Tobold's Blog
Saturday, May 22, 2010
 
Perfect MMORPG: Stats

As you might have glimpsed from my first "Perfect MMORPG" post yesterday, the basic motto of this hypothetic MMORPG which would be perfect for me is an old Sid Meier quote: "A good game is a series of interesting decisions". (I never found a source for that quote, although on the GDC 2010 Sid mentioned having coined the phrase "interesting decisions".)

In another post this week I asked for which enchant to put on the boots of a holy paladin. I was really interested in the answer, albeit more the "why" than the "what", but I also had an ulterior motive: Being well aware that the choice would only make a tiny difference, I also wanted to see how sure people were of the right answer, how authorative it was, or whether anyone would say "well, it's your choice".

Which brings me to the stats of my hypothetical perfect MMORPG: I believe in a perfect MMORPG there should be a lot of valid choices for what stats to choose. That is to say the stats should be designed in a way that there are very few completely useless stats for any class, and several "best" stats, which are all completely valid choices, depending on playstyle. Thus instead of having a simple formula telling you that for your class strength is better than agility, the two should be equivalent, and both be useful in different situations, so you could make an interesting decision.
Comments:
I don't think i've seen a rpg where min/maxing wasn't possible.
 
Well in WoW healers used to have genuine choices between regen and straight-up healing power.

Of course that's disappeared due to mudflation but it's been promised a return in Cataclysm.
 
I'd argue that we already do have plenty of near-optimal specs in WOW, it's just that no such system can withstand the relentless onslaught of the theorycrafters. There will always be a slightly better spec and a slightly worse one (even if they are close) and people will flock to that.
 
So the question would be one of playstyle... for example, if the choice was between attacking faster, or hitting harder, but where it was valid to go either way, as the dps would be the same.

But in some situations you'd need want to hit faster (bosses which have shields that come down for brief periods) or harder (bosses with low armor, where power would multiply more).

In the end, min/maxing would still occur, but the variables, and the different situations would make it difficult to have a single "way" of doing things.
 
I'd argue that we already do have plenty of near-optimal specs in WOW, it's just that no such system can withstand the relentless onslaught of the theorycrafters.

This.
Except for the fact that it is just the current WOW-system that has problems. Change encounters in raids in such a way that they require more from dd than dps, more from healers than hps and more from tanks than effective health (eh) and there you go.
 
I think Tobold could accept that it's not possible to make every stat 100% equal, but the current WoW dev philosophy (classes don't need to be exactly even, yet all must be competitive) applies pretty well here. Stunlockers and bursters don't need to be 100% equal in all scenarios... but they each need to have a clear use.

A rogue who chose to stack agility/evasion could tank a tough add, while a rogue who stacked crit could burst him down instead. This has been the argument behind the big-sword/small-shield or small-sword/big-shield debate, the same reason rogues with lots of stun/incap abilities aren't allowed to have high burst potential.

Meanwhile, if you have a burst healer, his job would be easy to heal the tanking rogue. Since a few heavy hits would kill said rogue, a bomber like a holy pally would fit the bill perfectly, such as with current-day raidtank healing. However, the encounter would have to be tuned to be completed with a regen/longevity healer. Tuning mob damage/hit-chance so low could trivialize the encounter if you have a burst healer... which I don't see a problem with, considering everything is always easier with a perfect makeup. You only need to make it possible for the longevity healer, not necessarily forgiving or easy.


