[DESIGN] How big is enough?

8 messages in this thread from mud-dev2 in 2009-01

  1.   Ian Hess <ianhess@ya...com> 01-14 20:45
  2.   Mike Rozak <Mike@mx...au> 01-14 21:38
  3.   Mike Oxford <szii@sz...com> 02-07 18:02
  4.   Vincent Archer <archer@fr...org> 02-09 12:08
  5.   Zach Collins (Siege) <siegemail@gm...com> 01-15 04:17
  6.   Threshold <mlist@th...com> 01-15 06:41
  7.   David Johansson <johansson.david@gm...com> 01-15 10:09
  8.   Roger_Durañona_Vargas <luo_hei@ya...es> 01-15 13:22
I've been hacking at a hobby project for some time,between playing and
administrating various games.  Ingoing back to work on my db schema
again, I've runinto questions about how big to make the world, andthe
obvious fallout in performance and storage, as wellas population
density.  

I've been thinking about a world size in terms of howlong it could take
a determined character to walk fromone side to the other.  I was curious
if there are othermetrics that might be better used to determine a
maximumscale to plan for.

Thanks,
Ian Hess

_______________________________________________
MUD-Dev2 mailing list
MUD-Dev2@mu...com
http://my.binhost.com/lists/listinfo/mud-dev2

Mike Rozak <Mike@mx...au>

2009-01-14 21:38:26
Ian Hess wrote:

> I've been thinking about a world size in terms of howlong it could take
> a determined character to walk fromone side to the other.  I was curious
> if there are othermetrics that might be better used to determine a
> maximumscale to plan for.
You first need to answer (at least) two questions:

(a) How many players will occupy a shard at one time?

(b) How many hours of content will there be? (Or, how many hours of play 
before most players get bored and go elsewhere.)


Here's how the numbers fit together:

Player density is important. If you have too many players in an area, the 
game feels more like disneyland, waiting in line. If there are too few, it 
feels like a poorly written single-player game. When I played WoW (a few 
years ago now) it felt like a good player density. If I stopped at a quest 
site and waited, another player (doing the same quest) would come along 
every 6-10 minutes. Inverting that: you need 6-10 players per hour of 
gameplay in each shard. What that means is that if I have a game with 500 
hours of content, I need a world with 3000-5000 players.

How many hours of content is important depending upon what market you're 
going after. Game playing time (even for non-MMORPGs) has gotten shorter 
over the past few decades because players are less-and-less hardcore. Fable 
1 was about 15-20 hours. Mass Effect about 20. Fallout 3'd main quest around 
30(?). Baldur's Gate was... immense, 50-100 hours(?).

So, if you want a game that can actually be finished by a majority of 
today's players, it needs no more than 20 hours of content. x 6-10 players 
per hour => 120 - 200 players in a shard. How large does your world need to 
be so that 120-200 people fit it in comfortably? Not that large.

If you target your game at hardcore players who can commit 500 hours, then 
you need 3000-5000 people in a shard, and a much larger world.

_______________________________________________
MUD-Dev2 mailing list
MUD-Dev2@mu...com
http://my.binhost.com/lists/listinfo/mud-dev2

Mike Oxford <szii@sz...com>

2009-02-07 18:02:35
On Thu, 15 Jan 2009 07:08:26 +0930, "Mike Rozak" <Mike@mx...au>
said:
> Ian Hess wrote:
> 
> > I've been thinking about a world size in terms of howlong it could take
> > a determined character to walk fromone side to the other.  I was curious
> > if there are othermetrics that might be better used to determine a
> > maximumscale to plan for.
<snip>

> If you target your game at hardcore players who can commit 500 hours,
> then you need 3000-5000 people in a shard, and a much larger world.
Read up on the C10k problem.  Some of it's dated but the concepts are
still there.

You're going to hit technical limitations for scalability as well as
your planned content base.  Doesn't do much good to plan for 500 hours
of content (projecting 3-5k players at a time) and then be capped at
1000 concurrent users.  C10k today may be C20k in reality with today's
hardware (I haven't done any profiling in a number of years,) but it's
just a heads-up to keep your pipes/memory/bandwidth/CPUs/etc in
consideration.

-mox
_______________________________________________
MUD-Dev2 mailing list
MUD-Dev2@mu...com
http://my.binhost.com/lists/listinfo/mud-dev2

Vincent Archer <archer@fr...org>

2009-02-09 12:08:53
According to Mike Oxford:
> You're going to hit technical limitations for scalability as well as
> your planned content base.  Doesn't do much good to plan for 500 hours
> of content (projecting 3-5k players at a time) and then be capped at
> 1000 concurrent users.  C10k today may be C20k in reality with today's
> hardware (I haven't done any profiling in a number of years,) but it's
> just a heads-up to keep your pipes/memory/bandwidth/CPUs/etc in
> consideration.
Note that you do not need to serve every single simultaneous player
from a single server front-end. Unless you plan for large-scale
battles in which the entire server gets concentrated on a single area,
you are better off with segmenting your world into individual zones,
with a transparent or specific zoning (transparent zoning means you're
handed off to a new server at "an appropriate time" without any
obvious effects, while specific gives you the loading screens almost
every MMO player is familiar with).

You usually need multiple servers to handle the game world itself,
so your architecture can either have a lightweight front-end that
essentially dispatches incoming connexions to the various zone back-ends
(which, if I remember right, is how UO worked), or you have a master
control system that tells your client to switch from one zone server
to another (which is how EvE handles its enormous world cluster, I
think).

