16 Jun, 2009, David Haley wrote in the 141st comment:
Votes: 0
It mainly went back to what we said a long time ago, namely that allowing people to change choices fluidly removes the perhaps undue weight placed on initial decisions. :smile:
16 Jun, 2009, elanthis wrote in the 142nd comment:
Votes: 0
David Haley said:
Not really; you could easily specialize in several things, at several points in time, with some specialties only becoming available later depending on various things that have happened. You can perhaps reduce this to some kind of complex path-sensitive multiclassing system, but a skill-less system can also be reduced to a multiclassing system, so, well, I'm not sure what has been gained.


You've gained a simpler system that has obvious paths to take instead of presenting the user with near infinite options in which only a small handful of said options are viable long term.

Quote
I think you're making a big assumption here, namely that a very small number of points in some skill is sufficient to make you "good" at it. Maybe you can swing around maces, axes, swords, knives, flails, spears and halberds without killing yourself, but that hardly means you'd be able to do much of anything terribly useful with it against very skilled opponents.


I don't think he meant that all skill systems work that only. Only his opinion that it _should_ work that way to be a worthwhile, non-confusing, player-friendly, balanced system.

Quote
I'd be interested to see a skill-based system that forces you to specialize, where all players of the same specialization don't end up with the same cookie-cutter build at the end.


I don't believe this is possible without making what would really just be a class system pretending to be a skill system and hence being complicated and ugly, or without making a system that only strongly encourages players to specialize instead of forcing them to do so (which is better than secretly wishing they specialized without indicating it in any clear way, of course). I could easily be wrong, of course.

Quote
You have stated a need for specializing (talent trees, again you can have however many different trees you wanted), and you have stated a need for if a person specs wrong without knowing what they were doing at creation they are screwed (why not just add respecing to the game).


That is a solution for a game that focuses mainly on earning the right to be powerful as the main advancement. For a lot of games that want to have at least a bit of RP involved, the idea of a man spontaneously changing from a crappy multi-element wizard into a perfectly optimized raging war machine is a little sketchy. It reduces immersion for both the character's owner and the other players, and reinforces the idea that the character is just a goofy avatar for a human player in an overly complex chat room.

The biggest problem for me is that what you propose isn't actually a real solution to the core problem: a broken set of mechanics. Sure, respecing works around what is by far the worst consequence of those broken mechanics (tons of lost time) but it doesn't fix any of the other issues. It still requires players to get on the forums and read and metagame the hell out of the mechanics to figure out the best possible combination instead of just making it intuitive, easy, obvious, and fun right from the beginning.

The built-in respecing in D&D 4ed is actually one of the many reasons I'm sticking with 3.5. ;)
16 Jun, 2009, David Haley wrote in the 143rd comment:
Votes: 0
elanthis said:
You've gained a simpler system that has obvious paths to take instead of presenting the user with near infinite options in which only a small handful of said options are viable long term.

You can make obvious paths without having classes; that's basically what you were advocating earlier with restricting choices based on other choices. You still don't have a proper classed system as you can easily cross "class boundaries".

elanthis said:
It still requires players to get on the forums and read and metagame the hell out of the mechanics to figure out the best possible combination instead of just making it intuitive, easy, obvious, and fun right from the beginning.

Here, I think you're making the assumption that this isn't part of the fun to begin with. A strategy game is in many ways about finding optimal (or at least superior) means to defeat your opponent by combining units of various types, building orders, and so forth. It's not much of a stretch to move the same game to "first-person" games. It just means that you end up with a game where the difference between experts and newbies is great, but that doesn't make an inherently bad game. As evidence it suffices to look at very many games we play IRL, chess being the most common example.
16 Jun, 2009, KaVir wrote in the 144th comment:
Votes: 0
quixadhal said:
By contrast, a class system lets you select one (or more) classes, often at character creation time, but not always. Once those classes are picked, they funnel you through their pre-defined skill set progression.

Not necessarily. A class-based system can offer as much (or little) variety as a classless system. It's possible to have classless systems in which every character is the same, or class-based systems in which every class supports dozens of different character builds.