Since the math equations for X agi = 1 str (for dps conversions) have been known for a long time now for wow, it would not be entirely unruly to simply up the value of agi to the classes where X is too high, til hitting ratio of 1:1. This would, of course, break pvp, but then I never had the illusion that pvp and pve should even be in the same game.
 
the problem here is that i think tobold is actually pretty much quoting blizzard's design philosophy.

wow's warlock class has been stacking haste rather than crit throughout most of the current expansion because the most reliable theorycraft suggested that is what is best. if a warlock stacked crit the chances are it would account for a 5% difference in their output which is too small a figure for anyone to notice at the end of a single raiding night, you would need far more data to be able to find a difference that isn't effected by RNG.

and yet, if that warlock's gear were inspected another player might decide that they don't know what they're doing.
so what i'm saying is that blizzard have basically achieved this goal and the players have rejected it.

from my own point of view, i nearly replied to tobold's 'spirit for pallies' post to say that while the choice is unimportant it's the reasons for making it that matter. returning to the warlock example the warlock that stacks crit is likely doing that because they haven't read up on the class and don't understand the implications of their choice.

as a raid leader i want players who do give their character some thought outside the game. who do read up on them and find out the intricacies. the actual choice they make at the end of that process is much less important to me than the enthusiasm for understanding.

so, returning to the 'pally with spirit' thing, if i saw a pally with that enchant i would probably be one of the ones who mentions it (though i'm not in the habit of inspecting everyone i meet). i wouldn't be saying "lol, spirit noob!" but i would assume that the pally has misunderstood how useful the stat is for them.

so, from a developer's PoV i can totally see the importance of interesting decisions. from a players PoV i think players get a lot of fun from understanding why one decision is better than the other and making it for an informed reason.
 
just wanted to add:
i started that post with "the problem here..." which sounds very negative.
i'm thoroughly enjoying the points you're raising with these posts tobold ^_^
 
Theory craft is a big "problem" for this kind of thing. You either choose between homogenized choices, which all make your character effectively the same, or you make things different, but they will never be precisely as efficient. When raiders are willing to spend the time and effort to make sure they are at 100% efficiency, there isn't much room for choice. Sure, Arcane and Fire mages might both be "viable" (this is just an example, I don't know what the numbers are), but if Arcane does 2% more damage than Fire, almost 100% of raiding mages will go Arcane.
 
The problem is purity of roles. The rogue who can survive a bit longer is useless, because a tank would do the job 10x better, and will do the job, so all the rogue did was gimp his damage. If players had to be more self-reliant, then we'd see more diversity of stats to take into account the defensive uses.
 
that would probably require a big shift away from the "holy trinity" of tank-healer-dps though. that would be interesting but would require some truly epic ingenuity the like of which we've not seen to date.
 
Another thing I'd like to see in a perfect MMORPG is stats that have an immediate and obvious effect on the gameplay.

In WoW, since there are so many items and so many upgrades, each individual decision is between, say, a fraction of a chance to dodge and a smidgeon of additional health. That single choice is extremely unlikely to make a difference by itself, but all of those choices, in the aggregate, will affect your playstyle.

This aspect of WoW's gameplay is yet another reason why gear decisions boil down to spreadsheets and theorycrafting. Since each individual decision has only a miniscule effect on your character, you need to do some math offline in order to figure out what choices you want to make in the aggregate.

What I'd love to see is a game where equipping those new boots made an obvious difference in how you played your character. I'm not sure how to achieve this, but one possibility is to make items that don't have quantitative effects like +10 stamina; rather, equipping some items could add completely new abilities.

Another possibility is to boil the bonuses on quantitative items down to a single stat. That way, each new item automatically improves the one stat you care about, and you can choose what that stat will be (dodge or stamina?). Now you can get small quantitative upgrades, but since you must apply them all to a single stat, the choice of stat is an interesting decision.

Actually, now that I think about it, Blizzard has been experimenting with both of these ideas. There are ICC drops like the staff that spawns Val'kyrs to fight for you; I think I remember reading that Blizzard was deliberately experimenting with "cool" weapons like this. They also have the mastery stat in Cataclysm, which is similar to the "one stat" above. However, I suspect that with these changes patched into place late in the game, it'll be too little to change the overall flavor of WoW's gear decisions. It's a great opportunity for their next MMO, though!
 