-- 
	Vincent Archer			Email:	archer@fr...org

All men are mortal.  Socrates was mortal.  Therefore, all men are Socrates.
							(Woody Allen)
_______________________________________________
MUD-Dev2 mailing list
MUD-Dev2@mu...com
http://my.binhost.com/lists/listinfo/mud-dev2

Zach Collins (Siege) <siegemail@gm...com>

2009-01-15 04:17:57
On Wed, Jan 14, 2009 at 3:45 PM, Ian Hess <ianhess@ya...com> wrote:
> I've been hacking at a hobby project for some time,between playing and
> administrating various games.  Ingoing back to work on my db schema
> again, I've runinto questions about how big to make the world, andthe
> obvious fallout in performance and storage, as wellas population
> density.
>
> I've been thinking about a world size in terms of howlong it could take
> a determined character to walk fromone side to the other.  I was curious
> if there are othermetrics that might be better used to determine a
> maximumscale to plan for.
The relative amount of attractive content comes to mind. How spread
out are items and areas that players will have interest in? Towns or
other gathering places will be attractive because shops and
guildhouses tend to go there. Newbie zones will be attractive as long
as new players are coming in. Item-gathering opportunities (ore,
herbs, vegetables, whatever) will get people moving, depending on how
that's handled (plants can be picked once and have to respawn, veins
in a mine might deplete over time, et cetera).

-- 
Zach Collins (Siege)
"If code can be speech, then software can be art."
_______________________________________________
MUD-Dev2 mailing list
MUD-Dev2@mu...com
http://my.binhost.com/lists/listinfo/mud-dev2

Threshold <mlist@th...com>

2009-01-15 06:41:21
Ian Hess wrote:
> I've been hacking at a hobby project for some time,between playing and
> administrating various games.  Ingoing back to work on my db schema
> again, I've runinto questions about how big to make the world, andthe
> obvious fallout in performance and storage, as wellas population
> density.  
>
> I've been thinking about a world size in terms of howlong it could take
> a determined character to walk fromone side to the other.  I was curious
> if there are othermetrics that might be better used to determine a
> maximumscale to plan for.
That is worth considering, but I think there are other (perhaps more 
important) issues when considering how big your world should be.

1) How many players do you think you might have? I ask this because 
player density is important. If players are too packed together, they 
are fighting over places to go and things to do. If player density is 
too sparse, they never run into each other and interact.

2) To what extnt is the focus of your game PvE, PvP, RP, etc. If it is 
all about PvE, a bigger world with more places to go, fight, etc. is 
better. If it is more about world PvP, you want the world not so huge, 
but with zones that are conducive to players fighting each other. If RP 
is the focus (yeah, stop laughing) then you want a small world so people 
are constantly crawling all over themselves.

3) How much content do you think you (or your team) can create? A huge 
world that is empty and dull is worse than a small world with the same 
amount of content but easier to reach it all. Big for the sake of big is 
not valuable.

_______________________________________________
MUD-Dev2 mailing list
MUD-Dev2@mu...com
http://my.binhost.com/lists/listinfo/mud-dev2

David Johansson <johansson.david@gm...com>

2009-01-15 10:09:54
On Wed, Jan 14, 2009 at 9:45 PM, Ian Hess <ianhess@ya...com> wrote:

> I've been hacking at a hobby project for some time,between playing and
> administrating various games.  Ingoing back to work on my db schema
> again, I've runinto questions about how big to make the world, andthe
> obvious fallout in performance and storage, as wellas population
> density.
>
> I've been thinking about a world size in terms of howlong it could take
> a determined character to walk fromone side to the other.  I was curious
> if there are othermetrics that might be better used to determine a
> maximumscale to plan for.
>
> Thanks,
> Ian Hess
In my opinion, do not strive to mark a territory and then fill it with
content, fill it with content from inside and out, then see after you have
added all your content, if the world feels large enough.

If it feels a little too small, you can stretch out your content, increase
the distances between the points of interest, generate random areas in
between or something like that, but I always found that assigning the size
of the area before placing the content ended up in catastrophies.

Traveling is boring since it "wastes" time. If traveling is eventful, the
time is not wasted, it is lived and experienced. Handcrafted content is
always more pleasant than randomly generated one. If you have the ability to
instantly travel long distances, whatever you place between the points of
interest, will not be experienced by the players, and then you don't even
have to have it there.
_______________________________________________
MUD-Dev2 mailing list
MUD-Dev2@mu...com
http://my.binhost.com/lists/listinfo/mud-dev2

Roger_Durañona_Vargas <luo_hei@ya...es>

2009-01-15 13:22:48
Ian Hess escribió:
> I've been hacking at a hobby project for some time,between playing and
> administrating various games.  Ingoing back to work on my db schema
> again, I've runinto questions about how big to make the world, andthe
> obvious fallout in performance and storage, as wellas population
> density.  
> 
> I've been thinking about a world size in terms of howlong it could take
> a determined character to walk fromone side to the other.  I was curious
> if there are othermetrics that might be better used to determine a
> maximumscale to plan for.
> 
>From a player's point of view, I can tell you that I hate large worlds.
Newbies usually have to ask how to go to a place, and often receive
vague directions, making them waste a lot of time walking, time that
could be better employed playing. Yes, I do like large worlds, but
splitted in smaller regions with an adequate mechanism to easy moving
from A to B.

-- 
Roger D. Vargas
Gentoo Linux 2008.0
http://gpnfn.blogspot.com, The news for game programming newbies

		
______________________________________________ 
LLama Gratis a cualquier PC del Mundo. 
Llamadas a fijos y móviles desde 1 céntimo por minuto. 
http://es.voice.yahoo.com

_______________________________________________
MUD-Dev2 mailing list
MUD-Dev2@mu...com
http://my.binhost.com/lists/listinfo/mud-dev2

8 messages in this thread from mud-dev2 in 2009-01