I discussed this a couple of years ago on ..., and came to the conclusion that "classless" and "class-based" are simply two conceptual extremes, with almost every real implementation falling somewhere between the two.
17 Jun, 2009, Runter wrote in the 145th comment:
Votes: 0
KaVir said:
quixadhal said:
By contrast, a class system lets you select one (or more) classes, often at character creation time, but not always. Once those classes are picked, they funnel you through their pre-defined skill set progression.

Not necessarily. A class-based system can offer as much (or little) variety as a classless system. It's possible to have classless systems in which every character is the same, or class-based systems in which every class supports dozens of different character builds.

I discussed this a couple of years ago on ..., and came to the conclusion that "classless" and "class-based" are simply two conceptual extremes, with almost every real implementation falling somewhere between the two.


Hrm. I don't remember where I saw it but…I played a game once that was pretty open ended in terms of picking skills/spells/talents/abilities, etc. but it placed you in a "class" based on calculations of what you seemed to use the most. (Which if I remember right could change as soon as your character was built more like another of the many classes.)
17 Jun, 2009, Kline wrote in the 146th comment:
Votes: 0
Daggerfall? It had pre-built "class" choices, that were just pre-selected primary skills.
17 Jun, 2009, Runter wrote in the 147th comment:
Votes: 0
Kline said:
Daggerfall? It had pre-built "class" choices, that were just pre-selected primary skills.

I'm not sure. I seem to remember it having no class choices and it just named you on the who-list based on your current proficiencies. (Maybe gave bonuses, didn't play for long to know.)
17 Jun, 2009, David Haley wrote in the 148th comment:
Votes: 0
Daggerfall wasn't a MUD, it's one of the Elder Scrolls series of games. I haven't played Daggerfall specifically, but in Morrowind and Oblivion you first pass through an intro area where the skills you use and/or the questions you answer help the game suggest a class to you. Oblivion actually makes it quite explicit by giving you a sampling of all skills at the beginning, and then watching which ones you use in the intro; the ones you use most determine your "class". Of course, in these games, "class" only affects which skills are easier to learn than others (and which attributes are by extension easier to improve), but you're still free to practice whichever skills you like.
17 Jun, 2009, quixadhal wrote in the 149th comment:
Votes: 0
David Haley said:
elanthis said:
You've gained a simpler system that has obvious paths to take instead of presenting the user with near infinite options in which only a small handful of said options are viable long term.

You can make obvious paths without having classes; that's basically what you were advocating earlier with restricting choices based on other choices. You still don't have a proper classed system as you can easily cross "class boundaries".


Many class-based games allow this. Of course, if your game allows multi-class characters (either at creation, or by joining guilds later), there is usually a limit of how many you can join, and sometimes classes are set in opposition to each other. If you join the fighters guild, the healers won't talk to you, etc…

AD&D, at least as early as 2nd edition, allowed you to purchase cross-class skills at double the cost, although that was limited to a smaller set of "generic" skills. A mage could learn how to pick locks, but a thief could not learn how to cast knock.

To me, the difference between a skill system and a class system is that the choices I make in a skill system shouldn't be forced. In a class mechanic, maybe fire mages all get fireball at level 6. In a skill system, fireball is one of the choices you have, but you could skip it and take rain of fire instead – provided the skill set doesn't funnel you in such a way that you end up being forced to take fireball eventually because it's a pre-requisite for "greater-fireball", which is your only choice at level 40.

In a skill system, as KaVir was describing, perhaps you took the big-gun route and have one mega-attack that takes away half my health, but it takes you 10 seconds to make it work. I took a bunch of simpler skills that don't do much, but they work in 2 or 3 seconds each. If we can both do things to interrupt each other, I can win as long as I don't miss my interrupt ability roll.

