[MUD-Dev] Expected value and standard deviation.

[Katie Lukas]

Me: After an absence of a couple of days, this particular thread has
piled up.  Thus, I am responding to an earlier message in the
thread, because I think that Mr. Koster's reply has pointed out some
interesting items, and would like to reply to those before tackling
the rest of the messages.  Apologies in advance for (the, holy crap
I wrote like miles over here, really extreme) length, but it seemed
that snipping much of the replies would remove some valid context.

>From Raph Koster:
> From: Katie Lukas
>>  And Raph Koster wrote:

>>>      There's no way to get rid of the "boring" way to play your
>>>      game. Players can always choose to play conservatively to
>>>      maximize return while minimizing risk.

>>  I'm not sure that this doesn't move us into Elephant Territory -
>>  as in the Elephant in the Room.  Personally, I think that the
>>  above statements reflect some central assumptions about what
>>  constitutes good (or fun) gameplay that I'm not so sure are
>>  accurate, or at least not accurate for enough people.

> I must say that the fact that players prefer to play a boring way
> that gives them advancement over a fun way that gives slower
> advancement seems to be well-proven over decades of online games.

> Here are the assumptions I am operating under: players seeking
> advancement will be driving towards optimal advancement. Optimal
> advancement will include making the activity as predictable as
> possible. Predictable activities become less fun over time.

> Generally, any given game has multiple near-optimal paths. One of
> these paths will be the one that minimizes risk to the point of it
> being non-existent. It's a slow but steady guaranteed return on
> time invested.  It's also "boring" in the sense that there is no
> challenge to it, no risk, and little variation in the activity.

> Most likely, this path is NOT one designed intentionally by the
> designer.  It's one that players find by manipulating the system.

I am absolutely *beyond* agreement about the statement that the
players will always find the optimal path to a goal, and that even
past that they will find any means to an end that is faster than any
other means regardless of whether that was a design choice.  I
personally believe that to underestimate the players and their
ability to heat-seek optimal (in terms of time) methods to any
game-related goal is virtually suicide on the part of the
developer(s).

That said, there are a couple of ideas that I find worth tackling
here.  First is pretty central - the fact that optimal in games is
measured in time.  Meaning - in other parts of life, the fastest
route is not always optimal (for various reasons, many of which are
purely subjective to the individual choosing the route).  Second,
that subjectivity alone is largely (not wholly, but largely)
eliminated in games - the goal itself is chosen by the developers.

Third, the risk-reward equation.  To reduce that to a simple
statement: The fastest route to an objective is a compound of the
length of time of the activites with the probability that the risk
is high enough to make delay nearly inevitable.  That is - players
have to consider the time that would be required in a failure
circumstance when considering the best route to take.  This also
relates directly to the "predictable = optimal = boring" statement
made above.

I personally believe that each of these items and their
ramifications has not been overwhelmingly explored in search of more
interesting game mechanics, especially as pertains to simply
multiplying possibilities.

  1. Shortest Time = Most Optimal Means to an End.

    The first problem with this statement is that game developers
    often seem to forget that time, to a player, means every single
    minute spent online, not just the time taken in a particular
    activity. There are many examples of game areas in MMORPGs that
    clearly had no small amount of time and effort put forth to make
    them attractive, full of mobs that drop useful stuff, and
    engaging at various skill levels - but that are desolate and
    deserted 90% of the time.  Why? Because it takes too long to get
    to them, especially for the lower-level characters they are
    designed for that have less money for transport mechanisms and
    speed spells/items.  Often, some more enterprising souls will go
    there just because they know it's always uncamped, but the
    numbers stay small.  Even areas that are relatively easy to get
    to, but that are surrounded by mobs that are higher than the
    mobs in the objective area will often be deserted.  There are a
    couple of examples of both in DAoC.

    The second problem with this is that I'm not sure it's an ideal
    condition.  Even apart from the purely business downside
    (shortest time = less money), it makes me wonder if we're
    looking at the whole equation wrong.  Since this is the case, it
    means we are more than adequately rewarding those who accomplish
    things more quickly than average, and inadequately rewarding
    those who take longer than average.  I think there are a lot of
    ways to resolve that problem, but that the problem itself is an
    incredibly threatening one to both developers and players, since
    it is impossible to bring up without the age-old question of how
    to reward players that play more (who are also those who can do
    things more quickly, go figure :) AND casual gamers so that both
    are engaged, interested, and feel adequately rewarded for their
    time.

