Dear All
As we all know IF reviews are scatered all over the web, and this newsgroups
seems to be the only common place to find them (apart from 1 or 2 major
reviewers own webpages) and as such I decided today to create a special
place to have them all together.
This place will be called www.ifreviews.org (i already have the domain), and
it will be based on PHP/MySQL, to enable speed searchs/access, but specially
online reviews editing :)
For what i could understand, there are already some, more or less, review
based websites, and also that reviewers might want to keep their own reviews
webpage anyway, so I decided to enable the reviewers (in each review at
ifreviews.org) to be able to show a link their own webpage's review. So that
there's a link exchange between ifreviews.org and the reviewers own webpage.
Im sure that even with the duplication of reviews, this is the best solution
for reviewers.
Later on (only after next tuesday im affraid) i will give you all more
precise details, in the meantime, should you have any ideas about this
subject, please feel free to contact me via email root...@netcabo.pt.
There are some things that were already thought about (by friends of mine),
but nothing was decided yet, things like:
- Fields you think are mandatory to be included in a review
- Search facilities that users might want
- Kind of ranking (0 to x, should x be 10 or 100, integer or not)
- Review size limit (words/kb)
Things that are already covered:
- Webmaster decision/power to cancel any reviewer's account (due to
unproper use) and even ban from site.
- Reviewer login/accounts
- Online creation/editing/delete of reviews
- Reviewers ranking (votes from other reviewers, so that reviewers take
special care on how they review the games, making it more important to have
quality rather than quantity)
- Games
ranking/category/system/language/author/Authoring_System/Links_to_game_files
(external)
- Author rankings/information/website
- Search by: game/author/word/category/rank/date/event
To prevent anyone from feeling that his/hers work was used/copied/stealed
without his/hers concent, a "terms of use" document will be implemented,
which *must* be agreeded prior to concluding the registration process by
means of a email confirmation. This email will serve as prove of
authorization to publish online any review that the reviewer decides to
publish.
I know that this should have been ready sooner so that this years IfComp.org
reviews could be entered, but i really only decided today to give it a go,
and let me tell you that it's a joy to be able to finally do something for
this world of IF that i so much love :)
Please forgive me if I sound to confusing, but it's late now (02:45), and i
just wanted to let the IF Comunity know about this new project, which *has*
to be ready before the end of the Spring Comp (according to my plans) :)
Im counting on you to go ahead with this project, where the newcomers can
find useful information about what games to try first :).
Any sugestion is welcome, since this site will be made for you but
particulary by you... The Intereactive Fiction Comunity!
Kind Regards,
RootShell
SOunds like a great idea. I remember spending ages last year scouring
the internet for reviews of my game and something like this would have
saved ages.
I will have to build it from scratch, but i will do it...
Lots of ideas are already arriving by email, so I will try to listen to
every possible contribution/feature and then layout a format here to see
what people can improve on.
Regards,
RootShell
Excellent Idea...
>
> This place will be called www.ifreviews.org (i already have the domain), and
> it will be based on PHP/MySQL, to enable speed searchs/access, but specially
> online reviews editing :)
>
> For what i could understand, there are already some, more or less, review
> based websites, and also that reviewers might want to keep their own reviews
> webpage anyway, so I decided to enable the reviewers (in each review at
> ifreviews.org) to be able to show a link their own webpage's review. So that
> there's a link exchange between ifreviews.org and the reviewers own webpage.
> Im sure that even with the duplication of reviews, this is the best solution
> for reviewers.
>
> Later on (only after next tuesday im affraid) i will give you all more
> precise details, in the meantime, should you have any ideas about this
> subject, please feel free to contact me via email root...@netcabo.pt.
>
> There are some things that were already thought about (by friends of mine),
> but nothing was decided yet, things like:
>
> - Fields you think are mandatory to be included in a review
> - Search facilities that users might want
Maybe accompanied by an alphanumeric index of the games?
> - Kind of ranking (0 to x, should x be 10 or 100, integer or not)
IMHO I recommend x = 10. 100 seems to high; just allow people to use
decimals like 9.5 or 6.25.
> - Review size limit (words/kb)
If bandwidth/size for site becomes a problem, sure. If it isn't I
recommend separating them into categories by size so visitors can
decide how indepth they want (spoilers, no spoilers, spoilers-detailed
etc...)
>
> Things that are already covered:
>
> - Webmaster decision/power to cancel any reviewer's account (due to
> unproper use) and even ban from site.
That should be a rule on every site of that type as the webmaster is
responsible for the site's rep.
> - Reviewer login/accounts
> - Online creation/editing/delete of reviews
Only by webmaster,admin(s), and authors.
> - Reviewers ranking (votes from other reviewers, so that reviewers take
> special care on how they review the games, making it more important to have
> quality rather than quantity)
Good idea... I like it
> - Games
> ranking/category/system/language/author/Authoring_System/Links_to_game_files
> (external)
> - Author rankings/information/website
> - Search by: game/author/word/category/rank/date/event
>
> To prevent anyone from feeling that his/hers work was used/copied/stealed
> without his/hers concent, a "terms of use" document will be implemented,
> which *must* be agreeded prior to concluding the registration process by
> means of a email confirmation. This email will serve as prove of
> authorization to publish online any review that the reviewer decides to
> publish.