As for having to get on the forums and meta-game… that can be part of the fun, but only if it isn't required until end-game. WoW actually did something right there…. you can jump on and roll a character and play however you like, using garbage gear because it looks better, taking "stupid" talents.. and until you hit the level cap, you can still have fun and be successful. Once you hit end-game, you then HAVE to start reading forums and collecting the right kinds of gear, and spec'ing your character in a small handful of ways.. or you will be substantially weaker than your peers. In PvP, that means you die often. In raids, that means you don't get invited unless they need warm bodies to fill up the spots. So the casual player can enjoy leveling, while the hardcore player tries to climb the hill and do all the hardest instances on "heroic" setting.
17 Jun, 2009, flumpy wrote in the 150th comment:
Votes: 0
Can we try and steer this back to the topic of the post a little?

A class vs classless discussion is useful, but we need to remember this is in the context of "are bigger or smaller numbers better".

Ok I'll ask a couple of questions that interest me: How do you prevent numbers, big or small, from dominating your class systems? What can be done to prevent exploitation apart from hiding them? What strategies can game designers use to prevent the paper scissors stone kind of games we see?

[edit]

Ooh and the most interesting question I have is how do you decide on what numbers to use in the first place? Would I have to draw bell curves and graphs and the like to make sure things are balanced?
17 Jun, 2009, KaVir wrote in the 151st comment:
Votes: 0
Kline said:
Daggerfall? It had pre-built "class" choices, that were just pre-selected primary skills.

That's also the approach I use. Every character starts with 50 stat points, 250 skill points, 250 fighting style points, and 3 talents.

You can go into custom character creation and distribute those points as you please, and afterwards the mud will allocate you a set of starting equipment optimised for your setup. It sounded good on paper, but I've yet to see first-time player create a character that was anything more than mediocre.

To get around that, I introduced an alternative character creation system, allowing players to select from a range of predesigned concepts. Each concept has the same total starting attributes as a custom character, but I carefully optimised them (in some cases actually basing them on the setups of veteran players) and designed each around a certain theme (Samurai, Dragon Rider, Pyromancer, etc).

Runter said:
I seem to remember it having no class choices and it just named you on the who-list based on your current proficiencies.

I do that as well (at least for characters that haven't yet classed), basing the titles on your talents. For example a player with Dark Lineage is listed as a 'Dhampir' (or 'Winged Dhampir' if they also have House NightWing), a player with Assassin Training is listed as an 'Assassin', etc. However I only really added it to provide some variety to the who list, as I'm not a fan of letting players set their own titles, descriptions, etc.

quixadhal said:
To me, the difference between a skill system and a class system is that the choices I make in a skill system shouldn't be forced. In a class mechanic, maybe fire mages all get fireball at level 6. In a skill system, fireball is one of the choices you have, but you could skip it and take rain of fire instead – provided the skill set doesn't funnel you in such a way that you end up being forced to take fireball eventually because it's a pre-requisite for "greater-fireball", which is your only choice at level 40.

In Diablo II, the Sorceress can pick from ice, lightning and fire spells, and there is no "one true way" of building a character; spells are chosen, not forced.

However the other classes (such as Barbarian, Necromancer, Paladin, etc) cannot learn any of the Sorceress' spells - each ability is clearly assigned to one specific class, and you cannot gain the abilities of a different class.

I would consider that a class-based system, wouldn't you?
17 Jun, 2009, quixadhal wrote in the 152nd comment:
Votes: 0
Assuming you're developing a brand new combat system, and not implementing an existing one in the context of a MUD, I think the first things you want to think about are:

Do you want players to gain numbers as they level, or different skills, or both?

By that, I mean you can design your system so that a level 10 player and a level 20 player do about the same physical damage, but the level 20 will swing more often, have lots of extra attacks they can throw between the auto-attack, and lots of extra mitigation skills to avoid incoming attacks. Or, you can stick to the tried-and-blue higher level players have damage that scales, so a level 10 hits for 20hp per swing, while the level 20 has stats + gear that let him hit for 200hp per swing.

How long should an average fight last?

Should a typical fight between two even-con opponents take 30 seconds, 2 minutes, 15 minutes? Should the fight be quick and decisive, or should the players have time to work on their tactics and adjust to one another?
Do you want a thief to be able to one-shot backstab people?
Do you want anyone to be able to one-shot things?
Should anyone have abilities that take control away from their opponent? IE: stun-lock, sleep, charm?
What's the ratio of hit to miss that your players will tolerate?
Do you have active mitigation skills (block, parry), or just "armour class"?
Does your armour actually absorb/deflect blows, or is it a part of the dodge calculation?

