attacking or other action. Fleeing you can spam.

[Jonathan]

As far as I can tell, nobody ever posted ideas about crafting. Here's my quick thoughts.

Crafting should cater to two different guys: an adventurer who knows what kind of specific eq he wants and is willing to shell out to make it custom, and a merchant who wants to sell eq to other players and is willing to produce a great deal. Therefore, your crafting mechanic should be workable if you want to produce just one sword or if you want to produce hundreds. This means you need optional automation. Also, you should be able to craft from loot or automatically from either 1. a standing order for loot in an auction house or 2. a controllable resource like a mine or NPC (which might possibly be subject to PvP battles). Craft merchants shouldn't enter the game to hunt, or assemble widgets, or train useless skills, but to determine their supply lines and product mix.

And I think the best crafting system would let people mix and match components to create nifty new artifacts.

[George]

Crafting should cater to ... an adventurer who knows what kind of specific eq he wants and is willing to shell out to make it custom

I'm not convinced this is practical, since a self-crafter would have to spend enormous amounts of time training his craft skill just to get one or two items out of it. Instead, I'd propose a varient of one of your other ideas. Let players create the "blueprint" for their custom item and hire another player or NPC to actually create it. Essentially, you open up the item creation interface to anyone regardless of skill level. They can set up the desired specs for the item, and the interface will generate a blueprint item. Each blueprint requires some minimum crafting skill and ingredients to build. You generate your blueprint, hunt down the needed ingredients, and find a dedicated crafter. You pay the crafter and give them our blueprint, which they can "use" to automatically create an item matching your specs. Now anyone can get exactly the custom equipment they want (if they have the money) without needing to train their skill or camp the auction house. You could also post blueprints at the auction house and let crafters bid to make it (kind of a more specific version of your request suggestion).

[George]

Something I thought of today. I was trying to figure out how to keep the game engine generic enough that the front end could be a telnet client, a custom text client, a 2D graphical client, or a full blown 3D client. There's the obvious problem of position, which I haven't fully solved yet. Movement aside, though, all other interaction between the client and server can be generic enough to allow any of the above clients, except for telnet. Telnet requires that the server parse the input and format the output. That means that it can't share an implementation with a 3D engine where all I/O happens on the client.

ANd then I had my brilliant idea. Implement a telnet proxy server. The game server communicates with the proxy using generic command and status messages, just like it would with a rich client. The proxy contains all of the game's text context. So when a status message comes from the generic server, the proxy translates it into the appropriate text for the telnet client. When the telnet client gives some input, the proxy parses it and translates it to a generic command message to the server.

Now, as long as the command and status messages are well defined and stable, players can develop their own clients (similar to WoW's ability to customize the interface). Players are free to use the basic telnet client to get the default interfaces, but you could also have a richer client that shows character status, inventory, the combat log, chat, and anything else in separate GUI windows.

Ideally I'd like to figure out a way that different players on the same server can use text, 2D, and 3D interchageably (mostly because it would make experimentation a lot easier), but I have to figure out the movement/position issue first.

[Peijen]

Hehe, that's what I have in mind.

I will create a Game.Command namespace that server as proxy between front end to back end. It not only allow me to switch text/2d/3d client but also allows for new system.

Of course the framework is far from complete, but I think we are thinking alike on most issues which is a good thing.