>
> I know that this should have been ready sooner so that this years IfComp.org
> reviews could be entered, but i really only decided today to give it a go,
> and let me tell you that it's a joy to be able to finally do something for
> this world of IF that i so much love :)
>
> Please forgive me if I sound to confusing, but it's late now (02:45), and i
> just wanted to let the IF Comunity know about this new project, which *has*
> to be ready before the end of the Spring Comp (according to my plans) :)
>
> Im counting on you to go ahead with this project, where the newcomers can
> find useful information about what games to try first :).
>
> Any sugestion is welcome, since this site will be made for you but
> particulary by you... The Intereactive Fiction Comunity!
>
> Kind Regards,
> RootShell
Wonderful idea root. HOpe it goes well
8
I just hope that people will grab this idea and contribute, because without
them... the site will be empty :(
RootShell
Every succesful ranking system I've ever seen is very simple.
IMDB.com uses a 1-10 scale. The IF ratings site at carouselchain uses
a 1-10 to much success. Movie critics use a 4 star system which
essentially has eight or nine different rankings.
The more possible rankings you allow, the more muddled the rankings
become. After all, how can a person really distinguish the difference
between 68/100 and 66/100? Or if you use a 1-10 scale with decimals,
you have the same problem. What's the difference between 7.2 and 7.1?
SPAG used to have a rating system similar, with the scale being from
0.0 to 2.0, and the ratings seemed inconsistent and hard to evaluate.
I would also avoid rating different parts of the game like story,
writing, plot, etc. as it allows the rater to think way too hard about
their ratings, and they will often underrate the game because of it.
Games that purposely don't have a plot, but are still excellent
nevertheless, would be hurt by such a rating system. All that really
matters is if a person liked the game overall, unless you're handing
out awards for specific aspects like the Oscars (or, in our case, the
Erins) does.
Softiron
[snip]
>For what i could understand, there are already some, more or less, review
>based websites, and also that reviewers might want to keep their own reviews
>webpage anyway, so I decided to enable the reviewers (in each review at
>ifreviews.org) to be able to show a link their own webpage's review. So that
>there's a link exchange between ifreviews.org and the reviewers own webpage.
>Im sure that even with the duplication of reviews, this is the best solution
>for reviewers.
Duplication of reviews is a feature, not a bug. We could do with
redundancy. If one site goes down, whether temporarily or
permanently, the review is still available.
[snip]
>There are some things that were already thought about (by friends of mine),
>but nothing was decided yet, things like:
>
> - Fields you think are mandatory to be included in a review
Not neceesarily mandatory, but these should be sought:
Item title, subtitle (details on series), author (as released),
author (real name), reviewer, review text, natural language, game
language, where to download, ***where to download game system
run-time***, year of creation, part of (for contest entries),
commercial or not.
Aside from the reviews:
Contact information for the authors and reviewers. These could
be used to validate some field entries. The contact information could
simply be "DND" (Do not disturb.) if the person does not want
correspondence.
Lists of contests (used to validate "part of" anyway).
> - Search facilities that users might want
By any field and by keyword. Keyword should be by field or by
all fields. Sort by any combination of short fields. (By "short
fields", I mean anything but the review text itself.)
> - Kind of ranking (0 to x, should x be 10 or 100, integer or not)
IFComp uses 1 to 10, but why limit it? I tend to think of letter
grades as better.
> - Review size limit (words/kb)
No. This is the sort of thing that will sort itself out
automaticaly.
>Things that are already covered:
>
> - Webmaster decision/power to cancel any reviewer's account (due to
>unproper use) and even ban from site.
A must-have.
> - Reviewer login/accounts
> - Online creation/editing/delete of reviews
Yes to both.
> - Reviewers ranking (votes from other reviewers, so that reviewers take
>special care on how they review the games, making it more important to have
>quality rather than quantity)
Not really needed.
> - Games
>ranking/category/system/language/author/Authoring_System/Links_to_game_files
>(external)
> - Author rankings/information/website
> - Search by: game/author/word/category/rank/date/event
Yes.
>To prevent anyone from feeling that his/hers work was used/copied/stealed
>without his/hers concent, a "terms of use" document will be implemented,
>which *must* be agreeded prior to concluding the registration process by
>means of a email confirmation. This email will serve as prove of
>authorization to publish online any review that the reviewer decides to
>publish.
Fine.
[snip]
Sincerely,
Gene Wirchenko
Computerese Irregular Verb Conjugation:
I have preferences.
You have biases.
He/She has prejudices.
Have a check-box to obfuscate email addresses in the contact info. For
example, Slashdot allows three choices:
Email Display
[ ] Do not display an e-mail address.
[X] Show your email address with random SPAM-armoring applied.
[ ] Show your real email address without obfuscation.
That middle choice (which I use) allows people to permit correspondence
without getting harvested by spam-bots.
> IFComp uses 1 to 10, but why limit it? I tend to think of letter
> grades as better.
Do you mean like in American schools? I think it is hard to understand how
good a grade is with that system. I suspect many non-americans think the
same.
Numeric scores have a big advantage: it's very easy to average them,
add them, comput estandard deviations, and other statistic
operations. With letter grades you have to have some translation
scheme to numbers before you can do that - and then what's the point
with the letters?
--
Magnus Olsson (m...@df.lth.se)
PGP Public Key available at http://www.df.lth.se/~mol
What's so bad about that?
I certainly applaud the effort to get more feedback to authors. But is
it worth asking why there's this big push lately to get reviews put on
websites instead of posted to the group? Won't there be more feedback
generated if reviews are posted *here,* since they can then be used as
a jumping-off point for discussion?