All of these will affect how your numbers play out. It's very possible to have a small numbers game that's interesting, but there have to be other rewards for players to gain. New skills, new stances or techniques to modify their existing skills, whatever.

As for rock-paper-scissors…. how bad is that? Obviously we don't like it when it's that simple, but do you want your game to allow any kind of immunity or requirements of a particular action to win? Should anyone be able to do anything solo? Do you want to make some content require a group, and should that group have different types of people in it?

Another question to ponder, because I worked out a system many years ago to try and answer it, is how do you deal with the maniac hero? That is, you have a hero who saunters into town and starts slaughtering villagers. How many villagers should he be able to fight at once? In most MUD's, he can fight however many there are, because they do 0-4hp of damage to him, and he kills them in one or two blows. That's not sensible. Yet it also doesn't make sense that a seasoned veteran in magical plate mail and with god-forged weapons shouldn't be able to kill a half dozen farmers with pitchforks without spilling his drink. How can you factor the number of foes in so that a tank can't pull dozens (or hundreds!) of foes at once, but CAN beat a small bunch easily? Do you WANT to solve this "problem", or embrace it as some games do?
17 Jun, 2009, elanthis wrote in the 153rd comment:
Votes: 0
flumpy said:
Can we try and steer this back to the topic of the post a little?


You must be new to the Internet. :p

Quote
Ok I'll ask a couple of questions that interest me: How do you prevent numbers, big or small, from dominating your class systems?


Don't rely on them much if at all. Too many games are based on the idea that power comes from having a bigger number than the other guy. The difference between an Orc and a Giant boils down to the Giant having 10x as much HP, doing 10x as much damage, etc. There's no reason at all that you have to use numbers for this, though.

For example, instead of giving the giant bigger numbers for damage, assign a size class to weapons. The giant is swinging a Huge club, which causes more serious damage when it strikes most characters, cannot be blocked (do you think a shield would stop a Mac truck from crushing you?), etc. Now, when players get more powerful, instead of giving them bigger numbers to counter-act the giant's big numbers, give them a selection of abilities that negates the giant's advantages. Give them an ability like Block Huge Weapons that lets their shield work against the giant's club. Give them an ability like Giant Slayer that lets them do normal damage against a Huge opponent. You can split it up quite a bit, of course, and mix-and-match which abilities are needed to make each opponent manageable (a giant fire elemental may not use physical attacks, so Block Huge Weapon is not necessary to be able to fight it, but Giant Slayer might still be needed). At some point it all breaks down to numbers because we're using an abstract game system, but you can reduce and remove numbers to a surprising degree. Get rid of HP and go with a wound/critical system, and damage then just becomes a weighted randomization to determine what level of wound or critical is received. Ger rid of Mana and use a D&D-style slot/preparation system, or a cooldown system, or a component/ritual system, or a combo/pattern system (a challenging minigame that must be completed to cast a spell).

By far the best thing (IMO) about using a largely ability-based system is that it gets rid of the content-invalidation present in most every other RPG, MMO, and MUD. It closes the gaps between new and long-time players, too. The problem with a number system is that in order for the player to get big enough numbers for the giant to become a reasonable challenge that player ends up getting numbers so high that the early-content challenges like orcs become entirely unchallenging. A level 20 player has no challenge at all in a level 5 area. Using the ability based system, the player gains the power to defeat more powerful opponents without invalidating the challenge of less powerful opponents, and so the player can still have fun exploring "newbie" areas. That in turn gives far more incentive for the game designers to introduce new areas that both high and low level players can explore instead of only ever introducing new top-tier areas like WoW does. That then makes it far more reasonable for me to join a game my friend has been playing for 6 months because there is going to be content that we can both enjoy playing together while I work on gaining the levels to adventure in the high level areas.