    Some of the ways I could see the problem being addressed are:

      - An incremental reward system that involves more player
      choice as per how and when the rewards are granted.  Meaning,
      a player could choose to "store" points, not unlike a gambling
      situation, and granted a larger reward as long as the player
      managed to stay alive. This is certainly correlative to the
      problem with risky activities causing longer timeframes, but
      adds an element of player choice directly, as well as
      introducing what could become almost a level of
      competitiveness with the mobs themselves.  Not unlike players
      saving cash and using poor items in order to be able to buy a
      specific item later on.  Or, allowing players to literally
      gamble, to a certain extent - I'll wager half-a-bub that I can
      win this fight.

      - An extreme multiplier being added to the number of
      reasonably equivalent development paths, such that a player
      could vary activities with absolutely no penalty whatsoever.
      Easiest example is PvE XP advancement: by multiplying the
      number of kinds of mobs (and kinds of fights) *in the same
      area* by a factor of, say, five, could scale into
      time-equivalent but vastly different advancement courses.

      - Advancement itself becoming a factor of more complex
      activities.  A small example of this is DAoC's camp bonus - if
      you stay in the same area for x amount of time, your xp drops
      by y% until it hits a floor or until you move.  I would
      advocate the reverse system: your xp maintains unless/until
      you do something (whatever that might be) differently, at
      which point it increases by a percentage.  This could also
      involve tying some level of xp to non-traditional activities,
      such as mentoring or exploration.  Varying the kinds of
      activities within a single category (different kinds of
      fighting, etc) would allow this to occur without angering or
      reverse-boring players that do not like some of the available
      activities (many players, I think, would balk at being forced
      to craft, or to explore the map, at least if that forcing was
      overt).

  2. Lack of Subjective Goals

    Nearly every game attempts to provide various subjective goals.
    Most don't do a very good job of it.  This is most often
    successful as pertains to avatar customization - one player will
    save up for pink dye, the other for blue.  In some restricted
    incidences, comparable items are offered in different varieties.
    To a certain extent, some skills can be considered subjective
    goals - one player loves crafting, the other hates it; one loves
    PvP, the other hates it.  To my mind, these are fairly paltry
    excuses for subjective goals, especially considering the above
    statements that players will rapidly figure out the "best"
    thing/path/end and begin using those exclusively.
    Non-equivalent items will always be measured on a scale,
    especially when taking into account the (im)balance problems of
    class and skill.

    Now, apart from the obvious and very time-consuming idea of
    adding more things for players to do (which is not only
    time-consuming but often futile, since players frequently reject
    those activities out of hand, even those that are otherwise
    clamoring for less boring gameplay), and trying to look at the
    problem largely from the standpoint of existing systems, I
    personally think the best solution lies in actually UNbundling a
    number of concepts within the game. Unbundling items and stats.
    Unbundling skills and XP in some instances.  Unbundling skill
    dependencies.  Unbundling cash and certain types of objectives.
    Unbundling loot and mobs to some degree.

    Unbundling in this circumstance could be as simple as vastly
    increasing equivalent but dissimilar choices (10 equal swords,
    in 10 different visual styles; 10 different endings to the same
    quest; offering stat multipliers across a broader spectrum;
    offering XP for more activities), or as complex as creating a
    system where much of the content is customizable, or a system
    where skill choices are vastly expanded and unbound to some
    extent from one another.

    In real life, one could take the example of a cross-country car
    trip. One person might choose to meander and view the scenery,
    while another would just cover the mileage as quickly as
    possible.  Now, the first person has a reward - pleasure from
    observing the beauty of nature.  The second person has the more
    obvious reward - a shorter trip.  Except that to the first
    person, the shorter trip would actually represent a punishment,
    and vice versa.  We have not yet built many equivalents to this
    concept in our games.  The person who meanders is almost
    universally punished, and the person who speeds is almost
    universally rewarded, and it is rare for one of those people to
    see the second's circumstances differently than the second
    person sees them.  Newbies become upset that they are moving
    more slowly, regardless of the lovely trees the 3D artist has
    cooked up.