--
Jason Melancon
>I certainly applaud the effort to get more feedback to authors. But is
>it worth asking why there's this big push lately to get reviews put on
>websites instead of posted to the group? Won't there be more feedback
>generated if reviews are posted *here,* since they can then be used as
>a jumping-off point for discussion?
You mean the entire review posted here? That could mean some high
phone bills for folks still on slow dial-up modems!
But all reviews could have pointers posted here so that people who
want to read them are aware of where they are.
Frank
___
Frank Lane
la...@eaglewing.org.uk
http://www.eaglewing.org.uk
It's often a pain searching through the groups for reviews of a game you
might want to play (or for reviews written about one of your own games). If
they were all together in one place, it would be a lot easier.
> I will have to build it from scratch, but i will do it...
Are you sure that's true? I bet there is an existing package that would
meet immediate needs.
In any event, I suggest you get the bare minimum working, see if they come
once you build it and add features based on what your visitors think is
important.
-Alan
Traditionally, A=4, B=3, C=2, D=1, F=0.
I think the letters have stuck because it's easier to remember "A is
best" than to remember "4 is best". But any reasonably bright
high-school student knows what a "4.0 average" means.
>In article <41ac5986$1...@news.broadpark.no>,
>Jan Thorsby <no_jthor...@broadpark.no> wrote:
>>
>>"Gene Wirchenko" <ge...@mail.ocis.net> skrev i melding
>>news:q77nq0h5uvnghpat2...@4ax.com...
>>> "RootShell" <root...@netcabo.pt> wrote:
>>
>>> IFComp uses 1 to 10, but why limit it? I tend to think of letter
>>> grades as better.
>>
>>Do you mean like in American schools? I think it is hard to understand how
>>good a grade is with that system. I suspect many non-americans think the
>>same.
Do you really need antything more than
A = Excellent
B = Very Good
C = Adequate
D = Marginal
F = Fail
YMWV anyway.
>Numeric scores have a big advantage: it's very easy to average them,
>add them, comput estandard deviations, and other statistic
>operations. With letter grades you have to have some translation
...and otherwise pretend that one is performing valid operations
on data.
>scheme to numbers before you can do that - and then what's the point
>with the letters?
But the statistics are rather meaningless. Is a seven for one
game the same as a seven for another?
Well, you still have to know that the scale comprises those letters
- unlike, for example, the pre-1960's Swedish system which used the
grades A, a, Ab, BA, B and C or something like that (I'm too young
ever to have experienced it).
>>Numeric scores have a big advantage: it's very easy to average them,
>>add them, comput estandard deviations, and other statistic
>>operations. With letter grades you have to have some translation
>
> ...and otherwise pretend that one is performing valid operations
>on data.
Yes. :-) Whether they're valid or not, people will be interested in
things like average scores, standard deviations and such. If nothing
else, they can pretend that these numbers mean anything.
(And I submit that they do, but in a rather fuzzy way, given that
enough people grade enough games. If one gazillion people give my
game a low grade, it probably means something.)
>>scheme to numbers before you can do that - and then what's the point
>>with the letters?
>
> But the statistics are rather meaningless. Is a seven for one
>game the same as a seven for another?
Aye, that's the rub.
But people will do statistics, if nothing else then just for the
sheer joy of data manipulation. Using letter grades won't stop them,
it will just add a meaningless level of complication.
> Magnus Olsson wrote:
> > Numeric scores have a big advantage: it's very easy to average them,
> > add them, comput estandard deviations, and other statistic
> > operations. With letter grades you have to have some translation
> > scheme to numbers before you can do that - and then what's the point
> > with the letters?
>
> Traditionally, A=4, B=3, C=2, D=1, F=0.
In one country out of nearly two hundred... :-)
--
Esa Peuha
student of mathematics at the University of Helsinki
http://www.helsinki.fi/~peuha/
I think the idea is great, and I think it will be worth the effort
you put into it. However, have you considered trying to make it a
part of the IF Ratings site? These are the advantages that I can see:
* IFRS already has a connection to Baf's guide built, ensuring that
new games in the archive make it into IFRS too.
* IFRS already has the necessary handling of database, accounts, scores,
statistics, searching etc.
* A large number of the people who write reviews already have an
account at IFRS. Less fuss for them.
* Baf's guide is rather tightly integrated with if-archive. IFRS is
rather tightly integrated with Baf's guide. Both of these integrations
are Good Things, but they would be even more powerful if the integration
was complete - if-archive linking to Baf's for each file where it's
applicable, Baf's linking to IF Ratings for each game etc. I think the
very best idea would be to have all three merged into one single,
seamless site. Assuming that won't happen for a while, adding a review
site that doesn't integrate with any of these is still a pretty good idea.
Making it integrate with one or more of them is even better. Making it
part of one of them is the best idea - that would make the reviews site
more useful, which would make people use it even more.
Of course, this requires that the developer of the IF Ratings site to
thinks this is a great idea. Also, this approach to the project can
only happen if you want to develop this in the way that's the most
useful to the community. If your motivation for this project is to
exercise your PHP skills and work on a pet project of your own,
without having to cooperate with someone else, this approach
just won't be possible.
/Fredrik
> In article <5gmqq098ff6vi9afa...@4ax.com>,
> Gene Wirchenko <ge...@mail.ocis.net> wrote:
>>m...@df.lth.se (Magnus Olsson) wrote:
>>
>> Do you really need antything more than
>> A = Excellent
>> B = Very Good
>> C = Adequate
>> D = Marginal
>> F = Fail
>>
>> YMWV anyway.