You can of course use a hybrid approach. I don't think a game that never allows players to get stronger against the early game opponents would fly with most RPG players, and I don't blame them either. While a high level player might be able to easily handle 10 orcs while a new player could only handle 1 or 2, that's a stark difference from most games where the high-level player can wade through an army of orcs with no fear at all.

A pure ability system does require a bit more ingenuity. Far too many RPG style games differentiate between the low-level spells and high-level spells purely with damage dealt, for instance. Fire Bolt does 20 damage while Fireball does 50 and Fire Storm does 100. Going with a more ability-based system, the higher level spells don't do a lot more damage (in fact, they may do less!) but they are more versatile. Fire Bolt hits one opponent for 20 damage, Fireball hits a small group of opponents for 15 each, and Fire Storm ravages an area for 2 damage a second for 5 seconds. (Specific numbers pulled out of my ass.) Or with a wound/critical system, Fire Bolt may have a higher critical rating than Fireball or Fire Storm.

That then has the further advantage of never invalidating abilities. Those abilities you learn early on remain useful even in the end game, unlike most RPGs where your wizard ends up with a spellbook with 100 spells in which only 10 are worth holding on to.

That is, of course, still entirely unrelated to my original question about big vs small numbers, so we're still off topic. ;)

Quote
What can be done to prevent exploitation apart from hiding them?


Hiding them prevents nothing. Go look up the FAQs for most console RPGs. The players manage to figure out relatively complex formulas, inclding things like (100 - (dmg - def / 2 + luck / 5)) * (crit ^ 2 / 3). It's ridiculous how much time people put into crunching the numbers, but people do it. Basically, security through obscurity DOES NOT WORK.

To answer your question, then, it's unfortunately going to be complicated, especially for complex system that allow a lot of customization. One of the biggest "fixes" I know of is to prevent excessive stacking of bonuses and multipliers. Some games put caps on bonuses, but that is useless IMO because all it means is that the min-maxers figure out how to reach the cap and stay there 100% of the time. A better approach I think is what D&D does, which simply disallows a good deal of bonus stacking. Basically, each bonus has a type, and you can have only one bonus of each type for each stat at a time, with the highest bonus taking effect. The item and spell system is then designed to make it difficult to get all the bonuses at the same time. For example, AC has the base armor bonus, the Dex bonus, a deflection bonus, a natural armor bonus, and a couple others. Magic armor provides a deflection bonus, as do 95% of AC-boosting spells, meaning that you can't try to min-max magic armor and spells. The only item to grant a permanent natural AC bonus uses the neck item slot, which is also the only slot that can give certain other types of bonus (CON bonuses, for example), so the player has to pick between getting a permanent natural AC bonus or a permanent CON bonus. Which one he picks will depend on his class and play style, so while there may be an "optimal" equipment for particular builds, but there is no way to get a perfect build for all purposes.

Quote
What strategies can game designers use to prevent the paper scissors stone kind of games we see?


I think a question to ask first is whether you want to prevent that. In a game where characters don't have different strengths and weaknesses there is little incentive for teamwork. At best it just turns into a "we need 10 people to do this" and off you go. Specialization might mean that character A is always going to lose in 1-on-1 combat against character B, but it means that team-oriented play focuses on building effective groups and players learning tactics suited to their party role.

If your game is mostly a simple non-team PvP environment, you may want to avoid the rock-paper-scissors issue, but otherwise I think you want to encourage it.

So far as fixing it, just don't use specialization. Encourage players to get all abilities, or make different specializations' abilities most equivalent. Don't make rogues worse tanks than fighters, don't make mages better damage-dealers than clerics, don't make clerics better healers than rogues, etc. You'll probably end up with the worst case of cookie-cutter characters you can imagine, though. :)

Quote
Ooh and the most interesting question I have is how do you decide on what numbers to use in the first place? Would I have to draw bell curves and graphs and the like to make sure things are balanced?


Depends on the importance of the numbers to begin with. In general, though, you're going to need to apply a lot of math, so yes. Most of all though you need a lot of play testing, and you need to be ready to tweak both the rules and the world itself. I would highly recommend building off of an existing tested and balanced system, preferably one that's already similar to where you want to go with yours.

