Should a slower demo get listed as WR

Started by thopa
13/11/2008 22:42
Page 1 of 2
thopa13/11/2008 22:42

Weird topic name. But should a demo with slower record time get accepted if the ingame kz timer is quicker; and if the margins are VERY small.

So. What is your opinion?

My decision about heL^_x's demo on kzno_xtremeblock came as a shock to many of you. I want to hear your opinions about this. Take heL^_x's demo vs Fact or Fiction's demo as an example. heL^_x's ingame timer stops at 0159 and FoF's at 0200. heL^_x is slower when you check the REAL TIME.

You know where I'm standing. I want to know your opinions for future similarities.

kzno_xtremeblock_heL^_x_0159.rar
Fact or Fiction's xtremeblock is 0200 @ 120.9761 sec
heL^_x's xtremeblock is 0159 @ 121.1514 sec
Picture

r0fl13/11/2008 22:47

on topic: dont know but
off topic: i think demos that have the same in-game timer as a world record but a faster time in fps death should be accepted so those useless 5 second demos will be finally beaten :D...

Dacca13/11/2008 22:49

I'm not quite decided, but I'm going to think about it and post my comment.

P.S. I think that the battle movie would solve this problem :D

thopa13/11/2008 22:58

Originally posted by r0fl
on topic: dont know but
off topic: i think demos that have the same in-game timer as a world record but a faster time in fps death should be accepted so those useless 5 second demos will be finally beaten :D...



I made this one. Continue here. :p
http://xtreme-jumps.eu/e107_plugins/forum/forum_viewtopic.php?129297.last

SSJGoku13/11/2008 23:07(edited 13/11/2008 23:10)

Well yes, this is kinda weird, but still, if we remember unreal skilled Tonic's time on @ cupiators ( the 58 sec which noone could beat untill the gay sc), where he was the first to get WR with a time of 58 sec, i actually geeked that map for like 2 weeks nonstop, and got 58 sec all the time, but i was like 0.5 sec faster than Tonic ( according to fps_death), than it was all about what timer says, if there isnt timerbugs or some signs or cheating.

However, this is a little different, and the chances of getting this kind of a result is just ridiciously small.

What i think, thopa made a right decision, even tho the time is xtremely awesome, ofc its obvious you can do better, you can see the explanation below:

For example, if something like this happens again, the fallowing step should be taken:
Check the maps length!
because in 2 min map , which is xtreme rated by itself, you can do better if you have the skill for it, thereby denieing hel's demo because of demo_analyzers results( this apply if its totally sure that analyzer can _NOT_ make miscalculations by any incoming data ) ,
however, if a person gets WR in a easy rated map which is ( for example ) below 1 min , and another gets a better time with slower analyze time, it shouldnt be accepted, because little maps are possible to make like perfect, just imagine zr_minimountain to get a better time than brian with slower dem_analyzing time, ofc the previous record should remain, just like thopa did.

About getting the 0.1 sec WR times, than shouldnt happen only because you can be faster at every single jump, but press the end key somehow 0.2 sec too late, which you couldnt even notice, and be a WR holder with a slower demo.
Concluding what my point was if some1 is to lazy to ready my story, thopa made the right decision no matter which side you look at it, why is that you can check above in my example.
:)

anny013/11/2008 23:11

I already talk to you about this @ irc thopa.. and as you said using the record time would be better `cause draw demos would be check !
So we need a new rule...

Ewoud13/11/2008 23:24

use the demolyser to check demos, people are getting and better in kz and in the future more demos with marginal time differences will be submitted, my 2 cents

PepoMerca14/11/2008 01:33(edited 14/11/2008 01:35)

hel had timerbug then?

Pulsar!14/11/2008 06:03

Originally posted by Dacca
P.S. I think that the battle movie would solve this problem :D

boom.mdl ? Where u ?

cheche14/11/2008 10:00(edited 14/11/2008 20:49)

I completely agree with your decision thopa and tbh you would have got flamed either way - and you probably knew that already, right?

The timer in KZ-maps is really not up to date with kz ano 2009.

And I really don't like the fact that someone might beat Bibo's coldbhop_v2 demo with almost a second, and if they are abit unlucky they would still not get the wr.

Or when you download the wr for clintmo_bhoptoon you would expect that this is the fastests demo ever on this(just as with kzno_xtremeblock) - but in fact you can just go to CG's website and download a demo that is atleast 0.5 seconds better with more skips - and in this case the demo has the same time as in FPS-Death as in the map.