>
> Well, you still have to know that the scale comprises those
> letters - unlike, for example, the pre-1960's Swedish system which
> used the grades A, a, Ab, BA, B and C or something like that (I'm
> too young ever to have experienced it).
For those as curious as me, here's the complete old Swedish scale
(I'm also too young, but I consulted my mother):
A = Excellent, Laudatur = 3
a = Passed with great distinction, Cum insigniore laude approbatur=2½
AB= Passed with distinction, Cum laude approbatur = 2
Ba= Passed with credit, Non sine laude approbatur = 1½
B = Passed, Approbatur = 1
BC= Not fully passed (no latin phrase) = ½
C = Failed, Non admittur = 0
In addition, there was a second scale, used for grading behaviour:
A = Very good
B = Good
C = Less good
D = Bad
And a third, for order:
A = Very good
B = Good
C = Less good
Rikard
Oh so there is no E? That explains a lot!
> "John W. Kennedy" <jwk...@attglobal.net> skrev i melding
> news:I98rd.3470$197....@fe12.lga...
> > Traditionally, A=4, B=3, C=2, D=1, F=0.
>
> Oh so there is no E? That explains a lot!
That's right. Otherwise it would be too easy to change F to E.
Hope noone gets upset at this but the international/universal interger
numeric rating was adopted (based on several email/opinions), meaning
reviewers will be using a 0 to 10 integer scale (no decimals), with 0 being
the worst and 10 the best.
This appears to be the best practice, and enables several reviewers
worldwide to just paste their existing reviews at ifreviews, without the
need to change their rate.
Regards,
RootShell
> Hope noone gets upset at this but the international/universal interger
> numeric rating was adopted (based on several email/opinions), meaning
> reviewers will be using a 0 to 10 integer scale (no decimals), with 0
> being the worst and 10 the best.
>
> This appears to be the best practice, and enables several reviewers
> worldwide to just paste their existing reviews at ifreviews, without the
> need to change their rate.
Except that most reviews I've seen tend to follow the IF Comp's practice,
which is a scale that starts at 1, not 0. A minor difference.
/====================================================================\
|| Quintin Stone O- > "You speak of necessary evil? One ||
|| Code Monkey < of those necessities is that if ||
|| Rebel Programmers Society > innocents must suffer, the guilty must ||
|| st...@rps.net < suffer more." -- Mackenzie Calhoun ||
|| http://www.rps.net/QS/ > "Once Burned" by Peter David ||
\====================================================================/
Depends on the school. The 4-point system described here is really a
college thing, often, but not invariably, and less in recent years
than in the past, borrowed by high schools. The relative point assessment
we used in my school district was:
A=90-100, B=80-89, C=70-79, D=60-69, E=0-59.
The F designation has traditionally replaced the E, but in many public
schools, it was deemed excessively unkind to explicitly label such
grades a 'Failure'
[snip]
> Except that most reviews I've seen tend to follow the IF Comp's practice,
> which is a scale that starts at 1, not 0. A minor difference.
[snip signature]
Yes... your correct... my mistake... let me refrase it:
Hope noone gets upset at this but the international/universal interger
numeric rating was adopted (based on several email/opinions), meaning
reviewers will be using a *1* to 10 integer scale (no decimals), with *1*
being the worst and 10 the best.
RootShell
IFReviews.org, will be 'populated' with every game that exists at ifarchive,
which will be done using a script that will process the entire ifarchive.org
directory structure (using the file "ls-lR"), gathering games names and any
information that is possible (filename, system, the rest of the information
will be added manually).
Any other game can be added by the reviewer itself, as long as some
obligatory fields are filled in (author, platform, system, date, etc).
IFReviews and IFRS (and Baf's), are diferent sites and use also diferent
databases, the only way to unite them would be them sharing their users
table (which i think is a bad idea), reviewers at IFReviews.org can have the
use the same login (if it's available) that they currently have at IFRS.
The purpose of this site (IFReviews.org) is *not* to copy/rip anyone's
hardwork nor to deviate users from their sites.
The *main* objective of IFReviews.org is to unite not only reviews, but
mainly reviewers, in a single place.
To east the searching for reviews on a particular
game/autor/comp/reviewer/year/system is also one of the main objectives.
And as stated before, reviewers will be able to link back to their reviewes
(at their own websites), should they choose to do so, so that it's a two way
winning situation.
I haven't contacted no other IF reviews webmaster yet, since im trying to
create something more usable first (but progress is going slowly) and then
i'll start contacting them, with something for they to look at, in an
attempt to obtain some sort of linking/sharing of information.
One of the things that made me create this IfReviews.org website, is the
fact that i love IF, and also that i have a two-year 1GB hosting available,
and im still only using something like 300MB, which leaves a lot of free:
server space, mysql databases and sub-domains to use for other projects.
Your sugestion that both IFRS and IFReviews/Baf's should cooperate is in
fact possible, and would be the next logical step. Who knows?
Let's see if IFReviews.org lives up to the expectations and get's up there
with the already long time implemented reviews websites.
As soon as i have something that's in beta stage, i'll share it with the
community so that the online testing can begin, and if someone (either of
them, or someone else) like's it, we can start cooperation/sharing other
informations (like games databases/ratings/reviews).