Last night I was continuing the discussion with some friends, and one brought up the case of Ragnorak Online. Apparently mages in that game get a choice of one of three high-end spells, breaking down an ice-based spell, a fire-based spell, and a thunder-based spell. The three were in general considered very balanced, yet almost everyone always picked the thunder-based spell. The reason for that was that almost all of the end-game monsters had a weakness to thunder magic.

So, the raw numbers don't matter, because the world itself can unbalance the numbers in surprising ways. Without doing a lot of manual documentation or building a very comprehensive set of game analysis tools, you're likely going to end up with that issue, even at levels below the end-game. For example, if the best mid-level area to gain XP or gain the bets equipment is fire themed, you can safely bet that most players will be optimizing for ice magic and ice weaponry at the mid levels. Unless you're prepared to generate content for every possible theme at every possible level, you either need to design the abilities around the world or design the world around the abilities. That means that either you just make all the mid-level abilities and items ice themed so there's no worry about screwing players who picked something else before getting to the fire dungeon, or it means avoiding excessive theme-based dungeons and making sure that any set of abilities is going to be enough to get through any dungeon.

On a similar note, one of my soon-to-be colleagues brought up a very astute observation. If your game is going to include PvP at all, design around that first, and the PvE content last. The reason for that is pretty simple and makes a world of sense. If you design for PvE and then find out that you have to tweak for PvP, then once you're done you have to go back and redesign all the PvE content around the changes you made for PvP. On the other hand, if you design for PvP first and get that balanced and squared around, you only need to design and develop the PvE content once based around the existing PvP abilities you gave characters. Almost every mixed PvE/PvP game designs around the PvE content first and hence almost every single one has been stuck with an almost endless cycle of rebalancing characters between PvP and PvE content. That advice is a bit hard for a lot of us because many of us have no interest in PvP, but we want to include it in some capacity in our games in order to appeal to the PvPers… we work on the PvE stuff first because it's more fun and add in the PvP later, and end up stuck with that endless cycle of imbalance.
20 Jul, 2009, KaVir wrote in the 154th comment:
Votes: 0
David Haley said:
elanthis said:
Put simply, a min-maxer should not be able to create a better character than someone who's just making choices based on how fun he believes the options are.

I heartily disagree with this statement. The logical conclusion of your argument is that essentially random choices should yield a character just as powerful as choices made very deliberately, which strikes me as a far more broken game design.

The above comments were made around five weeks ago, and they got me thinking. I put my ideas down on (virtual) paper, they turned into a design, and that lead to some prototype code, and last night I introduced a new class to my mud - Titans.

The concept is fairly simple. Whereas the effectiveness of most classes is based on specific combinations of powers, Titans can just add anything they like. This is because the Titan powers themselves don't add raw power - instead, they provide utility options or alternative attack types, things which are useful but which don't make the character any more powerful.

However Titans do gain a wide range of bonuses based on the total of all their power ranks. So, just like the other classes, they become more powerful as their train new powers - the only difference is that it doesn't matter which powers they train. While the other classes are trying to carefully balance each and every power rank to eke out every possible advantage, the Titans can just pick whatever they feel like without needing to worry too much about it.

I've also decided to incorporate another contraversial design decision that I normally disagree with - I'm hiding all the numbers. The help files contain only flavour text. However in this case I don't think it's such a big deal, simply because you can't really go wrong.

The class is aimed primarily at newbies and first-time players, who often have trouble building a decent character. Initial feedback has been positive, and although I think it unlikely that Titans will hold the attention of the veteran players for long (at least compared to the other classes), it does appear to be filling a niche. I wouldn't want to make this the standard approach for my classes, but as a one-off it seems to be working out quite well.
20 Jul, 2009, David Haley wrote in the 155th comment:
Votes: 0
KaVir said:
instead, they provide utility options or alternative attack types, things which are useful but which don't make the character any more powerful.

How does this work? That is, how can something be "useful" but not increase the character's power or effectiveness?
20 Jul, 2009, KaVir wrote in the 156th comment:
Votes: 0
David Haley said:
KaVir said:
instead, they provide utility options or alternative attack types, things which are useful but which don't make the character any more powerful.