I think that it is literally impossible, at least for damage dealing. Speaking of which, I don't know if you have ever said you are happy or want to change the whole three-headed monster of DPS, Tanking, and Healing or not. But assuming it has those roles, or ones close to them...

Tanking and Healing are probably possible. Healing has the dynamic between output, efficiency, and longevity that, if done right, could mean that all three are viable choices and yet not directly comparable to overall effectiveness. In the past this was true in WoW, and may yet be true next expansion.

Tanking is the dynamic between survival and threat, although this has changed and may yet change. Threat isn't popular for some people as a mechanic. Survival right now can be summed up as "time to live" sadly and various types (Avoidance, effective health...which is made of armor and health) can all be reduced down to your time to live. This, however, I can see being expanded if you have lots of very common damage types or something that are all commonly at play, though that may just lead to specialization and tanks having to farm different sets for each encounter...unless you make multiple damage types done at various times in each fight.

But damage dealers I just don't see how you can ever be summed up in any other way than a number. Even "burst" vs "longevity" can be made into an "effective DPS" if the length of the encounter is a known factor, although I guess if the length was HIGHLY variable in the same fight somehow they could maybe be separated. Not even survivability can really be separated from DPS since death as an outcome in the fight just means you are doing 0 damage for a lot of seconds. Maybe if you had more interaction between health levels and damage done (positively or negatively), but then I can easily see it being hard to avoid twitchy gameplay.

I honestly think that this may be a pet peeve you can't get rid of though, Tobold. Min/maxing has been around as long as stats have been around. It is such a sacred cow that I don't even think something can BE an RPG and not have any min/maxing. I mean, yeah, I guess you don't have to remove it entirely since you just want to have more choices, not tons of choices. Hmmm.
 

But damage dealers I just don't see how you can ever be summed up in any other way than a number.


It's actually pretty simple...
Remember Razorgore?

1) All calsses have defensive abilities that they can skill. Just do not balance heal output (and AE heal) around the assumption that they didn't skill any of them.

2) Do not always add (uncredible) enrage timers to bosses, but use e.g. spawning adds that need to be defeated, but cannot all be tanked. (Scale back aoe tanking/aoe dmg). Not every enemy in a raid should require to be tanked.

3) Make (untankable) adds have damage spikes that need to be countered by whoever has aggro: Blink, 1H+shield, etc.

The defensive and fun abilities are already there, because WoW would be rather .. boring without them - and because people like to do something else in PvP in addition to a perfect dmg rotation. These abilities might need to be balanced a bit, but not much.
 
Just by the way:

This also makes WoW a little bit more about the game mechanics and less about dancing.
 
I think the better way to do this might be to have lots of different stats and make it clear that the optimal distribution is to try to balance them all. But leave players lots of choice on how to do it.

Something like that (how to get to an optimal number) is more interesting to me than 'haste or spellpower? I'll just look it up'
 
i'm trying to relate these points to the WoW we have...

if everyone has a defensive cooldown they can pop and that becomes an expected part of every encounter (or even just several encounters) then is anything more interesting happening?
DPSers work out the best talent points to drop so they can get the defensive cooldown without losing too much DPS. a boss mods addon tells everyone when to press the button and that's it.
when doing hard modes personal responsibility is a big part of raiding including responsibility for the damage you take. at appropriate gear levels i think the same is probably true of normal modes tbh.

few bosses have tight enrage timers. the ones that do (festergut has an actual timer, blood queen has one that comes from the game mechanics) are DPS test fights.
ICC has a broad range of fights and i think it's only right that some of them test the DPSers ability to do their job under difficult circumstances.
as a DPSer i want those fights.

there are adds in various dungeons that have no threat table. armour and health values being what they are in the current game (this is changing in cataclysm) those mobs can't do lots of damage without one-shotting clothies.
if you're asking for personal responsibility for DPSers then i think it does exist and this isn't really a good way to achieve more of it. cloth wearers have to be able to avoid getting hit.