One thing im not looking after is to copy every possible review that i can
find and paste it at IFReviews.org, since i expect that this should be left
up to the reviewers to do, so that they can do it their way, but also being
held responsible for any review 'theft' that might come up online.
I will, eventually, place my own reviews at IFReviews.org (and will not be
considered up for IFReviewer of the year award).
Of course, that any of this (above mentioned) can/will modified/changed
according to any needs.
Well i have writen way too much, now i'll go and keep strugling with PHP to
get something workable fast.
Regards,
RootShell
Is there any reason why a simple "10 = best, 1 = worst" system can't be
used?
>For a time in the 1950s, my elementary school used:
>
>E xcellent
>S atisfactory
>N eeds improvement
>U nsatisfactory
For IF-relevancy, surely the last grade should be:
W oefully bad
Regards,
Graham Holden (g-holden AT dircon DOT co DOT uk)
--
There are 10 types of people in the world;
those that understand binary and those that don't.
E xcellent
S atisfactory
N eeds improvement
U nsatisfactory
--
John W. Kennedy
"Information is light. Information, in itself, about anything, is light."
-- Tom Stoppard. "Night and Day"
You can leave U in there as long as you find room for D.
e.g. Deadly good.
Andrew
That would be easy to remember by the acronym, ESNUW.
No, I meant REPLACE the last grade. Hence ESNW (or N, S, E, W).
Regards,
Heh. The cardinals.
For AIF, we could rewrite them:
E rotic
P ornographic
C heesy
T urn-off
Dreadful would be better.
> Andrew
>
>
--
------------------------
Mark Jeffrey Tilford
til...@ugcs.caltech.edu
If you read carefully... that's exactly what's going to be implemented at
IFReviews.org :)
Regards,
RootShell
>In article <5gmqq098ff6vi9afa...@4ax.com>,
>Gene Wirchenko <ge...@mail.ocis.net> wrote:
>>m...@df.lth.se (Magnus Olsson) wrote:
[snip]
>Well, you still have to know that the scale comprises those letters
>- unlike, for example, the pre-1960's Swedish system which used the
>grades A, a, Ab, BA, B and C or something like that (I'm too young
>ever to have experienced it).
You also have to know the numeric scale. If I were scoring a
game numerically, I would prefer to score out of ten. The IFComp
system is not out of ten. It is out of nine, adjusted up by one.
[snip]
>(And I submit that they do, but in a rather fuzzy way, given that
>enough people grade enough games. If one gazillion people give my
>game a low grade, it probably means something.)
It might merely mean that you have written a game for a
specialised interest group.
[snip]
>But people will do statistics, if nothing else then just for the
>sheer joy of data manipulation. Using letter grades won't stop them,
>it will just add a meaningless level of complication.
Chalks and cheeses.
>"John W. Kennedy" <jwk...@attglobal.net> writes:
>
>> Magnus Olsson wrote:
>> > Numeric scores have a big advantage: it's very easy to average them,
>> > add them, comput estandard deviations, and other statistic
>> > operations. With letter grades you have to have some translation
>> > scheme to numbers before you can do that - and then what's the point
>> > with the letters?
>>
>> Traditionally, A=4, B=3, C=2, D=1, F=0.
>
>In one country out of nearly two hundred... :-)
1) Canada is not the only country using that system.
2) Which countries of the over 200 are you leaving out, and why?
A continuous score between 1 and 10 inclusive would be "out of nine". A
discrete score between 1 and 10 inclusive is "out of ten", if you treat each
discrete score as covering a unit interval of continuous scores.
-- Ben
>Darn it! This is going to keep me up all night. One, two, three, four...
try 0=1
1=2 etc.
Frank
___
Frank Lane
la...@eaglewing.org.uk
http://www.eaglewing.org.uk
>In article <3ra9r0l3tr69rec1j...@4ax.com>,
> Gene Wirchenko <ge...@mail.ocis.net> wrote:
>
>> You also have to know the numeric scale. If I were scoring a
>> game numerically, I would prefer to score out of ten. The IFComp
>> system is not out of ten. It is out of nine, adjusted up by one.
>
>Really? I never noticed that. Let me count...
[snipped (Most of us know it.)]
>Darn it! This is going to keep me up all night. One, two, three, four...
Give someone no points at all, that is, 0%. What do you give him
on a 1 to 10 scale? 1. Give him 11.1%, and he gets 2. And on and
on...
> On Mon, 06 Dec 2004 21:54:08 GMT, Anson Turner
> <platyp...@yahoo.com> wrote:
>
>
>>Darn it! This is going to keep me up all night. One, two, three, four...
>
>
> try 0=1
> 1=2 etc.
Well yeah, but that'd violate the Nontrivially Assumption of Fields.
--
Michael
Yes, of course. And there will always be issues (as witness the
recurring discussions every Comp) like are you scoring on a curve
(i.e. should only the top X percent get a 10), does a 10 mean a
Platonic ideal of unattainable perfection, just a very good game, or
simply the best game you've seen so far?
>If I were scoring a
>game numerically, I would prefer to score out of ten. The IFComp
>system is not out of ten. It is out of nine, adjusted up by one.
Well, I see you've already drawn some flak for this, but I think I
see what you mean: if you're scoring on a continuous scale from 0
to 10, then you'd like to round to the nearest integer, but a system
with scores 1-10 would mean rounding down and adding 1, right?
