PUBG vs Fortnite vs H1Z1 Netcode Analysis
Hi my name is chris and this is Battlenonsense
Today I’ve got a very special Battle Royale
Netcode video for you where you won’t just
learn a bit more about PUBG, Fortnite and
H1Z1 formely know as King of The Kill, but
you will also find out which of these currently
provides the lowest network delay.
Now while always do my best to keep the netcode
analysis as simple as possible, you will still
need to have some basic knowledge about computer
networking, tick rates, update rates, network
models, super bullets, lag compensation, packet
loss and a few other things like how I do
my network delay tests.
So to keep the netcode analysis videos as
short as possible, I have put all this information
into a separate “Netcode 101” video which
you should have watched at least once before
you continue with this video here.
The card overlay in the top right corner of
this video as well as the link in the description
down below will both take you directly to
So lets talk about PUBG Fortnite and H1Z1
PUBG and Fortnite use the Unreal Engine 4
from Epic games, which is also the development
studio behind Fortnite.
H1Z1 on the other hand uses the ForgeLight
engine, which has been developed for MMO
and very high player counts.
It powers games like Plaentside 2, Payday
The Heist, DC Universe Online and Everquest.
When we talk about the player 2 player delay
in an online multiplayer game then we also
have to talk about the frame rate.
PUBG has an FPS cap of 144 which the player
can’t remove, while Fortnite and H1Z1 do
not cap the frame rate.
If you want to limit your frame rate at a
specific value so that you always have the
same amount if input lag, or to keep the frame
rate inside the G-Sync or FreeSync window
of your monitor, then PUBG sadly doesn’t
provide you with a built in FPS limiter.
Fortnite has such a feature, but you can only
choose between 3 presents, you can’t choose
a custom FPS limit like you can in Overwatch.
In H1Z1 you can choose a custom frame rate
limit, but you have to edit a config file
Now lets get to the networking of these games.
There have been rumors that PUBG would switch
to Microsofts cloud service Azure, and while
it’s possible that the game uses Azure on
the Xbox, the PC version of PUBG always connects
me to a server hosted by AWS.
Also I am not aware of a statement from the
developers where they announced that they
will switch to Azure.
If you have a link to such a statement from
the developers then please post the link in
the description down below.
Fortnite does not only use the same Engine
as PUBG they also use AWS to host the gameservers.
In H1Z1 you will primarily end up on Daybreak’s
But should it become necessary then they can
also utilise amazons cloud service AWS to
spawn additional game server instances there.
Now in which regions do these games provide
PUBG currently provides servers in more locations
than Fortnite or H1Z1.
However Epic Games announced that they will
add servers in more locations early this year.
Even though PUBG does seemingly have better
server coverage than the other 2 games, it
must be said that players could benefit from
PUBG utilizing additional server providers
to increase their coverage.
Another important point is that Unlike PUBG
and Fortnite, H1Z1 does not allow players
to connect to a region to which they have
a ping of more than 200ms.
It does that to mitigate the negative impact
that players with a lot of latency have on
the experience of other players who have a
low ping to the gameserver.
More on that a little bit later.
When it comes to the network delay, then the
rate at which the game sends and receives
data have a very strong impact on the players
So I used Wireshark to capture my network
traffic during multiple games so that I could
then find out more about these 2 update rates.
What I found was that in pubg the minmum send
rate of the server is 2Hz, the average is
18Hz and the maximum is 31Hz, in Fornite it’s
7Hz, 19Hz and 30Hz, and in H1Z1 it’s 16Hz
76Hz and 148Hz for the servers send rate.
So what does this tell us.
When you just look at the average rates then
the H1Z1 server clearly has the highest send
rate and so the highest tick rate as that
data has to come from somewhere.
If you then compare Fortnite and PUBG then
you could think that Fortnite beats PUBG by
But while you are correct that Fornite’s
server performce beats PUBG’s, its not because
of that 1Hz difference.
