Spawn rules and protection in ut4 and arenafps in general

This post started due to..

And I lost interest, then heard the beyond unreal podcast 20 and some unusual comments prompted me to flesh it out and post. I will preface this by saying I like arena FPS because the random aspects and like spawn selection being out of the players hands. However the current handling does not work all that well either, so lets look at some possible changes.

Enjoy and correct me where I need correcting please.

Recap TLDR version without reasoning

A spawn system appropriate for FFA, duel and TDM. The aim is to remove spawn protection with a better spawn system.

The basic idea is to make spawning players less predictable so they cannot be picked off instantly on spawn.

What is spawn protection

Spawn protection aims to give spawning players some sort of chance. Players should not spawn then die, or so the concept goes. It only sort of works in games I have played and I don’t think implementations in any other games is really that great – because if it was it would simply be copied.

Looking for Fortnite Double Pump help?

Players should not die because they happened to spawn next to a dude with a rocket launcher. They should not die because the spawn system can be gamed/exploited/skilled in order to make them spawn nearby a stacked player. They should not die to a % play of timing a shot on a nearby spawn point, unless that % is so small it is unimportant.

Basically players who are rejoining the game should not be sitting ducks and be obliterated instantly. Or so the proponents of spawn protection think.

Current spawn protection implementation

As of writing (11092014 and 21102014) the system works like this. Players spawn with 2500ms invulnerability (previously 3000ms), which is cancelled on attack. The player can move as far as they want in this time. They can run away from anyone trying to kill them. They can run and get a weapon.

And thats pretty much all. It is not a new take on spawn protection, it has been used in other titles in the past.

A question: I don’t know if things like locking down spawns in TDM or shooting close spawns (or known spawns if a system worked that way) in duel is actually good gameplay or adds much to the skillset. Yeah it takes player knowledge but is it good to have? Should it be a skill? Should timing a shot and shooting a spot on the ground after killing a player yield a % possibility to get a free kill?

The current implementation of spawn protections does not function well – I haven’t even played with it and know it is flawed. Battlefield 3 is a great example here. Spawn protection works in a similar way to ut4, the player is invulnerable (also does not take suppression) however protection is cancelled on attack or movement rather than just attack. It is also significantly shorter in BF3 due to the spawn in black screen delay. You can spawn then rotate view with mouse and aim down sights. Anything else cancels it – movement, medpack drops, weapon changes (I think), grenade throws. Given the short ttk and loadouts in BF3 this is quite powerful – as a spawner you have enough time to almost casually line up a shot and kill the player attacking you – optional for even better performance wait until the player shooting at you starts to reload/in between bursts in order to avoid suppression. Spawn protection did not ship with BF3 and was added later due to how respawns have a delay with black screen when popping into the map. Or it shipped but was much “weaker” initially. It was still quite short in comparison to 3 seconds.

Sounds good right?

However, and this is the problem, few people knew how it worked. Few knew the game had spawn protection or even used it. The vast majority of players simply spawned, +forward and were mowed down if they were on a contested point. They spawned on their teammates under fire (derp) and died due to pressing W. Players being unaware of the mechanic was good because if everyone used it properly spawn protection would have detracted rather heavily from the game. As it was not really ever used on pubs. Players died on spawn because they moved straight away.

The same will happen in ut4 duel and other game modes – the vast majority of players will shoot when they spawn if there is someone to shoot at and cancel spawn protection. If they do not cancel it will be because there is nothing to shoot at. Now the obvious argument here is that “well then spawn protection will not break the flow of the game as most players will exit invulnerability unwittingly anyway!”. Unfortunately its not that cut and dried. Most will cancel it and as a result they will not benefit from protection.

Thats fine – like the BF3 example it will not detract from the game for most players.

Yeah there are probably ways to make it more obvious, etc. But it is counterintuitive. Much like holding fire to wait for spawn protection to run out when a player spawns in front of you.

It will be abused by the minority that understand it, like BF3. Personally I feel that abuse of this type of mechanic is just as bad as spawn killing it is trying to fix. This is why discussion on the forum about spawn protection specifically is pointless. We should be discussing ideal spawn systems (plural for different game types) rather than trying to work out how to cludge invulnerability or some other nonsense in a way that is not too distasteful.

A Better Spawn System

Is the problem spawn protection or is the problem the spawn system?

