Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

IF Indicators ?

3 views
Skip to first unread message

rgrassi

unread,
Feb 15, 2006, 7:17:56 AM2/15/06
to
I'm considering the following 'indicators' to help "evaluating" a game
(not in terms of good/bad, but as reference for discussions)
--------------------------------------------------------------------------
1. Number of Locations / Number of NPCs
(It is 0% if the second term is 0, It is 100% if result is >than 100%)

This indicator should tell us something about "how much" the world is
populated. NPCs are intended has objects with Inform-like "life"
property.
--------------------------------------------------------------------------
2. "Accessible" Locations / "Not accessible" Locations
(It is 0% if the second term is 0, It is 100% if result is >than 100%)

This indicator evaluates "how much" the map has to be unveiled. It must
be calculated at the end of the game. "Not accessible" Locations cannot
be considered "Accessibile" after they've been unblocked.
--------------------------------------------------------------------------
3. (Scenery+Static) Objects / Dynamic Objects
(It is 0% if the second term is 0, It is 100% if result is >than 100%)

This indicator gives some indications about "how much static" and
location based is the game.
--------------------------------------------------------------------------
4. Reusable Objects / Dynamic Objects

This indicator can be used to evaluate "how much" objects are reused
during the game. Can be some hints about a game with "get x use x leave
x" syndrome.
--------------------------------------------------------------------------
5. Puzzles / Locations

This indicator could give an overall idea about the
"puzzlefest-puzzleless" spectrum. Puzzles categories must be cited in
the report. At first, we could use
http://www.mxac.com.au/drt/PuzzlesProblemSolving.htm.
--------------------------------------------------------------------------
6. LockandKey Puzzles / Other Puzzles

Using the previous list this indicator can give an idea on "how much
lockandkey" are expected.
--------------------------------------------------------------------------
7. Moving NPCs / Static NPCs

Are NPCs alive?
--------------------------------------------------------------------------
8. Topics / NPCs

For conversation topics coverage. Galatea could be the 100%.
--------------------------------------------------------------------------

They're just notes. Will try to use it for the new pages about articles
and reviews in English I'm preparing in my homepage.
Rob

Kevin Venzke

unread,
Feb 15, 2006, 9:45:48 AM2/15/06
to

"rgrassi" <rgras...@gmail.com> wrote in message
news:1140005876.5...@o13g2000cwo.googlegroups.com...

> I'm considering the following 'indicators' to help "evaluating" a game
> (not in terms of good/bad, but as reference for discussions)

The only thing like these that I've calculated is objects (including NPCs
and scenery) divided by locations.

> --------------------------------------------------------------------------
> 1. Number of Locations / Number of NPCs
> (It is 0% if the second term is 0, It is 100% if result is >than 100%)

Is this backwards? It will usually be greater than 100%.

> 5. Puzzles / Locations
>
> This indicator could give an overall idea about the
> "puzzlefest-puzzleless" spectrum. Puzzles categories must be cited in
> the report. At first, we could use
> http://www.mxac.com.au/drt/PuzzlesProblemSolving.htm.

This is hard. To me it isn't always clear what counts as a puzzle.

> 8. Topics / NPCs
>
> For conversation topics coverage. Galatea could be the 100%.

This is another one that should be greater than 100% most of
the time.

Kevin Venzke


rgrassi

unread,
Feb 15, 2006, 10:10:43 AM2/15/06
to
Hi Kevin,

> > 1. Number of Locations / Number of NPCs
> > (It is 0% if the second term is 0, It is 100% if result is >than 100%)

> Is this backwards? It will usually be greater than 100%.

Could be. Let's start gathering data, for the moment.

> > This indicator could give an overall idea about the
> > "puzzlefest-puzzleless" spectrum. Puzzles categories must be cited in
> > the report. At first, we could use
> > http://www.mxac.com.au/drt/PuzzlesProblemSolving.htm.

> This is hard. To me it isn't always clear what counts as a puzzle.

I know. THat's why I'm suggestin at least a "common" document for
puzzles classification and a list of puzzles that must be attached with
this indicator, in order for readers to be aware of wha'ts been
considered as a puzzle.

> > 8. Topics / NPCs

> > For conversation topics coverage. Galatea could be the 100%.

> This is another one that should be greater than 100% most of
> the time.

Yes, you're right.
Maybe the right indicator could be "Topics (NPC) / Topics (ALL)"
considering which topics are covered by NPC with respect all (?) topics
in the game.

Rob

Quintin Stone

unread,
Feb 15, 2006, 12:14:52 PM2/15/06
to
On Wed, 15 Feb 2006, rgrassi wrote:

> I'm considering the following 'indicators' to help "evaluating" a game
> (not in terms of good/bad, but as reference for discussions)

These would be more appropriately referred to as statistics or metrics.

> 1. Number of Locations / Number of NPCs
> (It is 0% if the second term is 0, It is 100% if result is >than 100%)
>
> This indicator should tell us something about "how much" the world is
> populated. NPCs are intended has objects with Inform-like "life"
> property.

Every game has at least 1 room, but not every game has an NPC. These
should be reversed if you want to compare them, and should be allowed to
range over 100% if necessary, in order to give an accurate representation.
If a game has 1 location and 3 NPCs, why not say 300%?

> 2. "Accessible" Locations / "Not accessible" Locations
> (It is 0% if the second term is 0, It is 100% if result is >than 100%)
>
> This indicator evaluates "how much" the map has to be unveiled. It must
> be calculated at the end of the game. "Not accessible" Locations cannot
> be considered "Accessibile" after they've been unblocked.

Again here we have a situation where the first term will never be 0 (your
start room should surely be considered "accessible" right? Or at least
the end room would be accessible, since that's where the game ended) and
the second one might be 0. So reverse, since division by 0 isn't 0 and it
doesn't make any sense to try and represent it that way.