To understand why we must take a look at the
network statistic to see the real difference.
This is the send rate of the H1Z1 server which
stays well above 50Hz most of the time.
Then we’ve got the Fortnite server, which
starts at around 8Hz, then goes up to 10Hz
and steadily increases it’s send rate as
the match progresses.
There is not a single significant drop in
the servers send rate.
Now lets have a look at PUBG.
Just like in Fortnite, the PUBG server also
sends about 8 updates per second when the
match starts, and then increases its sendrate
as more and more players die.
However it’s send rate has significant drops
throughout the match, and if you have a gunfight
while such a drop occurs then you will be
affected by a lot more delay – which is
one of the reasons why you might get shot
around a wall.
So generally speaking, the fewer players are
on the server, the better the servers performance
and the higher its send rate.
However in PUBG I also noticed that sometimes
the servers send rate would suddenly start
to decrease later in the match which also
increases the network delay.
And this is the case on both the old and the
So when we compare this to Fortnite, then
not only does its server performance steadily
increase as the match progresses, its send
rate also does not suffer from these massive
You will also notice that despite an occasional
spike, the Fortnite server peaks at a send
rate of 20Hz as the developers limit it to
that rate, while in PUBG they allow the server
to run as fast as it can which might even
be why we see those send rate drops as that
could indicate a stall of the server.
Epic games said in one of their Fortnite dev
blogs that their goal is to achieve a stable
Which means that the developers of the Unreal
Engine have set 20Hz as their goal for Fortnites
Battle Royale mode.
So with that in mind it seems logical that
the Unreal Engine is indeed a limiting factor
here for both Fortnite and PUBG, as it simply
wasn’t developed for large playercounts
– unlike the Forgelight engine used by H1Z1.
So I think that the PUBG developers can’t
really do all that much about the low update
and tick rate of the server as it seems to
be the engine that is holding them back.
But they could surly do something about these
drops as these do not occur in Fortnight.
Besides the send and tick rate of the server,
the update rate of the client is also important.
Here we see that all 3 games use quite similar
rates – but what is interesting is that the
H1Z1 client sends fewer updates than it receives
from the server, which is a bit unusual.
Now, how can a player find out if the issue
that he experiences in the game is a result
of a problem with his internet connection
or maybe a performance problem of the server.
PUBG does not provide the player with any
kind of network information.
While in Fortnite you can enable a Network
Overlay which shows the ping to the gameserver.
How much up and downstream bandwidth the game
It even shows you the send rate of the server
and the client
As well as inbound and outbound packet loss.
So Epic games provides the player with a lot
of very usful information here.
However I am not quite sure if that ping here
is actually the round trip time between my
client and the server as it is about half
of what I get when I ping the AWS datacentre.
And when I artificially increase my ping by
200ms, then this value here only increases
So you have to that that ping value with a
grain of salt.
In H1Z1 you see the ping to the server in
the top right corner, where the developers
even use colors to tell you if your ping is
good or bad.
In addition to that you can use the N key
on your keyboard to enable the network overlay.
Inside the overlay we can see how much bandwidth
the gameclient uses.
Ping 1 tells you the round trip time of the
UDP game data between the client and the match
You will also notice that it is lower than
the ping 2 value as that also includes the
time the match server needs to process a data
So should Ping 2 be significantly higher than
Ping 1, then this could indicate a performance
issue on the server.
And lastly the overlay shows you inbound and
outbound packet loss – however in my tests
I noticed that these do not update very frequently,
which is why they might not pick up occasional
I have forwarded that information to the developers.
Now lets check out the results of my network
delay tests, where I measured the delay between
2 players who had a ping of 25ms to the gameserver.
To get conclusive results, I tested PUBG,
Fortnite and H1Z1, 5 times per day on 4 different
I also made sure that I finished each test
run during the first 5 minutes of the match
so that each game was tested when the server
was under the most stress.