i think i'm just not grasping the whole issue here. the talk of disliking choreography. disliking the "dancing", but also not liking the idea of settling into perfect rotations.
10 or 11 out of the 12 bosses in ICC prevent me from doing my ideal rotation for more than 20 seconds at a time. some have me "dancing" or moving out of the way of environmental damage. some have me reacting by grouping with (or avoiding) other specific players. some have me reacting to debuffs. some have me using cooldowns ar special abilities. some have me simply moving my character to a different position in the room.

i guess i'm saying that i don't see how these are bad (or, in fact, particularly different to what is being asked for) and i certainly don't see how the alternatives presented would be better.
 
Alf, there are several advantages of making class abilities important in raiding:

1) You can switch classes and suddenly need to do something else. This adds to 'replay' value.

2) Player feel like they actually play their class.

3) It is an incredibly difficult, actually outright impossible, mathematical problem to determine the best spec if defensive abilities matter. The only reason EJ can do it today is, because they only need to maximize one number. Even that is mathematical difficult, but no challenge for stochastic/evolving algorithms combined with 3 GHz quadcores.

4) Encounters aren't the same every week you do them. Your actions are (slightly) chaotic: They differ each time; even emergent behaviour becomes possible.

if everyone has a defensive cooldown they can pop and that becomes an expected part of every encounter (or even just several encounters) then is anything more interesting happening?

You don't really understand: Fighting Razorgore I could 'save' sombody else, by frostnova/sheeping/snaring/silencing, and even drawing aggro. This kind of interaction between characters has been removed. The only interaction left is heal (one-directional) and some dance moves that require you to (not) stand where anybody else stands.

a boss mods addon tells everyone when to press the button and that's it.

That would be possible only if the encounter is trivial. If some add has silenced a mage, he cannot blink. He would need to do something else, ask for healing in TS, get help etc.
If somebody who needs to do something very specific has been CCed somebody else needs to fill his role..

Remember Razorgore, first phase. These kinds of fights are chaotic without being confusing (in a good raid).

Perhaps the pace would need to be reduced, but that's the price if you want players to think during the fight and judge situations that have been to many to think through in advance.
 
don't get me wrong, i understand that clearly defined classes is a good thing but i don't really see that this is lacking in the current game. i'd accept that it could be stronger but it is there.

i have 7 level 80s (no-lifer etc ^_^) 5 of which play in the top level content. i have very different things i do on each of those characters.
going from one to another doesn't at all feel like i go from pressing 1-2-3-1-2-3 to 3-2-1-3-2-1.

a couple of examples:
-my druid healer has to remove curses on lady deathwhisper. if i get CCd (mind control) i have to call on vent for a mage to do it. he concentrates on raid healing with HoTs.
-my pally healer can use divine sacrifice to reduce the damage taken by the whole raid for the next 12 seconds. he concentrates on healing the tanks with direct heals.

that's just the healers. my rogue has stuns that he uses every raid. he has tricks of the trade to give threat to the tank. he can vanish to remove agro. my hunter has a silence he uses every raid. he has an ability to prevent incoming damage for 5 seconds. my warlock has a teleport which gets used every fight. he even tanks a boss in one of the encounters!

all of them have different specs that bring different things so my warlock has a raid buffing spec and a survival/burst spec.
in terms of choreography, while each class might not have to stand in very different places each role does and each encounter will ask for specific reactions from different players that they don't have to do any other time.

i think that a lot of what you're asking for exists in the game right now. the fact that most people (PuG groups) usually wait until they over-gear the encounter so that they can brute force the fight without using any of these things doesn't mean it's not there and not required for progression.

as for the point about theorycraft not existing when defence is important, you're basically talking about PvP there. that is the way it is due to the unpredictability of the opponent. i'm pretty sure theorycraft would still be very strong in a game where defence mattered as long as the encounters were still scripted.
that maybe returns to yesterdays points but i don't want to play in a sand-box. in this case i want the theme-park. i want my experience to be tailor-made to be a fun and exciting challenge.
 