  3. The Risk-Reward Equation

      (Or, that players handily compute in the time-to-destination
      risk as well as the death risk when measuring a reward for
      desirability)

    I don't think we take nearly as much advantage of this as we
    could. Developers tend to see players' ability to locate and
    compute the fastest way to a goal, up to and including risk
    levels, as a frustration, and usually put their efforts into
    making finding that route as difficult as possible.

    What if we exposed those mechanical systems to a far greater
    degree than we currently do? For one thing, I think that to do
    so would force developers to be more rigorous in examining the
    precise degrees and effects of the risk-reward systems in place.
    What I am wondering is if we exposed those systems to such an
    extent that the players themselves would be better able to adapt
    the systems to their purposes, could we not then move our focus
    away from concealing those mechanics and towards expanding them
    in ways that work better for the players?

    I mean, it seems obvious to me that a central concept to game
    design is high risk = high reward, and that, theoretically, this
    seems like a fabulous idea.  In practice, however, this system
    rapidly becomes the "absolutely minimal risk for the absolute
    maximum reward" - meaning, players quickly discover that the
    risk factor detracts more heavily from their game than the
    reward adds to it.  Not to mention the fact that due to the
    general and overall design of most games, any player can, at any
    time, return to another, lower-level mob to produce a specific
    drop, even if the experience points are not useful at that time
    (and, so doing, usually transfer that item to other characters
    of his or his friends').  Additionally, games which reinforce
    and encourage grouping (most of them), become doubly troublesome
    in the risk department, since for risk to be usefully high at
    that point it must threaten a significant number of the group.
    I'm pretty much saying that I don't think the traditional
    risk-reward implementations work very well, quite honestly - at
    least not in MMORPGs (I think that in P&P games, and even to a
    certain extent, in old-school MUDs (due to the caliber of the
    audience, for one thing) it works much better).

    I'm not 100% sure how to replace what we've got when it comes to
    high-risk/high-reward systems.  It's certainly something that I
    am and have been thinking about pretty constantly for some time,
    and I have some ideas that I'm working to explore.  There is
    really an enormous conundrum to cope with when you want to
    examine that system - the powergamer vs. casual gamer conundrum.
    Risk = Time :: High-Risk (and associated High Reward) = More
    Time.  I'm trying to think along the lines of developing a
    risk/reward system that takes into consideration travel time and
    tangential risks as well as primary risks, and from that point
    offers a bevy of possible activities in such a fashion as to
    allow the players to choose from a pretty large number of
    possible timeframes and rewards (maybe even customized rewards).
    By doing so I think we could identify the risk-reward
    combinations that really, honestly work for the players, and
    expand on them.

    I also believe that we need to examine WHY the optimal route
    must be predictable activities which must mean boring
    activities.  Mr. Koster mentions that increasing the risk
    decreases the number of people choosing that activity, and that
    eliminating the risk pretty much eliminates the point.  That's
    not exactly where I'm disagreeing.  I think that the problem is
    that the system is broken.  We're spending huge amounts of time
    developing risk/reward combinations that are endlessly and
    stubbornly ignored by the player base, all but a few
    combinations.  I could literally, off the top of my head, list
    for you the exact best places to fight in DAoC Hibernia from
    levels 1 - 50 (other games too, but not off the top of my head -
    it's the one I happen to know the best).  Does this not indicate
    that we're wasting our time, whether it's because the system is
    inherently erroneous or because the implementation is erroneous?
    Clearly, the players are finding the most appropriate routes.  I
    don't think they're doing so out of some love of repetition, or
    wish for all activities to be predictable.  I believe they are
    doing so because that is how WE have defined success.  WE
    determine the rewards - not the players.  WE determine the
    factors of measuring skill and knowledge and powergamer-hood.
    Not them.  Given that fact, and given that the players are
    subsequently totally ignoring our carefully-crafted deep and
    broad methods for obtaining the rewards defined by us, wouldn't
    you think that something is going wrong, and it ain't the
    players?