By the way, why do you consider a system with grades A,B,C,D and F to
be superior to numeric scores?
You raise some interesting issues about statistics, which I'll
address in another post.
--
Magnus Olsson (m...@df.lth.se)
PGP Public Key available at http://www.df.lth.se/~mol
Continuing what I assume is your pattern:
0% - 11.1% => 1
11.1% - 22.2% => 2
...
88.8% - 99.9% => 9
99.9% - 100% => 10
This doesn't make much sense: hardly any of the percentage range maps to a
score of 10. A more sensible mapping is
0% - 10% => 1
10% - 20% => 2
...
80% - 90% => 9
90% - 100% => 10
which is what I was trying to say in my last post to this silly thread...
-- Ben
> Gene Wirchenko <ge...@mail.ocis.net> wrote:
>
> >If I were scoring a
> >game numerically, I would prefer to score out of ten. The IFComp
> >system is not out of ten. It is out of nine, adjusted up by one.
>
> Well, I see you've already drawn some flak for this, but I think I
> see what you mean: if you're scoring on a continuous scale from 0
> to 10, then you'd like to round to the nearest integer, but a system
> with scores 1-10 would mean rounding down and adding 1, right?
But if you'ld round to the nearest integer, 0 and 10 would only get half
as much of the spectrum as the rest. You only have enough of the number
line for ten integers; demanding scores from 0 to 10 is a typical case
of a fencepost error.
> By the way, why do you consider a system with grades A,B,C,D and F to
> be superior to numeric scores?
For The Chiiillldruuuunnnn? That was, in any case, the excuse that was
used in .nl for changing the numerical test scores to verbal
descriptions, in some schools. Their poor, impressionable brains would
become depressed from being scored - maybe the same is true for poor,
impressionable Implementors. (Needless to say, all such arguments
completely misunderstand the juvenile brain. And I don't know, in fact
don't even think, that it really is Gene's reason.)
Richard
>Gene Wirchenko wrote:
>> Give someone no points at all, that is, 0%. What do you give him
>> on a 1 to 10 scale? 1. Give him 11.1%, and he gets 2. And on and
>> on...
>
>Continuing what I assume is your pattern:
>
> 0% - 11.1% => 1
> 11.1% - 22.2% => 2
> ...
> 88.8% - 99.9% => 9
> 99.9% - 100% => 10
Actually, that interpretation isn't correct. The actual mapping is:
0% - 11.0% => 1
11.1% - 22.1% => 2
...
88.8% - 99.8% => 9
99.9% - 100% => 10
>
>This doesn't make much sense: hardly any of the percentage range maps to a
>score of 10.
That is correct, but the mapping is still suitable for those who want to
reserve 10 to be either the perfect game, or the best game submitted in the
comptetition.
>In article <3ra9r0l3tr69rec1j...@4ax.com>,
>Gene Wirchenko <ge...@mail.ocis.net> wrote:
>>m...@df.lth.se (Magnus Olsson) wrote:
>>
>>>In article <5gmqq098ff6vi9afa...@4ax.com>,
>>>Gene Wirchenko <ge...@mail.ocis.net> wrote:
>>>>m...@df.lth.se (Magnus Olsson) wrote:
>>
>>[snip]
>>
>>>Well, you still have to know that the scale comprises those letters
>>>- unlike, for example, the pre-1960's Swedish system which used the
>>>grades A, a, Ab, BA, B and C or something like that (I'm too young
>>>ever to have experienced it).
>>
>> You also have to know the numeric scale.
>
>Yes, of course. And there will always be issues (as witness the
>recurring discussions every Comp) like are you scoring on a curve
>(i.e. should only the top X percent get a 10), does a 10 mean a
>Platonic ideal of unattainable perfection, just a very good game, or
>simply the best game you've seen so far?
Quite.
>>If I were scoring a
>>game numerically, I would prefer to score out of ten. The IFComp
>>system is not out of ten. It is out of nine, adjusted up by one.
>
>Well, I see you've already drawn some flak for this, but I think I
>see what you mean: if you're scoring on a continuous scale from 0
>to 10, then you'd like to round to the nearest integer, but a system
>with scores 1-10 would mean rounding down and adding 1, right?
That is about it. If the IFComp scale were 11 to 20, would
anyone seriously consider that he was giving a lowest-ranked game 11
points by giving it an 11? 11 means nothing, zero. And there are
nine points that can be earned. Subtract 10 and do the obvious for
the IFComp scoring as it is.
>By the way, why do you consider a system with grades A,B,C,D and F to
>be superior to numeric scores?
Because it is good enough. A game is excellent, very good,
average/acceptable, marginal, or bad. It is just an opinion. How
digits of precision do you hold an opinion to?
If an opinion gets justified, it is a review. That gives much
more data than a number does.
A, B, C, D, and F scuttle the debate on what an 8 is. A B is a
very good game. No different scaling involved.
>You raise some interesting issues about statistics, which I'll
>address in another post.
Computerese Irregular Verb Conjugation:
>In article <lq5ar05ckulli5f49...@4ax.com>,
> Gene Wirchenko <ge...@mail.ocis.net> wrote:
>
>> The IFComp system is not out of ten. It is out of nine, adjusted up
>> by one.
>>
>> Give someone no points at all, that is, 0%. What do you give him
>> on a 1 to 10 scale? 1. Give him 11.1%, and he gets 2. And on and
>> on...
>
>Zeno of Elea you are not. None of your fancy talk is going to make me
He is dead, no? That is, if he ever made it to the grave.