For the sake of this article let us assume that the spawn system should not be influenced by other players. Players should not be able to force, or at least make it more likely that the player they just killed will spawn in front of them or at a known location. On top of this simply knowing spawn points and aiming at/shooting at should not yield kills reliably.

In short the goal is to have a spawn system that cannot be influenced by players and dead players spawning will have a very low chance to instantly die. After this all bets are off.

The first main consideration is stopping players being killed as soon as they spawn then there a number of easy fixes. Adding limited invulnerability is not the solution, at least in my opinion. It is somewhat jarring, breaks the flow as many other posters on the dev forum have pointed out and also allows the player too much freedom if they know what they are doing. It also does not really protect players as highlighted by the BF3 example. It is there but because players have not been educated they will just cancel it.

In short it is both useless and exploitable. If the goal is to alleviate spawn killing due to players exploiting that particular system, adding another that can also be exploited makes no sense.

Invulnerability can cause some other unintended side effects besides interrupting flow. If the spawnee plays correctly and does not attack back then they may be able to get a weapon without taking damage. Is this ok? Is this fair? This is akin to spawning them with a rocket launcher or shock rifle. It is negative feedback for killing players. You can argue that it is ok and they should be given a chance and I agree, but I do not think it should be at the detriment of the player who is trying to kill them.

I understand the argument from the pro-spawn protection crowd, and even agree with it to a point (thus this post), however we need to look at better solutions. Better spawn systems.

Whatever solution/fix is picked needs to be seamless and not negatively impact either player and most importantly not be gameable if the goal is also to remove this aspect, lets assume it is.

The current solution stops player being killed. This is all it does, and could barely be said to achieve even this. It disrupts the flow of the game. While many seem to be of the opinion that this does not matter to the pro-spawn protection group, my question is this: If there is a solution that does not have the same downsides why not use that instead?

Plus the current implementation has problems and is exploitable. If the goal is to stop newer players from getting killed as soon as they spawn tying invulnerability to +attack is the wrong way to go about it. Players who are not interested in learning how the spawn system works* currently will not make the connection between shooting and cancelling protection. They are not the type of players that will understand spawn protection. In this way they will always be killable because they will attack back upon spawn. On the flip side their spawn protection informed brethren will be able to do the inverse – not attack and run towards a gun, dragging out invulnerability as long as possible. The way of the world means the better players may be the ones to know the mechanic and the less regular, dare I say casual, players will not be aware of it.

So we end up with the current situation reversed – rather than the killer exploiting new spawning players the new spawners are exploiting protection to damage/kill other players. Imagine being the less savvy player, working hard to finally get a kill then the other player seemingly not dying to your fire? Omg lag, hax, whatever. Frustration.

It could be argued that this player should then go and look into WHY this occurred but we know that is unlikely to happen.

*no negative connotations intended, this is a fairly niche aspect.

And so we have a situation where instead of chain dying, the less keen players will die when they spawn (because they +attack) and when they manage to kill the better players they will have less opportunity to kill them again because the better player will be more likely to be aware of the mechanic. This might be a marginally better situation than chain death but it is not idea.

Fifteen years ago spawn protection in this manner was possibly acceptable, less players were looking at games at this level. Obscurity like this does not work anymore.

In duel this could be because the spawn system is setup and the other player can “force” the spawn to where they want. In TDM this might be due to locking down part of the map, covering spawns with some members of the team and killing respawning players that the rest of the team kill. In CTF it might be killing players who are spawning in your way out of the base with a flag. I don’t have the experience to comment on CTF.

A Better Spawn System before spawn protection

This is my opinion on what a spawn system should be. This sort of setup alleviates many of the problems associated with spawning and spawn killing in arena FPS, particularly in duel and ffa.

Firstly spawns should be random. 100% random. Any algorithm that spawns players based on X, Y or Z can be reverse engineered. Spawn farthest, or safest spawns or whatever can and will be worked out and abused. Also some systems promote not-killing players as it may give away items – it may also promote /kill in order to ninja in on a particular spawn. Obviously the latter can be addressed with a different spawn system for self kills but the point stands, if the goal is to remove spawn killing and manipulative aspects of spawning the only real solution is random. Removing mechanisms that makes the spawn system predictable/gameable in any way completely levels the playing field.