"I also wanted to see how sure people were of the right answer, how authorative it was, or whether anyone would say "well, it's your choice"."

I missed that thread, but if I'd have seen it, my reply would have been "Who cares?".

Mrs Bhagpuss and I were chatting today about why we tend to lose interest in progressing our characters once we reach maximum level. One of the key reasons is that this is the point when improving your stats ceases to be about really obvious choices and starts to require thought. And research. And planning.

I don't know about the rest of you, but when I come home from work and sit down to relax, I don't look forward to a good evening's statistical analysis. I especially don't want to have to get out a calculator and read five websites just to work out which pair of boots I should wear. I just want to run around and hit things, look at pretty scenery and go "gosh! wow! that's amazing!" a lot.

I don't want to make "interesting decisions". Interesting decisions aren't interesting in this context, they are fiddle-faddling tedium. Far from making any decisions where gear is concerned, I want it to be inevitable what stats I should focus on from the moment I choose my class and blindingly obvious which of any two items is better for any given character.

The ideal MMO would have a very limited number of stats. They would be extremely simple to understand. They would change relatively infrequently but when they did it would make a large and immediately obvious difference.

MMOs are far to complicated without ever being sufficiently complex. Less really is more, ask Alex Toth.
 
don't get me wrong, i understand that clearly defined classes is a good thing but i don't really see that this is lacking in the current game. i'd accept that it could be stronger but it is there.

I agree that DD classes (we are talking about DDs right now, are we?) do not feel identical. And, like you, I say that they could feel more distinct.

There is a difference between dd-warrior and rogue, but it just isn't very big in raids.



as for the point about theorycraft not existing when defence is important, you're basically talking about PvP there. that is the way it is due to the unpredictability of the opponent. i'm pretty sure theorycraft would still be very strong in a game where defence mattered as long as the encounters were still scripted.


I do not talk about sandboxes here at all.
PvE can be made more unpredictable - even without using more random numbers than today (crits,resists etc).

Seems like this really touches more subjects than I thought.

- classes feel more distinct
- less dance-feeling
- less predictability
- less min-maxing, because it's simply much harder/impossible to achive.
- encounters feel more different each week.

The most important thing here really is the min/maxing. There will always by min/maxing, but there are two different kinds:

- Min/maxing without knowing a theoretical optimum: Razorgore.
- Min/maxing while knowing a theoretical(and practical optimum).

I condemnt the last type. If min/maxing leads to measureing of a 'deviation from perfection' cookie-forced cutter specs find their way into the game.

I've written extensively about the deviation from perfection on my own blog.
 
I don't want to make "interesting decisions".

In that case we simply need to play different games :).

I also miss some 40 man raids that allowed you to slack if you wanted. However, I also want to have the possibility to use my brain when playing a game.

Ironically, I used my brain much more often in vanilla/parts of BC, when there was no RAWR around. But since most people didn't, nobody was forced to do it.
Nowadays you just load your char in RAWR and get your answers. It's trivial, but without alternative.

In my opinion a game should allow me to spend some time without any challenges (farming) and also allow me to get into some statistical analysis if I want to. Vanilla WoW was like that, by the way.
 
hehe, that's all pretty understandable though, massive geek that i am, i'm a sucker for a graph or two and i love the theorycraft.

however, continuing to use wow as a reference point most can relate to, in the case of tobold's paladin there aren't that many decisions to make _really_.
the gear that is clearly made for paladin healers (plate gear with spellpower) doesn't have any spirit on. if a player were to simply take any gear they get that fits the plate+spellpower model, upgrading any piece if it dropped from more advanced content, then they'd do perfectly well.

blizz added enchanting and gems (and forthcoming reforging) to give people the choice to customise things if they want to. if you're not interested in that then you could, again, do very well by gemming anything that you already know to be good, the stuff that's already on all your gear.