>forget how to count.
>
> 1 - 1
> 2 - 2
> 3 - 3
> 4 - 4
> 5 - 5
> 6 - 6
> 7 - 7
> 8 - 8
> 9 - 9
> 10 - 10
>
>Well, what do you know? Still 10!
Look up "fencepost error". It is a computing term.
>But why am I wasting my time? You're obviously not quite right in the
>head -- you double-space between sentences.
^ ^
If you are going to be picky about spaces, remove the beam from
your eye. The noted spaces should not be there. Now, I think that it
looks better with them when using a double-hyphen to represent a long
dash, but there you go.
>Anson, who now has a pretty good idea how the "baker's dozen" got
>started.
Ah, no. Think of the complication with pans. It is rather easy
to keep 12 and 13 distinct in that context.
Is 90% a 9 or a 10?
I prefer
0 0%
1 10%
2 20%
3 30%
etc.
> Esa A E Peuha <esa....@helsinki.fi> wrote:
>
> >"John W. Kennedy" <jwk...@attglobal.net> writes:
> >
> >> Traditionally, A=4, B=3, C=2, D=1, F=0.
> >
> >In one country out of nearly two hundred... :-)
>
> 1) Canada is not the only country using that system.
I didn't know it was used in Canada (in addition to USA). So that's
two countries.
> 2) Which countries of the over 200 are you leaving out, and why?
Over 200? There are 191 members of the United Nations, then there's
Vatican and Taiwan, so the total is 193. How can you get over 200?
--
Esa Peuha
student of mathematics at the University of Helsinki
http://www.helsinki.fi/~peuha/
> A, B, C, D, and F scuttle the debate on what an 8 is. A B is a
> very good game. No different scaling involved.
Not true. You still have to draw the line between very good and
adequate (and excellent on the other end), so there is still some
fuzzyness where the person rating gets to use his (or her) judgement.
This is unavoidable unless you create a very specific list of goals a
game must reach in order to get a B.
I agree that labels to describe what the points are supposed to mean
and fewer points makes the lines clearer, but I don't see that
difference between the two below:
A = Excellent
B = Very Good
C = Adequate
D = Marginal
F = Fail
5 = Excellent
4 = Very Good
3 = Adequate
2 = Marginal
1 = Fail
It's not the B that makes the rating more exact, it's the "Very
Good" definition that comes after it, and the smaller resolution.
(What's the difference between an 88% and an 87% score?)
Rikard
>Gene Wirchenko <ge...@mail.ocis.net> writes:
>
>> Esa A E Peuha <esa....@helsinki.fi> wrote:
>>
>> >"John W. Kennedy" <jwk...@attglobal.net> writes:
>> >
>> >> Traditionally, A=4, B=3, C=2, D=1, F=0.
>> >
>> >In one country out of nearly two hundred... :-)
>>
>> 1) Canada is not the only country using that system.
>
>I didn't know it was used in Canada (in addition to USA). So that's
>two countries.
>
>> 2) Which countries of the over 200 are you leaving out, and why?
>
>Over 200? There are 191 members of the United Nations, then there's
>Vatican and Taiwan, so the total is 193. How can you get over 200?
wikipedia does have 201 candidates. They include what you stated
plus
"* 6 de-facto independent countries (Abkhazia, Nagorno-Karabakh,
Northern Cyprus, Somaliland, South Ossetia, Transnistria)
* 2 entities recognized by many countries but de-facto not
independent, Palestine and Western Sahara."
I thought there were more. One site claimed over 260, but I did
not get into the details.
> This place will be called www.ifreviews.org (i already have the domain), and
> it will be based on PHP/MySQL, to enable speed searchs/access, but specially
> online reviews editing :)
I would suggest mandatory session transcript for the review.
Rob
--
Posted via Mailgate.ORG Server - http://www.Mailgate.ORG
> wikipedia does have 201 candidates. They include what you stated
> plus
> "* 6 de-facto independent countries (Abkhazia, Nagorno-Karabakh,
> Northern Cyprus, Somaliland, South Ossetia, Transnistria)
> * 2 entities recognized by many countries but de-facto not
> independent, Palestine and Western Sahara."
> I thought there were more. One site claimed over 260, but I did
> not get into the details.
Well, there's the Sovereign Military Hospitaller Order of St. John of
Jerusalem of Rhodes and Malta
<URL:http://www.orderofmalta.org/domrisp.asp?idlingua=5>, which owns a
couple of buildings in Rome.
--
John W. Kennedy
"The bright critics assembled in this volume will doubtless show, in
their sophisticated and ingenious new ways, that, just as /Pooh/ is
suffused with humanism, our humanism itself, at this late date, has
become full of /Pooh./"
-- Frederick Crews. "Postmodern Pooh", Preface
>Gene Wirchenko wrote:
>> Esa A E Peuha <esa....@helsinki.fi> wrote:
>>>Over 200? There are 191 members of the United Nations, then there's
>>>Vatican and Taiwan, so the total is 193. How can you get over 200?
>
>> wikipedia does have 201 candidates. They include what you stated
>> plus
>> "* 6 de-facto independent countries (Abkhazia, Nagorno-Karabakh,
>> Northern Cyprus, Somaliland, South Ossetia, Transnistria)
>> * 2 entities recognized by many countries but de-facto not
>> independent, Palestine and Western Sahara."
>
>> I thought there were more. One site claimed over 260, but I did
>> not get into the details.
