Respawn Times Math

Kerostasis · 03-12-2006, 10:23 AM

A while back I posted in the "Gorges in Res Nodes" thread about how long it took to respawn under various conditions. Then I went and did some testing of my numbers, and found out a few of them were off. I cant post the update there though, because its locked. So here you go. Just correcting my misinformation.

After any player dies, his status is frozen as "dead" for approximately 3 seconds. I thought this used to be longer at some time in the past, but it might have changed at some point without me noticing. After 3 seconds, he is added to the spawn queue.

Once in the spawn queue, a marine waits for an IP not busy processing another dead player. Each IP can process one player every 10 seconds. Once an IP has begun processing you, you are stuck with that one. If it gets destroyed, you go back to the end of the spawn queue, even if there are more IPs built. If a new IP is built, it can process its first marine almost instantly, but it wont steal one already being processed by another IP.

An alien in the spawn queue waits for a Hive not busy spawning another player, using the same logic. It's often harder to see the process at work though, since an alien doesn't visibly emerge from a hive the way a marine does from an IP, and Hives are much harder to build and destroy than IPs are. Each Hive can process one alien every 8 seconds, normally. However, in games larger than 6v6 with less than 3 Hives, that time is partially reduced to bring the spawn rate closer to that of a 6v6 game. I haven't timed this one exactly, but the spawn time reduction is a little less at 2 Hives than at 1 Hive, and disappears entirely at 3 Hives, or with 6 or less players.

Example time: For anyone who was confused, now you can be even more confused with mathematical examples!
So if one marine dies in the field, he will be back alive in 13 seconds. If 3 marines die (on 1 IP), they will return 13, 23, and 33 seconds later. If a second IP is built at the 20 second mark, the third marine will respawn instantly, actually beating the second one back to life. If a second IP is built at the 24 second mark, its too late and the third marine still has to wait his full 10 seconds.

Now say 3 marines die in the field, on 2 IPs, but the first IP is being attacked. It is destroyed before marine 1 respawns. Marine 2 respawns from IP2 at 13 seconds, Marine 3 respawns at IP2 at 23 seconds, and poor Marine 1 has to wait for last!

Finally, note that the order you build your IPs (or hives) decides which one you will spawn from. It isn't random. 3 seconds after a player dies, the first built IP or Hive will claim him and begin to respawn him. If a second player dies whie the first is still spawning, he will be assigned to the second built IP/Hive. This is really only important to know when you have IPs in 2 different places on the map (something I do occasionally in very small games), and which IP you spawn from can be very important. If IP or Hive #1 is killed and rebuilt, it goes to the back of the list.

**Wyzcrak** · 03-12-2006, 05:14 PM

Very helpful, Kero. Many thanks.

rs_al · 03-20-2006, 03:53 PM

Great post.

Quote:

Originally Posted by Kerostasis

Finally, note that the order you build your IPs (or hives) decides which one you will spawn from. It isn't random. 3 seconds after a player dies, the first built IP or Hive will claim him and begin to respawn him. If a second player dies whie the first is still spawning, he will be assigned to the second built IP/Hive. This is really only important to know when you have IPs in 2 different places on the map (something I do occasionally in very small games), and which IP you spawn from can be very important. If IP or Hive #1 is killed and rebuilt, it goes to the back of the list.

This has a pretty big impact for aliens. It means they have a very high chance of respawning at their first hive, and also means they'll never spawn at their third hive unless there are at least two already dead aliens when they die. This is a big reason why second/third hives are usually easier for marines to attack, aliens almost never spawn there giving marines an easier time doing the whole ninja thing should aliens be lax in their patrolling.

I also did the two sets of IPs thing recently on ns_ayumi, 1 at MS, 2 at server outlet (?) just south of Pressure hive. That was fun

Kerostasis · 03-20-2006, 08:39 PM

Right--I hadn't mentioned that because I thought that MCs would make it irrelevant which hive you spawned at, but thats a good point. If you are trying to sneak in a ninja phase, the #2 and 3 hives will have far fewer aliens spawning at them.

**Wyzcrak** · 03-20-2006, 09:32 PM

Ohhhh... good point.

Irish Pirate · 03-21-2006, 01:48 AM

Ninja phase aside, in a rush where marines are moving into a hive I think it's better to spawn in the distant hive and movement it. It lets you control when you enter the battle and you can prepare yoruself instead of anxiously waiting for your appearance, and the good chance you will be shot before you can move and figure out whats going on.

Also gives you time to get upgrades if you have any available that would help in defending the hive.