It just doesn't seem right, imo.
But if you can make this decision, it really should open up for some other changes aswell.

leggie14/11/2008 10:03

i want boom to make a movie so you can see who hits the end timer first. and if hel does then what then? i know you had to do it thopa, just sucks that it had to be on such a good map/time :/

boom.dll14/11/2008 10:30

i`m @ work now... i`ll make it this evening... i`m very curious myself... tho i think that the ingame timer is pretty much the REASON why people geek wr... if i remeber right... most of cordyline`s demos are slower than the previous wrs... i`m not sure tho...

i`ll make the battle movie AS soon as i get home...nothing fancy just the movie!

this is gonna be pretty awesome imo :D

qewi14/11/2008 10:43

nice boom.dll! :D

i think that you made the right decision thopa, but i dont think that you should be able to beat the wr just cause of the "real time". i mean like, if someone got kz_xxx in 10 secs in both real timer and maptimer and someone else gets like 11 secs on the maptimer but 9 on the realtimer i think that the old record should remain.

Grrr14/11/2008 10:53(edited 14/11/2008 10:53)

maybe fps is the reason ? O.o

Om114/11/2008 12:01

the wr should be set by the map-timer, just dont accept demos from a server with more than 60min uptime

trb14/11/2008 14:15(edited 14/11/2008 14:16)

The main question i have is the thing which i don't really unterstand.

point1: exact time?
many times i heard this "record time" on demolyser/fps death is the right time, and the kz timer isn't 100% accurate. so 1.00000000000.... seconds in real time isn't 1.0000000000... second on kz timer.

point1: timerbug?
the chance that a timerbug happens after 60 minuts running time will happen is big, but its not sure that it will happen.
the chance that a timerbug happens before 60minutes uptime is small, but it could happen too.
when u have a timerbug, than the kz timer is more inaccurate then it normally is (afaik).

so the point is, from which point we see if its a timerbug or just the inaccuracy from kz timer.
from beginning in kz we use kztimer, which we now that its not 100% accurate, but we decline timerbug demos (which are less accurate then normal kz timer).

what i would do on this demo?
So if it's a timerbug in his demo, then decline the demo, if its no timerbug, then accept it.
don't understand it wrong way, theoretically all demos have a "timerbug" in sence "bug in time" because normal kz timer isnt 100% accurate. i mean in this example (accept or decline example) the "big timerbug", for us known normal as "official" timerbug.

things to add?
to get back to point 1, if record time is the real time, so how can it be that fairy_island +secret in real time would have 5 seconds?
I didn't watch the demo now but i don't think that 11 seconds on kz timer will be 5 realtime seconds.
And why not every demo has a "record time" shown in demolyser/fps death? some demos have, some don't.
Btw i accept with some persons on chatbx, which said that if it would be from demolyser time and not kz timer, then many actual demos would have this bug.
And i have to say it probably is true.
I just downloaded 4 short demos:
rpz_downfall[boxes]_faker_0005
risk_compound_Supa_0010
kz_fairy_island[+secret]_brian_0011
kz_man_bhopsnow_HerloTz_0000
3 of this 4 demos had this problem too.
faker demo was correct record time in demolyser (5.x)
fairy_island (11sec ingame, 5sec demolyser) http://img370.imageshack.us/my.php?image=fasteryz8.jpg
bhopsnow (0 sec ingame, no time demolyser) http://img91.imageshack.us/my.php?image=notimeah3.jpg
compound (10 sec ingame, 11 sec demolyser) http://img389.imageshack.us/my.php?image=slowerti2.jpg

so the main point is, what is the exactelly bug in hel's demo? just the normal knowing inaccuacy of kz timer like in many other demos too?
if we should take the demolyser times? 5 seconds on fairy island then? (this would be a much more bad thing than the xtremeblock thingy, because it would make the time here to less than the half of the kz timer time ):-D

I know too less about the engine, the calculation of time there, the calculation of time in fps death and so that i could say "do it that way and do it that way then it's right", but i think the some things above can give some hints which should be involved in making a descision.



edit: god damnit why my posts everytime end up in that much text :D

mucholote14/11/2008 15:53

Well if you're 100% sure that fps death calculates the exact time and its the real time well that's the time to follow.
I still don't get the fairy island one... 6 sec is a huge difference. What about taking a crono and measure the time.
Sounds lame and old fashion, but it will give you a better look, on fairy at least since the dif is so big.

pd15/11/2008 11:27

Stop recording with cracked NON Steam.

PepoMerca15/11/2008 11:53

Originally posted by pd
Stop recording with cracked NON Steam.

why? there is ppl with low fps and no steam has more fps than steam

pd15/11/2008 12:08(edited 15/11/2008 12:09)

its illegal.
and maybe this bug´s comes from this.

Page 1 of 2

Please log in to post replies.