    Additionally, I wonder if it's time to much more strongly
    consider segregating (not physically, systematically) the
    powergamer from the casual gamer.  What I mean by that is -
    there are very few games in which one can feel one is truly
    accomplishing much in limited periods of time.  It's extremely
    unrewarding.  Casual gamers often play to spend time with
    friends, and in those instances can easily become frustrated
    when their friends swiftly outdistance them in experience and
    level.  I wonder if there are ways to tier out character
    development (and, quite possibly, pricing plans) in a way that
    will allow casual gamers to feel like they're having fun while
    the powergamers do too?  I think DAoC's pre-50 Battlegrounds are
    an attempt in that direction, but it's not very good at it.
    Many, if not most, of the population in those battlegrounds are
    players with other, level 50 characters.  First, I think we need
    a way to allow high-level players to group with low-level
    characters *as if they were that level again.* Not
    powerleveling, just hanging out with friends.  I also think we
    need a way to play MMORPGs as if they were something like an FPS
    - meaning, say, a state in which all players could compete,
    albeit briefly, on a level playing field.  Even past that, I
    think it would be interesting to create meaningful (truly
    meaningful) ways for players to impact the world at lower
    levels. This latter idea is something that I have yet to see
    done in a way that the players actually, honestly believe is
    meaningful to the game.

[Much Snipping]

> Eliminating cases where you can kill the monster without any risk
> is a dubious proposition. We all try, of course. The fact that we
> have the term "bottomfeeding" reflects how widespread the practice
> of optimizing advancement is. Players will always drive towards
> this point. It is very likely that there will be one optimal
> strategy per challenge. Once this strategy is codified, it meets
> the criteria you defined: repetitive activity and obvious outcome.

> I'll go further and say that it doesn't have to be a guaranteed
> return--it just needs to be the highest possible return on the
> risk in the possibility space. A 70% chance of return is still
> enough of a pattern that people will see it as predictable. Humans
> are incredibly good pattern-matching machines.

Agreed, as mentioned above.

As for eliminating risk in terms of PvE, I do agree that games have
tried, but I don't think that it's been pushed very far, frankly.
We all define an encounter very, very narrowly (usually some number
vs. some generally equivalent number resulting in the "death" of one
side's members).

Risk can be avoided in a number of ways in life.  Sure, some
risk-avoidance is simply "do not go to the place with the scary
person," and "do not attack person bigger than you unless you have a
very big gun."  But what about, for example, the concept of
preparation? Why can't I include as a portion of my risk strategy
the idea that if I put x amount of preparation into a fight, it
removes y amount of risk?

Additionally, you mention people avoiding risk in real life
(choosing boring careers - got snipped for space).  I think this is
an inaccurate comparison.  Players are avoiding risk in games not
because they're little scaredy-cats, they're doing so because of
what you state elsewhere - they have figured out the fastest way to
the top.  People in real life do not choose boring careers because
it's the fastest way to the top.  We are clearly and overtly
rewarding risk-averse behavior by enforcing that behavior as the
optimal means. Sure, we can't remove risk, and nor can we make
everything wholly unpredictable.  But, we can dramatically increase
the rewards when it's practical to do so.  We can redefine the
rewards, we can redefine predictable (especially in combat.  For all
the D&D tradition, there isn't, for example, a "saving throw."  Nor
is there much variation in tactics, that dreaded word.), we can
allow players to scale up and down their rewards and risks far, far
more than we do now.  We can even - gasp - surprise them once in a
while.

I am certainly aware that for all their screaming for excitement and
variety and better gameplay, most players lean heavily towards "what
they know."  I don't think that's bad, really.  I think it's just a
place where we honestly need to take a better look at what's going
on and why.  I think we can do a better job of helping players be
able to USE "what they know" in new or more unpredictable
situations.  I think we can increase and decrease risk far more
incrementally than we do now.  And I even think we can maintain
risk/reward levels while simultaneously drastically improving
gameplay.

> That's because most games, across cultures and across the ages,
> are about teaching players how to solve a specific problem. As
> Dave Rickey recently put it in his Skotos column, they are puzzles
> to be solved (puzzles of varying degrees of complexity). One of
> the reasons why most long-lasting games have been player vs player
> in the past is because it allows the puzzle to change constantly
> and dynamically.

Er, this email of mine isn't getting any shorter, is it?  But I keep
seeing some really excellent things to try to talk about.