**TheAdj`** · 03-21-2006, 04:50 AM

On a similar note: Phase gates do not necessarily follow this ordered logic. I know we spent 4 hours in a playtest session trying to figure out how they work and it gets really chaotic after the first 3 or so.

rs_al · 03-21-2006, 09:30 AM

Quote:

Originally Posted by TheAdj`

On a similar note: Phase gates do not necessarily follow this ordered logic. I know we spent 4 hours in a playtest session trying to figure out how they work and it gets really chaotic after the first 3 or so.

I've noticed this too. I would assume they are in entity list order. When half life creates an entity, it assigns it an integer ID that isn't currently in use (half life stores entities in a plain old-fashioned array). Usually when you look up entities in half life, you say "start with this index, find me the next entity of type X". So if phase gates are entities 10, 20, and 30, then when you use phase 20 it tries to find the next pg starting with 20, iterating until it hits 30 and phases you there. But the next time you build a pg, it could be assigned 25 if the entity that was previously 25 has disapeared (maybe it was a building that got destroyed, or any number of temporary entities such as welder effect). So now pg #25 will appear in the order between 20 and 30.

That is just my theory it's been a while since I muddled about in the hlsdk, but I've got a good feeling that's what it is.

Kerostasis · 03-21-2006, 02:14 PM

*tests PGs on a local server for awhile*

I think rs_al is pretty close to correct. I havent checked items like welder effects or marine equipment because I'm not sure how, but I tested the buildings fairly thouroughly.

Every time a building is created, it gets assigned a number. Every time a building is destroyed or recycled, that building number is freed up again. The next building created will pick up the smallest available free number, creating excellent possibilities for phase gates to get out of order.

Edit: I hadn't considered this earlier, but theres a good chance that IPs and Hives actually follow the same logic as Phase Gates, but you don't see it as often because IPs and Hives arent destroyed and rebuilt quite as often as PGs. But you do see some games where Hives are destroyed and rebuilt a couple of times, and I'll bet if you pay attention you'll sometimes see them get out of order the same way Phase Gates do.

Edit 2: After some testing on the hives idea, I cant come to an solid conclusion. Its too complex. Feel free to offer your own evidence.

rs_al · 03-22-2006, 09:55 PM

I bet that hives and ips are probably done seperately, because of the complexities of the respawn queue, like how a marine "blocks" an ip while he is respawning there. The default half life respawn process does work like the phasegate thing (iterating over all player starts, starting with a global to track the last player's respawn spot, until it finds a spot not currently occupied by some other entity), and there's a good chance that ns uses that, but only for spawning at the start of the game and beacons, or for selection of a spawn point within a hive after the hive has been selected, for alien respawns and mcs.

They probably do not use that logic for the higher level calculations like picking a non-busy ip/hive. It's pretty likely the game just uses a seperate list of ips/hives that it looks at whenever it needs to respawn a player.

**TheAdj`** · 03-27-2006, 10:02 AM

We tried pretty much everything you can think of and it was still random. We built other buildings, built nothing, dropped nothing, and no one was on aliens. It was an exercise in futility once the number of PGs went too high (over 3-4). If you can find a pattern I'd be interested.

Darkilla · 03-29-2006, 12:58 AM

I always thought PGs worked like : You phase to the lastest, then second latest, then first latest..

perhaps a developer could answer this?

Kerostasis · 03-29-2006, 08:41 AM

Nope. You phase to the first, then second, then third, and so on. It doesn't go in reverse order.

However, it often seems like it goes in reverse order, because the games way of keeping track of which one is "first" gets confused so often.

**zibalatz** · 03-29-2006, 10:09 AM

Quote:

Originally Posted by Darkilla

perhaps a developer could answer this?

BWahahahahahahahahahahahahahahahahaha!

Oh wait, were you serious?

I'm sure they'll get back to you on that, within one NS Week*.

*One NS Week is equivalent to one standard Earth Week +/- NS Forums downtime.