i understand that you may have bad experiences where "that guy" tells you you're an idiot for making a decision you made but any social interaction can leave you with good and bad experiences. there's always bad eggs in every group.

i guess i'm defending the current model again and perhaps coming across something of a fanboy.
i do think, though, that most of the points being raised are actually fairly well catered for and we're blaming the developers for social problems. there's no doubt that the dev's can have a big impact on social interaction but we have to take some responsibility too. however it is that you want to play when you get to max level, i guarantee there's a guild out there somewhere that wants to do the same. if there isn't you could make one and i'm sure you'd find people who wanted to be part of that kind of atmosphere.
 
the previous post was @baghpuss

now @nils
the talk on cataclysm that's interested me most has been around the talent trees losing all the mandatory but boring talents (+5% damage etc.) leaving more space for fun ones.

i think their thinking is that cookie-cutters will always exist (they've given up the fight ^_^), however they said that their aim is to have the cookie-cutter creator say "here's what's best then you can put the last 10 points any where you like".

this will allow us to take defensive abilities or just fun ones that we would like to take but need to get that 2% damage or lose our raid spot (for example).

if their plan for larger health pools and relatively smaller heals and greater onus on efficiency and longevity pan out then we might end up quite close to what you're hoping for.

there's a big IF here though. it's a fine balance to strike and i'll be very interested to see how they do with it.
 
if their plan for larger health pools and relatively smaller heals and greater onus on efficiency and longevity pan out then we might end up quite close to what you're hoping for.

I also have been hopeful about Cataclysm. I also think that Blizzard partyly agrees with my analysis.

As soon as healers can have mana problems in PvE again, DDs will automatically feel a pressure to not take is much damage.

Skills of arkane mages that grant 140 resi against everything (~30% less magical damage) will suddenly look OP and might even have to be nerfed; while nowadays those who specc into these skills are cooled noobs ;)
 
It would be great to customize my class to whatever specs I want and not have people tell me that my character is built wrong.

Having interesting decisions that actually matter makes for a good game. Decisions that don't make a difference are pretty boring.
 
As long as the AI is predictable and you can prepare your equipment for a specific pre-defined encounter I don't see how what you say is possible. The only way to make it work is have some kind of metagame, which can be done by:
-The market: if X equipment becomes popular is becomes more expensive which makes using it less enticing (only works if equipment needs replacing).
-PvP: if X equipment becomes popular people start looking for counters to that equipment.
-More factors to juggle: the more factors you have to juggle the less clear any one "right" source becomes.
-Less information: the less information that players have about what they're facing the harder it is to min-max their equipment, with raids people know EXACTLY what they'll be facing but if you inject more uncertainty things become much less clear which opens to door to more interesting choices.
 
As long as the AI is predictable and you can prepare your equipment for a specific pre-defined encounter I don't see how what you say is possible.

That is why this post isn't about changing WoW stats or EQ2 stats or some other stats of existing games. Interesting decisions in stats require combat not to be 100% predictable, e.g. various spells and abilities doing different amounts of damage at different ranges. Thus not only is it more interesting to decide which ability to use, but also the choice of the stats affecting those abilities gets more interesting.

If combat is highly dynamic instead of always the same for every mob, there are no spreadsheets and formulas to calculate the optimum, because there *is* no optimum.
 
This comment has been removed by the author.
 
As long as the AI is predictable and you can prepare your equipment for a specific pre-defined encounter I don't see how what you say is possible.

We need to differentiate between theoretically predictable and, well, predictable.
Chess is theoretically predictable, but not actually predictable; not by a human mind.

If the PvE game mechanics are well chosen, unpredictability is possible.

Important to note: Rules do not even need to be many or complex to lead to unpredictability. Again, chess is an example.
 
Tobold: "That is why this post isn't about changing WoW stats or EQ2 stats or some other stats of existing games."

