Tales from a mad student

JacqJacq Join Date: 2002-04-04 Member: 378Members
in NS General Discussion
<div class="IPBDescription">Estimating respawn time</div>May be i´m mad but i´m studying "Queueing Systems" for my September exam, i don´t know how is respawn time estimated for this mod, but if you give me stadistical data of the absolute deads in an game, i could obtain maximun respawn time allowed for a maximun waiting time in the respawn queue, it could be estimated for any numbers of hives/portals. All stadistical data about the game will be very usefull (number of players, time between direct confrontation...)

May be all this study will be usefull after all  <!--emo&:p--><img src="http://www.natural-selection.org/iB_html/non-cgi/emoticons/tounge.gif" border="0" valign="absmiddle" alt=':p'><!--endemo-->
Off Topic Disagree Agree Awesome

Comments

  • DOOManiacDOOManiac Worst. Critic. Ever. Join Date: 2002-04-17 Member: 462Members, NS1 Playtester
    I'm not a dev, but this is something I've talked with flayra a bit on cause I spent some time trying to reproduce a now squashed bug that caused long respawn time.

    the NS respawn system is based on a queue (FIFO - First in First Out) system, meaning the first to die is the first to respawn. However, the issue gets a litle more complicated than you'd think. When a player "dies" there is a 2 second black screen period before you pop into observer mode (are officially "dead"), so you are not popped into the queue until this happens. meaning that if nobody else on your team is waiting to respawn and the respawn time happens right in that 2 second space between life and death, then you miss the respawn. *I COULD BE WRONG ON THAT THOUGH*

    Also, the number of hives or infantry portals affect how many spawn at once. 3 hives = 3 spawns, 2 inf portals = 2 spawns. everybody else just keeps waiting in line...

    there used to be an old old bug of never respawning players but flayra has just recently squashed it! :D
    Off Topic Disagree Agree Awesome
  • FlayraFlayra Game Director, Unknown Worlds Entertainment San Francisco Join Date: 2002-01-22 Member: 3Super Administrators, NS2 Developer, Subnautica Developer
    Yeah, you're wrong about that Doom.  There's no special "respawn time" that you can miss.  If a portal or hive sees a dead player, it starts respawning them, and pops them out after 10 seconds.  There are no "waves" or special times for respawning.
    Off Topic Disagree Agree Awesome
  • JacqJacq Join Date: 2002-04-04 Member: 378Members
    So, you don´t need any stadistical information about your respawn queue?
    mean time to respawn
    mean number of players waiting to respawn
    probability of finding 0,1,2,... players in the queue
    ...
    Off Topic Disagree Agree Awesome
  • LockNLoadedLockNLoaded Join Date: 2002-09-05 Member: 1282Members
    jacq: are u studying quality assurance? it does sound like it- the words 'mean', 'median'..etc =D
    Off Topic Disagree Agree Awesome
  • DOOManiacDOOManiac Worst. Critic. Ever. Join Date: 2002-04-17 Member: 462Members, NS1 Playtester
    there aren't waves? oh.. heh.   silly me :)

    not the first time i've been wrong, and won't be the last time either :P
    Off Topic Disagree Agree Awesome
  • Stan_RStan_R Join Date: 2002-09-13 Member: 1311Members
    Eh, I doubt that you can alalyze this with queuing theory my friend.  Player life time does not depend on a probability distribution, but on skill.  

    So lifespans and "deathspans" (i.e. waiting times until next respawn) are not uniform or Poisson, but unpredicatable.  You can try saying  "an average player will stay alive for so long, on average," but averages depend on map, and on any given server there will be players of various skill.  In addition, deathspans also seem to be affected by the number of hives and amount of resources.

    So, to summarize:  interarrival times of players are unpredictable, the "server" (read: hive or base) properties depend on the progression of the game. Therefore queueing theory is useless.

    Nice thought though.  Let's bring more science to NS (I can help if necessary).
    Off Topic Disagree Agree Awesome
  • CMEastCMEast Join Date: 2002-05-19 Member: 632Members
    He isnt trying to work out how long the average marine will live, he is trying to work out how long it takes on average to respawn, from when you first die to when you spawn back in.
    Off Topic Disagree Agree Awesome
  • JacqJacq Join Date: 2002-04-04 Member: 378Members
    Stan R. is right, the problem in estimating queueing time is like in many situations estimating the human do. He knows what queueing systems are about. But the problem with different maps and players is already with the fixed respawn time, and developers can´t know how will be the waiting in any type of game. Some sciencie always is god  <!--emo&:p--><img src="http://www.natural-selection.org/iB_html/non-cgi/emoticons/tounge.gif" border="0" valign="absmiddle" alt=':p'><!--endemo--> , death time depends surelly on skill but in many other facts (lucky, team, daylight...) so is there where fdp are usefull, anyway having stadistical data about deaths and using a fdp to model it will be interesting
    . And if you know any secret about queuing systems that may help in my exam i will love to hear you  <!--emo&:D--><img src="http://www.natural-selection.org/iB_html/non-cgi/emoticons/biggrin.gif" border="0" valign="absmiddle" alt=':D'><!--endemo--> .
    Last but not least, WELCOME to the forummmm, and sorry my spanglish
    Off Topic Disagree Agree Awesome
  • MavericMaveric Join Date: 2002-08-07 Member: 1101Members
    The average life expectancy of a marine in combat is a grim 9 seconds. <!--emo&:)--><img src="http://www.natural-selection.org/iB_html/non-cgi/emoticons/smile.gif" border="0" valign="absmiddle" alt=':)'><!--endemo-->

    Put that into your tables. <!--emo&:p--><img src="http://www.natural-selection.org/iB_html/non-cgi/emoticons/tounge.gif" border="0" valign="absmiddle" alt=':p'><!--endemo-->
    Off Topic Disagree Agree Awesome
Sign In or Register to comment.