**Wyzcrak** · 03-29-2006, 10:29 AM

lol one ns week

03-12-2006, 10:23 AM	#1 (permalink)
Kerostasis Join Date: May 2005 Age: 24 Posts: 2,438	Respawn Times Math A while back I posted in the "Gorges in Res Nodes" thread about how long it took to respawn under various conditions. Then I went and did some testing of my numbers, and found out a few of them were off. I cant post the update there though, because its locked. So here you go. Just correcting my misinformation. After any player dies, his status is frozen as "dead" for approximately 3 seconds. I thought this used to be longer at some time in the past, but it might have changed at some point without me noticing. After 3 seconds, he is added to the spawn queue. Once in the spawn queue, a marine waits for an IP not busy processing another dead player. Each IP can process one player every 10 seconds. Once an IP has begun processing you, you are stuck with that one. If it gets destroyed, you go back to the end of the spawn queue, even if there are more IPs built. If a new IP is built, it can process its first marine almost instantly, but it wont steal one already being processed by another IP. An alien in the spawn queue waits for a Hive not busy spawning another player, using the same logic. It's often harder to see the process at work though, since an alien doesn't visibly emerge from a hive the way a marine does from an IP, and Hives are much harder to build and destroy than IPs are. Each Hive can process one alien every 8 seconds, normally. However, in games larger than 6v6 with less than 3 Hives, that time is partially reduced to bring the spawn rate closer to that of a 6v6 game. I haven't timed this one exactly, but the spawn time reduction is a little less at 2 Hives than at 1 Hive, and disappears entirely at 3 Hives, or with 6 or less players. Example time: For anyone who was confused, now you can be even more confused with mathematical examples! So if one marine dies in the field, he will be back alive in 13 seconds. If 3 marines die (on 1 IP), they will return 13, 23, and 33 seconds later. If a second IP is built at the 20 second mark, the third marine will respawn instantly, actually beating the second one back to life. If a second IP is built at the 24 second mark, its too late and the third marine still has to wait his full 10 seconds. Now say 3 marines die in the field, on 2 IPs, but the first IP is being attacked. It is destroyed before marine 1 respawns. Marine 2 respawns from IP2 at 13 seconds, Marine 3 respawns at IP2 at 23 seconds, and poor Marine 1 has to wait for last! Finally, note that the order you build your IPs (or hives) decides which one you will spawn from. It isn't random. 3 seconds after a player dies, the first built IP or Hive will claim him and begin to respawn him. If a second player dies whie the first is still spawning, he will be assigned to the second built IP/Hive. This is really only important to know when you have IPs in 2 different places on the map (something I do occasionally in very small games), and which IP you spawn from can be very important. If IP or Hive #1 is killed and rebuilt, it goes to the back of the list.

03-12-2006, 05:14 PM	#2 (permalink)
Wyzcrak Join Date: May 2003 Location: Memphis, TN Posts: 14,138	Re: Respawn Times Math Very helpful, Kero. Many thanks. __________________ Steam Community? Add me. \| Join #tacticalgamer \| Search Results Legend \| New Posts Forum Filter \| Postbox Toggle \| Live Thread Review \| One Line Results \| Free Remote, Encrypted Backup Darkilla: In short, NS is pretty much really fast chess. With guns. Apophis: I haven't seen anyone say that SM's are better than non-SMs. Nordbomber: This is THE first server I've seen where either side can comeback from out of seemingly nowhere with the right teamwork. en4rcment: I have NEVER experienced the type of gameplay that I have found here. Nightly I am amazed at the personalities and gaming talent. Root: Welcome to TG. Feel free to punctuate your sentences correctly. Monkerz: Its gonna take all my skills to beat those boys off in the future.

03-20-2006, 08:39 PM	#4 (permalink)
Kerostasis Join Date: May 2005 Age: 24 Posts: 2,438	Re: Respawn Times Math Right--I hadn't mentioned that because I thought that MCs would make it irrelevant which hive you spawned at, but thats a good point. If you are trying to sneak in a ninja phase, the #2 and 3 hives will have far fewer aliens spawning at them.

03-20-2006, 09:32 PM	#5 (permalink)
Wyzcrak Join Date: May 2003 Location: Memphis, TN Posts: 14,138	Re: Respawn Times Math Ohhhh... good point. __________________ Steam Community? Add me. \| Join #tacticalgamer \| Search Results Legend \| New Posts Forum Filter \| Postbox Toggle \| Live Thread Review \| One Line Results \| Free Remote, Encrypted Backup Darkilla: In short, NS is pretty much really fast chess. With guns. Apophis: I haven't seen anyone say that SM's are better than non-SMs. Nordbomber: This is THE first server I've seen where either side can comeback from out of seemingly nowhere with the right teamwork. en4rcment: I have NEVER experienced the type of gameplay that I have found here. Nightly I am amazed at the personalities and gaming talent. Root: Welcome to TG. Feel free to punctuate your sentences correctly. Monkerz: Its gonna take all my skills to beat those boys off in the future.