Exactly, that's what I was saying. Basically raid difficulty comes from various sources (as you've pointed out in previous points) but NOT from having to deal with anything out of the ordinary. If you make raids (as we know them) harder it'll just focus people down more and more narrowly focused lines.

Even with something like Civilization, the random map can bork a lot of plans and the interplay between different AI adds a lot of complexity in ways that MMORPGs mostly lack.
 
sically raid difficulty comes from various sources (as you've pointed out in previous points) but NOT from having to deal with anything out of the ordinary. If you make raids (as we know them) harder it'll just focus people down more and more narrowly focused lines.

Mmh, I disagree. Quite obviously (?) you can make raids harder by adding non-ordinary tasks for the players. I dislike that, but that doesn't mean that it is impossible (?).

Also, the most prominent reason for those 'narrowly focused lines' is not that they are necessary, but that they are possible.

The people who search for other players with GS>5500 for ICC 10 do not do so, because it is necessary. They do it, because they can.

The combination of openly known itemlevels, numerical simulations like RAWR and some common sense allow mods like GS to judge people. GS isn't perfect, but unlessd you deliberately try to cheat it, it works quite well.
 
Tobold says:

"Which brings me to the stats of my hypothetical perfect MMORPG: I believe in a perfect MMORPG there should be a lot of valid choices for what stats to choose. That is to say the stats should be designed in a way that there are very few completely useless stats for any class, and several "best" stats, which are all completely valid choices, depending on playstyle. Thus instead of having a simple formula telling you that for your class strength is better than agility, the two should be equivalent, and both be useful in different situations, so you could make an interesting decision."

I'm not sure that offering this sort of bifurcation is desirable or even original. We already have class and spec distinctions; adding "playstyle" seems like just another layer. So instead of Hunters using the "agil for Survival, ArP for Marksmanship" rule of thumb, it would become something like "agil for small/fast hit Survival, ArP for big/slow hit Survival, haste for blah blah blah".

Making different stats useful in different kinds of fights could, however, be interesting (but maybe annoying). To some degree, tanks do this now by stacking stam/armor/avoidance/resistance depending on the kind of damage involved in a fight. Though stamina is usually the right choice, there are occasional fights where other stats serve a use (eg, shield block vs Anub, frost resist vs Sindragosa, threat gear vs Hodir). Providing this kind of decision-making for dps would, I suspect, be hard to do without some classes being favored in a given fight.
 
you could make stats a more interesting decision if they changed what your character did. Predominately strength stats? You're a warrior. Predominately intellect stats? You're a mage. So instead of starting character development on the class-selection screen or practicing "skills," you could have progression move by stats instead. That would be an interesting decision!

But assuming that some random mish-mash of stats will affect your "playstyle" though is kinda dumb. First, because there will always be something that's x% "superior" in a fight and everyone will min/max to that. And even if you made things totally random or unpredictable (all stat-playstyles do the exact same output, all boss moves are totally random) then it really doesn't matter what players choose to begin with, and thus the choice isn't very interesting at all. It's just random preference/ignorance.
 
nd even if you made things totally random or unpredictable (all stat-playstyles do the exact same output, all boss moves are totally random) then it really doesn't matter what players choose to begin with, and thus the choice isn't very interesting at all. It's just random preference/ignorance.

Mmh, seems like I am writing against a wall of ignorance, no offence ;)

Let me give you a MMO example of 100% unpredictability und 0% randomness.

5 players, 5 mobs, no tanks, 3 waves.

The first 5 mobs run to the players. Players select mobs by using TAB or the mouse. Mobs always run to the guy with the most aggro. Some of the players need to avoid getting hit, others have a dead zone.

Since humans aren't machines they will select different mobs and they will do slighty different amounts of damage to these mobs, because they press the 'dmg' button with different delays. You do not even need latency here. Human reaction time is enough. The mobs run to different players on the first wave.