I'm confused by the concept of "not accessible" calculated at the end of
the game. Are these rooms that were formerly accessible and have since
been blocked off? Could you be more specific in what you mean here?

> 3. (Scenery+Static) Objects / Dynamic Objects
> (It is 0% if the second term is 0, It is 100% if result is >than 100%)
>
> This indicator gives some indications about "how much static" and
> location based is the game.

Another reversal. It might also make sense to say "dynamic objects / all
objects" for a nice 0 to 100 range.

> 4. Reusable Objects / Dynamic Objects
>
> This indicator can be used to evaluate "how much" objects are reused
> during the game. Can be some hints about a game with "get x use x leave
> x" syndrome.

Okay, but this seems an odd comparison. Why compare reusable to dynamic?
Is this really a useful or relevant statistic? Why not contract reusable
with the count of all objects?

> 7. Moving NPCs / Static NPCs
>
> Are NPCs alive?

Why not Dynamic NPCs / All NPCs, which would then give a range of 0% to
100% (unless there are no NPCs at all, which would be N/A).

==--- --=--=-- ---==
Quintin Stone "You speak of necessary evil? One of those necessities
st...@rps.net is that if innocents must suffer, the guilty must suffer
www.rps.net more." - Mackenzie Calhoun, "Once Burned" by Peter David

rgrassi

unread,
Feb 15, 2006, 12:19:27 PM2/15/06
to
All your comments are right Quentin.
I've written those few notes too quickly and most of 'metrics' must be
reversed.
I'll work on better specs for them and repost later.
Rob

rgrassi

unread,
Feb 15, 2006, 12:27:51 PM2/15/06
to
Hi Quintin,

> I'm confused by the concept of "not accessible" calculated at the end of
> the game. Are these rooms that were formerly accessible and have since
> been blocked off? Could you be more specific in what you mean here?

By "accessible" I mean locations that can be reached without troubles
only typing the usual direction commands, i.e. the 8 in compass rose
plus up and down. In this category I add "jumping, climbing and flying
it the action moves me to the desired location without troubles".
In that sense in "not accessibile" should fall other locs that require
a "puzzle" to be solved in order to proceed further in the map
exploration (for example, the classic lock-and-key).
Rob

BrettW

unread,
Feb 15, 2006, 6:39:08 PM2/15/06
to

This is a cool idea, although I worry about the specifics.

I'm figuring it's probably better to consider these as (truncated) decimal
numbers. Percentages might have a psychological effect that a higher
percentage means a better game.

Also, for example, a one-room game with a single NPC is not really the same
as a 100 room game with 100 NPCs. I think instead of compressing the data
into a single number, you should keep it as a ratio.

On the specific metrics (I don't intend to be overly critical, just
pondering
the finer details):

> 1. Number of Locations / Number of NPCs

I don't think this will be so easy to do. You can have a game with a
strong NPC
who doesn't actually exist as a different object (implied NPCs or NPCs "in
absentia")

> 2. "Accessible" Locations / "Not accessible" Locations

I didn't quite get this one, especially with calculating at the end. I
would have
thought you'd want a measure of "without solving any puzzles, how many
rooms can I
see?" Counting blocked-off locations at the end of a game is kinda weird.

> 3. (Scenery+Static) Objects / Dynamic Objects

Does this include NPCs? And by dynamic, does it count that you can move it
from
container to container, but not necessarily from room to room?

> 4. Reusable Objects / Dynamic Objects

Does this count *potential* reusability? Suppose you had a game
implemented finely
enough that you could solve all the puzzles with a single object if you
used it
creatively.

How about dynamically-created objects (like a book of matches, or cupsful
of water from
a stream)?

> 5. Puzzles / Locations

It might be hard to cleanly separate the puzzles. Take, for example,
"Enlightenment".
There's a main goal ("Get rid of all the light"), subgoals ("Get rid of x,
get rid of y"),
and for each of them, there are subgoals (since it may not require a
single step).

A good way (perhaps) is to count the instances when you get points. If the
author thinks
it's time to award points, then presumably the author thinks that a single
puzzle has been
solved.

> 8. Topics / NPCs
>
> For conversation topics coverage. Galatea could be the 100%.

This might favour games with loads of unimportant but implemented
responses.

Also, how do you count topics? A game could implement ASK GUY ABOUT STUFF
IN GENERAL and have
different responses (all about different actual topics) for the several
times you tried that
command. Alternatively, you could ASK GUY ABOUT APPLE and get several
responses all about apples.

Does it count if you ask a topic at point A in the game and get a
different answer if you did so
at point B?


All in all, with a lot of tweaking and careful consideration, a database
like this could be very
interesting. (Especially to those addicted to stats :) )

BrettW

Poster

unread,
Feb 15, 2006, 7:53:05 PM2/15/06
to
How about running a few games through the stats and seeing what they
generate? That could tell you (us) if they're useful for what you want
to do. :)

-- Poster

www.intaligo.com Building, INFORM, doom metal

rgrassi

unread,
Feb 16, 2006, 3:51:20 AM2/16/06
to
Hi,

> This is a cool idea, although I worry about the specifics.

Me too... :D
Infact I'm going to improve them.
Rob

BrettW

unread,
Feb 16, 2006, 7:32:23 AM2/16/06
to


Perhaps it's better to store and display all the base stats (number of
rooms etc)
and *then* the derived stats. Recording just the derived stats is
not-so-good.

BrettW

rgrassi

unread,
Feb 16, 2006, 8:27:49 AM2/16/06
to
Hi Brett,

> Recording just the derived stats is not-so-good.

Yes, I know.
Rob

0 new messages