As we can see, a spawn in front of the player below is opened up by backing into a corner.

ztn abuse able spawn

So make it 100% random for duel. Make it 100% random for FFA and TDM. CTF is different and someone who has experience in that department should comment on it. I like the idea of spawning on teammates but I doubt that would go down well 😉

Even in a random system players could opt to upload rockets at the closest known spawn, giving an almost guaranteed conversion frag at (100/spawnpoints)% kill based on timing the shot, knowing where the spawn points are and simply shooting the closest/most appropriate.

This can be solved in a number of ways. Remove the “instant spawn” that ut99 (and from memory later titles) had and replace it with a delay the user can cancel to spawn instantly. In this way players can spawn up to 10* seconds after they die. Why is this important? Because it removes any chance of timing shock balls, rockets or anything on nearby spawn points. It puts the choice of when to pop in the spawners hands. Also make sure not to introduce a Xms delay between death and spawning that quakelive has – if the player wants to spawn instantly, let them, if they want to delay 5 seconds let them. Obviously cap this delay to something reasonable, the goal is to stop players timing shots on nearby spawn points.

Allowing delayed spawns may cause concerns with the existing ut commnity. ql has a similar mechanic and it has never been a problem. In fact it should be added regardless of random spawning – put the time to spawn in the player that is dead hands and run the spawn algorithm when they choose to respawn, again putting more distance between the killer and the information on the spawning player.

Secondly add enough spawn points to a map so a player in each area has more than one place to shoot. This way it becomes less likely to receive a point for simply shooting at a spot.

The alternative to adding lots of spawn points is changing spawn points to spawn volumes and randomly spawning players in them. This way there is no “point” to direct rockets or projectiles at. I like this idea over many spawn points as it allows infinetly more variation on spawn locations but it may be harder to implement.

Of course both more spawn points and spawn volumes require mappers to implement them, just like spawn points need to be placed in logical safe places and not in the open or at the end of corridors that allow easy killing of fresh spawns.

Finally remove respawn sound and visual cues. In this way even if a player spawns close to their death location they are “safe” from the other player provided they opt to proceed stealthily. In this way the killing player can still check for spawns but they “spend” time (the currency in duel) checking nearby LOS hidden spawn points rather than simply knowing the other player spawned there from audio.

Maps and spawn

The spawn system should not try to make badly designed maps work or make maps that are designed for four players work with 16+. Putting twice the recommended number of players on a map should never be a consideration of the game core rules/mechanics. Of course its not going to be ideal.

code187: What about arcane temple in 99 rocket launcher area facing out 2spawn spots spawn kill then they spawn 3meters to the left or right. And again and again and again monster killing… fun as hell for the camper not so much for the spawner***

Firstly this is exacerbated excessively by ut99s weaponstay system (unlimited ammo via dropweapon) and the way most weapon spawns had 2 ammo boxes next to them. Off the top of my head Peak also had this problem in the belt room as well with the spawn in the doorway. On arcanetemple the rocket pickup had 30 ammo including the packs and even if the packs were gone due to another player you could easily weapondrop -> pickup to refill ammo indefinitely.

Also it was also problematic due to 16player games on small maps. When players opt to play in this way all bets should be off.

abuse able arena fps spanws

Bad maps and spawn issues like this should not be a reason to implement an across the board cludge to spawning. The system should deal with this up to the number of players the map is designed for and if it breaks beyond that then that is reasonable. The spawn system should not be designed to deal with content creators having a weapon with infinite ammo + two spawns in front of it (there are actually four since two additional points are in the corridors nearby as well). These spots were terrible for spawns anyway even if the rocket was not where it was. The spawn system should not have to deal with a room with a single door and a spawn in the doorway – peak. This is beyond what the spawn system should be dealing with.

For problems like this look to the mappers and content creators not the base game ruleset for fixes. Look to the developers who allowed unlimited ammo and weird weapon pickup rules. This actually highlights a problem with ut – for all the posters on the dev forum rabbit on about mutators and customisation maps with problems were rarely changed or fixed.

Going back to the arcane temple example with 100% random spawning. There are 20+ spawns on the map. If we drop this to 10 (due to two spawns at the rl) the chance of spawning there twice in a completely random spawn system is very low (one in one hundred). From a “not spawning players multiple times in the same spot because they die instantly” standpoint simply having completely random spawns fixes the problem. Dying once every now and then is fine, dying repeatedly is not – but the likelyhood of it happening two spawns in a row in the same place is quite low, even with only 10 spawn points. Add more and it would occur so rarely that it is not a problem.