When the second wave starts, the players already stand in different positions. They select different mobs with TAB/mouseclick. They do different amounts of damage and need different amounts of time to select a mob.

Finally, the positioning of the players when the third wave starts is 100% unpredictable. Now, imagine this game mechanic: The guy, next to the mobs of the third wave when they spawn, is attacked until they are dead; he cannot lose their aggro.

Unless players actively try to create order here, the fight will be 100% unpredictable.

It looks chaotic and yet it is not random. This is called chaos theory. The butterfly that causes a hurrican some 10000 km away.

This example even works without crits/miss/etc.
Add internet latency if you want.

Actually, Blizzard goes to great lengths to actually make encounters predictable. It is actually HARD.


About theoretical unpredictability and randomness: Within Newtonian pysics there isn't even randomness. You need to go to quantum physics to even find it - and even there it is questionable whether we see true randomness or the unpredictability of underlying, but unknown, processes.

Even generating random numbers with a computer is a science. It is not easy. And those involved in this science will tell you, that these random numbers aren't even random at all. They are just unpredictable by humans and adhere to some stochastic distribution.

So:
1) Ignoreing quantum physics there isn't even randomness anywhere in the world, let alone in an MMO.

2) Even ignoring this, MMOs do not even need 'random number generators' to create unpredictability. Slight differences in human behaviour create unpredictability on their own within a very short amount of time.

3) Blizzard actually uses single boss mobs and tanks to create some kind of preditability: If every bit of an encounter is unpredictable it is totally unfun; even if there is no randomness at all.


And once again the easiest example of them all: Although chess is totally not-random, every match is totally unpredictable.
 
Short explanation why I consider the distinction important:

Randomness smells like "arbitrariness".
Players hate this.

What most people who ask for randomness actually want is not randomness, but unpredictability.

The random number generator of the server isn't even necessary to generate unpredictability. Even if it could generate 'random' numbers, which it cannot.
 
I've talked to guildies about this issue, and I'll highlight the class/spec where Blizzard has got this right: the Holy Priest.

A Holy Priest values 4 stats almost equally: Spellpower/Haste/Int/Spirit.
Crit/Mp5 are considered "2nd rate".

There will be Holy Priests who value throughput (Spellpower/Haste) but will do it differently: some gem Spellpower, others gem Haste. Then, there are some Holy Priests who prefer regen (Int/Spirit), although this line of thinking is less common (I have a feeling the EJ Holy Priest thread favours regen though).

In terms of playstyle, there are similarly 2 schools. The direct healing style: Flash Heal to build Serendipity stacks then Prayer of Healing. The other style is the renew spam style.

It is interesting that the renew-style was looked down upon by some with scorn back in Ulduar; so much so that some Holy Priests would say that people who used Renew were bad players! This playstyle has now come to the fore, and I believe most of the top Holy Priests have switched to this method.
 
For some reason "casual" players always complain about min-maxing.

Thing is you can never avoid gimp templates (or overpowered ones) unless you do away with whole choice altogether . Which essentially is what wow did . That removes the entire layer from the game - layer of designing your character

I for one always enjoyed the part of designing a new template, something which fits best my play style yet still powerful. It is probably one of my favorite gameplay elements . -I enjoy it everywhere I play. mechwarrior was one of my favorite fps cause you could do lots of customization (something most fps lack)


Out of MMOS example shadowbane had incredible class customization you could build wildly different things. Magic casting rogue bard? dwarf warrior dagger thrower? -werewolf warrior ranger shooting crossbows? -you got it (and all those were very viable specs) - that on top of your "standard" classes (and it had it all - pures ,hybrids and all flavors in between and on edges)

Of course some things were gimp- its inevitable when you have choice. Solution is to provide easy way to fix mistakes (SB failed here - there was no way to respec up until game was dead, which obviously was too late ) , not to remove choice altogether
 
Post a Comment

<< Home
Newer›  ‹Older

  Powered by Blogger   Free Page Rank Tool