How does this work? That is, how can something be "useful" but not increase the character's power or effectiveness?

They increase the character's effectiveness, but not their raw power. For example 'Tremor Sense' allows you to sense creatures around you (both invisible and outside line-of-sight), 'Hammer Fists' strengthens your unarmed hand attacks (but only enough to put them on-par with weapons), 'Mighty Leap' provides you with a pounce attack (on-par with weapons) and allows you to leap great distances, and so on.

Obviously it's useful being able to use your fists if your weapon gets broken, or being able to sense other players nearby, or being able to leap great distances when exploring - but these abilities won't directly make you more powerful. Pit two Titans against each other in a typical toe-to-toe combat situation and their choice of powers should be fairly irrelevant.
20 Jul, 2009, David Haley wrote in the 157th comment:
Votes: 0
Oh. I guess we're not meaning the same thing by 'power'. I care little for size of numbers when measuring 'power'; to me the only thing that matters is effectiveness.
20 Jul, 2009, elanthis wrote in the 158th comment:
Votes: 0
The general gist of what KaVir did and what I was advocating, David, is that abilities should not just be a pile of numbers, but should have some kind of effect. If you have two Fire Attack powers, one that does 10 damage and one that does 100 damage, then one is clearly more powerful and selecting the former is a waste of time, implementing the former is a waste of time, documenting the former is a waste of time, and the power either gets ignored by veterans or mistakenly chosen by inexperienced players, neither of which is in any way a useful contribution to the game experience.

If instead your two Fire Attack powers do identical damage but differentiate themselves on effect – say, one does a straight line of damage 100' long by 5' wide while the other does a 40' radius sphere of damage – then suddenly the selection of the two powers is far more interesting. One is more useful in some circumstances while the other is more useful in different circumstances. At that point your further game design choices may invalidate one or the other, of course, since if most of your dungeons are straight corridors then the spherical fire attack is far less useful, which is why game design is something that needs to be done top-down and not piece-by-piece in isolation.

KaVir's game is a numbers-heavy game, however. Nothing but effects would be useless to a class. So he made it so that simply gaining power makes the character more powerful (shocking, I know) instead of requiring the player of the Titan class to choose between "powerful" or "lame" (but where the options are not clearly spelled out, of course, and the player is required to spend hours reading forums and sites looking for optimal builds by people with too much time on their hands who've tried a bazillion different builds).

It's a good compromise that lets the people who just want to play without investing huge amounts of time into understanding low-level mechanics enjoy his game while simultaneously allowing the tweakers to get their rocks off playing "advanced" classes. It's really pretty similar to a class-based game that has a "build your own class" feature, which I've always been a fan of (the concept, at least – I haven't seen a game really pull it off well yet).
20 Jul, 2009, David Haley wrote in the 159th comment:
Votes: 0
elanthis said:
If you have two Fire Attack powers, one that does 10 damage and one that does 100 damage, then one is clearly more powerful and selecting the former is a waste of time, implementing the former is a waste of time, documenting the former is a waste of time, and the power either gets ignored by veterans or mistakenly chosen by inexperienced players, neither of which is in any way a useful contribution to the game experience.

I never disagreed with this. It's almost a non-statement as a truism. :wink:

My issue is with systems where choices made at random will be just as effective as choices made deliberately. In this case the choices are meaningless and nothing more than illusions of different options, or chrome on top of all the same stuff, basically.
20 Jul, 2009, KaVir wrote in the 160th comment:
Votes: 0
David Haley said:
My issue is with systems where choices made at random will be just as effective as choices made deliberately.

That's actually pretty much what I've done; you can pick any combination of powers, and overall your character will be (approximately) as effective in combat as any other Titan of the same 'level' (bold used for emphasis).

David Haley said:
In this case the choices are meaningless and nothing more than illusions of different options, or chrome on top of all the same stuff, basically.

But they're not - the choices provide you with utility and variation. In some ways the Titans have even more freedom of choice than the other classes, as they're not forced to pick their abilities based on character optimisation.
140.0/213