The system could also be setup to not spawn players at the same point twice in a row.

Add spawn volumes and the rocket launcher camper on arcane temple can no longer spam at two tightly defined points. But beyond this the problem goes back to the map design. Arcane has a dead end with a pool of water in it. With spawn points. The setup is really asking for it.

Extending spawn ideas

Some random things.

In an attempt to make duel spawning “fair”, remember close/far spawn order then give the opposite player similar spawns when they kill the other player. In this way streaks can be given to the comeback player and spawning close/far. Obviously this steps away from the whole.. “random” aspect I have outlined previously but it could be worked in.

The ability to define some spawns as “initial” spawns that are used for the start of the game. In this way a mapper has more control over what players have access to at the start of a match. This means the “delayed item spawn” that ut features can be removed. Players are given similar powered stacks by the map designer. Not all spawns are suitable for first spawns due to item locations.

[QUOTE=conX5;127043]Armors should certainly be delayed off initial spawn, but I don’t think anything else should really be necessary.[/QUOTE]

Realistically armors/mega should not be delayed. The “no armor” start of the game simply skews it in favor of weapons that are more efficient with lower health pools or weapons that are more suitable for the upcoming armor spawns, etc. The problem is still there, just shifted. Pretty sure this point has been discussed to death over the years.

A better solution to this is giving mappers the ability to define spawns as initial starting points. These are used for the first spawn of the game. Spawn points not flagged as first spawn cannot be spawned on at the beginning of a match.

More spawn points may make sense as the game progresses but disallowing some during the start of the game should cut down on dm6 pillar spawn type problems.

This is how quakelive handles things and from the current pool of popular maps there are only a handful of “bad” starts. YA on aero for example – but even this is good compared to some examples mentioned around here for ut. Others like dm6 pillars have been ironed out. Sure there are still better/worse spawns but not quite as bad as getting belt/boots/jacket.

Another thing that must be done is update maps. Maps are not “finished” when they are released. ZTN for example has had [URL=””]at least three[/URL] spawn setups in recent memory. To my knowledge it was never updated in q3. That particular post is concerned with all spawns in general but the same can be applied to initial spawns easily enough. Basically the mapper cannot be expected to “see” every possible out come and problem that can arise – but they should be able to get it partially right and fix problems that do come about after lots of play time.

That is one way to approach it. Disallow spawns that make no sense or are wildly imbalanced. There is no way to have “perfect” split system because in ut stacking/armor/health does not really lend itself to both players being even – one is always weaker. This “initial spawn” system is validated reasonably well by quakelive, provided maps are tweaked after being played on enough.

However that might not be far enough? Players in ql still whine about some spawns and while the above cuts down problems heavily it does not remove it 100%. As an example high lower YA vs plasma on ZTN. This is more messy than anything else and turns the opening into a large dice roll (based on decisions) more than other spawns.

Solutions? Have “pairs” of spawns setup in a way to give both players some pickups. Create 3-5 “pairs” per map and you have a solid set of not-unbalanced* starts. The previously suggested half time is not needed.

*balancing this perfectly is not the goal. Aiming to remove 100/belt/mega vs nothing setups is the intended outcome.

Sounds like a good solution right? Give players somewhat even splits and remove 100%stacked vs unstacked occurrences via forcing pairs of spawns. The unfortunate problem is this breaks part of the early game that is overlooked and is quite important. By having pairs of spawns players simply know the opening and this becomes very fixed. Some might be considered a good thing but personally I love the dynamic nature of duel and the decision making process at the start is interesting.

If you have two spawns that are ALWAYS the same spawn A will have an optimum opening and a sub-optimal (but probably more aggressive) opening. The sub-optimal, aggressive option becomes less viable when the opposing spawn is fixed as they will know where the aggression will come from.

There are many ways to go about fixing things that are not as overboard as delaying initial item spawns or setting two spawn points in stone.


One main difference in quake compared to ut is these problems are very specific (in ql) and players quite happy to point out the problem on the maps in question. In ut, at least on these forums so far, players tend to post very broad comments about opening spawns. eg “You can get all the armors on deck” or similar. Being specific when discussing this is key. Even if no changes to the initial spawn system are made can spawns XYZ on deck be dropped or moved and a better map result?

It is important to not confuse things. A player making a bad call and not getting items is fine. When there are situations where one player can pillage all important pickups before the other can even show up needs fixing.

No responses yet

Leave a Reply