03-21-2006, 01:48 AM	#6 (permalink)
Irish Pirate Join Date: Jul 2005 Age: 20 Posts: 483	Re: Respawn Times Math Ninja phase aside, in a rush where marines are moving into a hive I think it's better to spawn in the distant hive and movement it. It lets you control when you enter the battle and you can prepare yoruself instead of anxiously waiting for your appearance, and the good chance you will be shot before you can move and figure out whats going on. Also gives you time to get upgrades if you have any available that would help in defending the hive.

03-21-2006, 04:50 AM	#7 (permalink)
TheAdj` Join Date: May 2004 Posts: 370	Re: Respawn Times Math On a similar note: Phase gates do not necessarily follow this ordered logic. I know we spent 4 hours in a playtest session trying to figure out how they work and it gets really chaotic after the first 3 or so.

03-21-2006, 02:14 PM	#9 (permalink)
Kerostasis Join Date: May 2005 Age: 24 Posts: 2,438	Re: Respawn Times Math tests PGs on a local server for awhile I think rs_al is pretty close to correct. I havent checked items like welder effects or marine equipment because I'm not sure how, but I tested the buildings fairly thouroughly. Every time a building is created, it gets assigned a number. Every time a building is destroyed or recycled, that building number is freed up again. The next building created will pick up the smallest available free number, creating excellent possibilities for phase gates to get out of order. Edit: I hadn't considered this earlier, but theres a good chance that IPs and Hives actually follow the same logic as Phase Gates, but you don't see it as often because IPs and Hives arent destroyed and rebuilt quite as often as PGs. But you do see some games where Hives are destroyed and rebuilt a couple of times, and I'll bet if you pay attention you'll sometimes see them get out of order the same way Phase Gates do. Edit 2: After some testing on the hives idea, I cant come to an solid conclusion. Its too complex. Feel free to offer your own evidence. Last edited by Kerostasis; 03-21-2006 at 06:06 PM.

03-22-2006, 09:55 PM	#10 (permalink)
rs_al Join Date: Oct 2004 Posts: 499	Re: Respawn Times Math I bet that hives and ips are probably done seperately, because of the complexities of the respawn queue, like how a marine "blocks" an ip while he is respawning there. The default half life respawn process does work like the phasegate thing (iterating over all player starts, starting with a global to track the last player's respawn spot, until it finds a spot not currently occupied by some other entity), and there's a good chance that ns uses that, but only for spawning at the start of the game and beacons, or for selection of a spawn point within a hive after the hive has been selected, for alien respawns and mcs. They probably do not use that logic for the higher level calculations like picking a non-busy ip/hive. It's pretty likely the game just uses a seperate list of ips/hives that it looks at whenever it needs to respawn a player.

03-27-2006, 10:02 AM	#11 (permalink)
TheAdj` Join Date: May 2004 Posts: 370	Re: Respawn Times Math We tried pretty much everything you can think of and it was still random. We built other buildings, built nothing, dropped nothing, and no one was on aliens. It was an exercise in futility once the number of PGs went too high (over 3-4). If you can find a pattern I'd be interested.

03-29-2006, 08:41 AM	#13 (permalink)
Kerostasis Join Date: May 2005 Age: 24 Posts: 2,438	Re: Respawn Times Math Nope. You phase to the first, then second, then third, and so on. It doesn't go in reverse order. However, it often seems like it goes in reverse order, because the games way of keeping track of which one is "first" gets confused so often.

03-29-2006, 10:29 AM	#15 (permalink)
Wyzcrak Join Date: May 2003 Location: Memphis, TN Posts: 14,138	Re: Respawn Times Math lol one ns week __________________ Steam Community? Add me. \| Join #tacticalgamer \| Search Results Legend \| New Posts Forum Filter \| Postbox Toggle \| Live Thread Review \| One Line Results \| Free Remote, Encrypted Backup Darkilla: In short, NS is pretty much really fast chess. With guns. Apophis: I haven't seen anyone say that SM's are better than non-SMs. Nordbomber: This is THE first server I've seen where either side can comeback from out of seemingly nowhere with the right teamwork. en4rcment: I have NEVER experienced the type of gameplay that I have found here. Nightly I am amazed at the personalities and gaming talent. Root: Welcome to TG. Feel free to punctuate your sentences correctly. Monkerz: Its gonna take all my skills to beat those boys off in the future.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Good Times.......	TheFeniX	Natural Selection - General Discussion	12	01-06-2005 05:09 PM
Good times...good times...	TG_JBRanger	Joint Operations - General Discussion	5	09-24-2004 11:53 AM
Math through the years... (Where's Skylark?)	CingularDuality	The Sandbox	10	08-03-2004 08:55 PM

Sponsored links

Sponsored links

Sponsored links

Sponsored links