>
>Well, there's the Sovereign Military Hospitaller Order of St. John of
>Jerusalem of Rhodes and Malta
><URL:http://www.orderofmalta.org/domrisp.asp?idlingua=5>, which owns a
>couple of buildings in Rome.
And IIRC, it is limited to a permanent population of three.
> Hi,
>
> > This place will be called www.ifreviews.org (i already have the domain), and
> > it will be based on PHP/MySQL, to enable speed searchs/access, but specially
> > online reviews editing :)
>
> I would suggest mandatory session transcript for the review.
Which would mean that existing reviews cannot be included in the new site
unless the reviewer decided to keep the associated transcripts. I seem to
recall RootShell saying he really wanted people to be able to plug their
old reviews into his new system.
/====================================================================\
|| Quintin Stone O- > "You speak of necessary evil? One ||
|| Code Monkey < of those necessities is that if ||
|| Rebel Programmers Society > innocents must suffer, the guilty must ||
|| st...@rps.net < suffer more." -- Mackenzie Calhoun ||
|| http://www.rps.net/QS/ > "Once Burned" by Peter David ||
\====================================================================/
> > Which would mean that existing reviews cannot be included in the new
> > site unless the reviewer decided to keep the associated transcripts
>
> Well, we can make exceptions for old reviews, can we? ;)
> Furthermore...
> Don't you think that transcripted session is useful?
> Reviewing an IF game is not like reviewing a book, which is "the same",
> considering just the text to be read, for all readers.
Useful? To the author, surely. To me? No. I doubt I'd ever read one
(if I wasn't the author of the game in question). Clearly if I haven't
played the game I wouldn't want to spoil it by reading the transcript.
On the other hand, if I have played the game I don't have a lot of
interest in rehashing the entire experience statically just to see where
someone else's choices differed from mine.
This is just my view and certainly others may have a differing opinion.
==--- ---==
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
From the author's point of view: transcripts are great as they allow you to
see what people are trying with your games. They give you a real insight
into improving your game to best fit the things people attempt to do.
From a player's point of review: they're a waste of time. I pore over
transcripts people produce for my own games but I've never bothered reading
any for other people's games.
[snip]
> From the author's point of view: transcripts are great as they allow you
> to see what people are trying with your games. They give you a real
> insight into improving your game to best fit the things people attempt to
> do.
>
> From a player's point of review: they're a waste of time. I pore over
> transcripts people produce for my own games but I've never bothered
> reading any for other people's games.
After carefully reading the posts and thinking about it, I really must agree
with David Whyld (and others) that transcripts are not a
necessity/obligation for a review, nevertheless, i think that if available
they should be acessible by users.
Meaning that transcripts will be shown (if available) as a link at the end
of the game review (so that only if the user clicks on the link he will see
it).
The *main* reason for this posture, is to prevent users (who will play the
games if they havent already) from reading transcripts inadvertly which (and
i think is general idea) will/could ruin the game experience.
This could also lead to the discussion of another feature that has been
requested by some of you (by email) about including a 'possible' solution
(since there might be several diferent ones to some games) also with the
review.
While not on the same level of spoilness as the transcript, i think that it
could also be very interesting to know what the rest of the comunnity thinks
about this.
The criteria would be somehow similar to the transcript, meaning that a link
(for viewing/download) would be shown (if available) rather than the
solution text.
That way, only who wants to see them could.
Something like:
==========================================
|| Name of The Game: X
|| Game Autor: W
|| Reviewed by: Y
||
|| Review: bla bla bla blabla bla bla. bla bla blabla bla bla.
|| bla bla bla blabla bla bla,bla bla blabla
|| bla bla bla blabla bla blabla bla blabla.
||
|| Rated: *****
||
|| Transcript available ( view | download )
==========================================
Or...
==========================================
|| Name of The Game: X
|| Game Autor: W
|| Reviewed by: Y
||
|| Review: bla bla bla blabla bla bla. bla bla blabla bla bla.
|| bla bla bla blabla bla bla,bla bla blabla
|| bla bla bla blabla bla blabla bla blabla.
||
|| Rated: *****
||
|| Solution available ( view | download )
==========================================
There's still plenty of time to decide this but i guess this would be a
suitable for everyone (autors/reviewers/players) :)
Regards,
RootShell
[snip]
>That way, only who wants to see them could.
>
>Something like:
>
>==========================================
>|| Name of The Game: X
>|| Game Autor: W
>|| Reviewed by: Y
>||
>|| Review: bla bla bla blabla bla bla. bla bla blabla bla bla.
>|| bla bla bla blabla bla bla,bla bla blabla
>|| bla bla bla blabla bla blabla bla blabla.
>||
>|| Rated: *****
>||
>|| Transcript available ( view | download )
Or
Additional information:
____________________ ( view | download )
...
where the reviewer fills in the blank, say
Additional information:
transcript ( view | download )
additional hints about the treasure chest puzzle ( view |
download )
walkthrough ( view | download )
Easter Eggs ( view | download )
The tresure chest is a puzzle? Arrgh, you just spoiled it for me.
:-)
With all the hair on the ground around the treasure chest, it was
a giveaway that it was a puzzle.
Well, we can make exceptions for old reviews, can we? ;)
Furthermore...
Don't you think that transcripted session is useful?
Reviewing an IF game is not like reviewing a book, which is "the same",
considering just the text to be read, for all readers.
Rob