I'm not sure that "puzzles" is the right term.  And I'm really not
sure that the satisfaction players gain is from solving the
equivalent of puzzles - or, if they are, they ain't the puzzles we
intend.  Where I see the greatest amount of puzzle-related
satisfaction is in fooling (or "getting") the developers.  Finding
bugs and exploits, locating the optimal route to advancement, even
just feeling like they understand the devs' motives.  THOSE are the
puzzles players like to solve.

I don't think that players of MUDs and MMORPGs are finding
puzzle-related satisfaction at all, really.  First, there's the fact
that a puzzle implies that by accomplishing or ascertaining a
specific set of tasks, sometimes in a specific order, one will
always win.  This is highly untrue of MMORPGs, and even less true of
PvP combat.  It's not a puzzle, it's COMBAT.  It's the gamer
equivalent of RL fencing or boxing.  The kids running around
screaming "I pwnz j00" are not crowing about their brilliant
solution to a puzzle. Myst is a puzzle.  MMORPGs and MUDs often
include puzzle-games within the games, but I truly don't see the
relationship to the games as a whole.

If MMORPGs were puzzles, then, in fact, the presentation of optimal
routes to the perceived end would invalidate much of the game.  The
satisfaction in solving a puzzle is *solving* the puzzle - not using
someone else's solution.

>>   When the players themselves focus solely on the details, game
>>   designers have failed.

> That is like saying that the poet is right for looking across a
> verdant landscape and seeing Nature's Beauty, but the scientist is
> wrong for seeing photosynthesis in action. You are espousing a
> worldview, an aesthetic of game design, but not an absolute.

Hm.  Well, that's probably true.  However, I stick by the statement,
mostly because I cleverly included the modifier "solely" :) - we're
trying to create worlds, universes, are we not?  Besides, to a
scientist, photosynthesis is NOT a detail - it is a meaningful
system with many details in its own right.  The scale of the
holistic system is irrelevant; the level of complexity is not.
Staring at "a+b=c" is boring and stale.  Staring at "e=mc squared"
is less so.

>>   When the players are unable to see the game as a holistic idea,
>>   one that either appeals or does not, the designers have failed.

> And this one is even an aesthetic that I AGREE with. :)

>>   When the questions and answers involve mathematical equations
>>   rather than what is honestly interesting about a game, the
>>   designers have *especially* failed.

> What of the (sizable) portion of people to whom the mathematical
> equations ARE what is interesting? There's a good case to be made
> that this is the *majority* of game players. In Bartle's typology,
> both achievers and explorers will fall into this category, and
> come to think of it, every killer I've ever talked to was
> similarly focused on the minutia because it gave them an edge.

I did not mean that the mathematical equations were invalid or
useless - I mean that I don't see why they are among the
heavily-asked questions.  An equation is an equation, dammit.  When
they are questioned, it is because the result of the equation is not
appropriate, whether it is because the equation was erroneously
solved or because the result is not totally applicable.  In the
former case, it's a bug.  In the latter case, it's a balance
question (usually).  What I meant by that statement is that I
believe that the equations themselves - while easily an intellectual
pursuit for many people - should be a given, not a question.

>>   Are most current games derived from D&D-style play? Yes, of
>>   course.  But why do we not use the technology and the talent at
>>   hand to abstract those concepts? Why do we have gamers behaving
>>   as if the game is actually rolling dice rather than immersing
>>   themselves in the world?

> Two answers:

>    1) Because they are not stupid, and they know that the game is
>    actually rolling dice

>    2) Because they prefer to see the game world as a puzzle to be
>    solved

1. Why is it actually rolling dice?  And if it must be so that it
is, why is it not more interesting (at least to a majority) to think
about causes and effects, rather than staring at the little tiny
"-14 HPs dmg +2 bonus" in the little tiny window?  Especially if all
I'm doing during the fight is twiddling my thumbs or poking F4 fifty
times?

2. I disagree :)

Okay, I cannot write any more at the moment, and I am quite
overwhelmed by the fact that about fifty new messages have appeared
in this list's mailbox while I have been writing it. I will attempt
to finish up later, if it still seems relevant to do so.

-k
_______________________________________________
MUD-Dev mailing list
MUD-Dev at kanga.nu
https://www.kanga.nu/lists/listinfo/mud-dev