So lets start with H1Z1 which provided very
consistent results throughout all testruns.
In the damage test I measured an average delay
of 42.75ms, in the gunfire tests I measured
an average delay of 61ms and in the movement
test I measured an average delay of 77.75ms.
So the gunfire and animation delays are quite
a bit longer than the damage delay, which
we have seen in other games as well.
However I did reach out to the developers
already and maybe they can bring those delays
down a bit more in future updates.
Then we’ve got Fornite which also provided
very consistent results in all testruns, but
due to its lower update rates suffers from
much longer delays.
Which are even a bit longer than in my previous
Then we’ve pot PUBG.
Sadly the only consistent thing about its
test results is that they are very inconsistent,
which is not really a surprise when you remember
that the servers send rate is all over the
place, no matter on which map you play.
So the results that I will show you now are
from a test run that represents the most common
results that I got.
It’s not the best run nor the worst.
So at a ping of 25ms to the gameserver players
can expect an average damage delay of 173.3ms,
an average gunfire delay of 172ms and an average
movement delay of 138ms
These are by far the worst results that I’ve
measured in PUBG since I tested it for the
first time back in march 2017.
So while the developers managed to increase
the frame rate on the client and added many
very useful game mechanics, the network performance
actually got worse.
We have reached a point where you will see
a network delay of more than 1 second while
such a drop in the servers send rate occurs.
So as I’ve said before, the Unreal Engine
is most likely a limiting factor here as not
even Epic Games can currently get Fortnite
to run at 20Hz from the round start to the
round end, but that does not change that the
Fornite servers runs a lot smoother.
Neither does it change the fact that PUBGs
network delay continues to get worse.
In the longrun the PUBG developers probably
don’t have a lot of options here.
They can either start to rework parts of the
Unreal Engine on their own, or wait for EPIC
to make their engine fit for games which put
more than 50 players on a server.
Short term the PUBG developers must work on
those drops in the servers send rate ,as these
seem to cause that massive delay increase.
They might also want to consider to limit
the servers tick and update rate to a value
that it can maintain, like EPIC currently
caps the Fortnite servers send rate at 20Hz.
I also hope that PUBG will eventually include
a network overlay similar to what we see in
And it would be great to get an unlocked frame
rate as well as a built in FPS limiter where
players can set a custom FPS limit, like we
can in Overwatch.
Now there are 2 more things left that I would
like to show you, first is how much traffic
these games generate.
As you can see here, PUBG, Fortnite and H1Z1
create surprisingly little traffic.
But please keep in mind that this is not a
ranking nor a rating.
And what kind of netcode analysis would this
be if we didn’t check out some lag compensation
As I mentioned before, H1Z1 does not allow
you to connect to a region when you ping to
that region is higher than 200ms.
However should a player suffer from network
congestion or buffer bloat, then this could
increase his ping to more than 300ms at any
point during the match.
Should this happen then the H1Z1 server will
still confirm that hit
and so a player with a ping of 25ms to the
server, will receive that shot far behind
In the description down below you can find
a link to a video where I tell you more about
buffer bloat and what you can do about it.
Fortnite and PUBG do not have a ping lock
yet which prevents you from joining a region.
But even if they had, the server will still
confirm a hit even when the shooter has a
ping of more than 200ms.
Non of these games seem to have a limit for
how much latency they compensate, which is
one of the reasons why you will receive damage
far behind cover.
So I hope that this netcode analysis of PUBG,
Fortnite and H1Z1 helps you go get a better
understanding of what is going on in these
games, what the problems are and what needs
I will keep and eye on the progress that the
developers make and release update videos
as these changes become available.
In you enjoy these netcode analysis videos,
then it would be great if you could support
me on patreon.
It’s because the awesome support that I
get from my patrons that I can continue to
bring you that kind of content.
So if you like my work then you can find a
link to my patreon in the description down
below, where you can also do a one time donation.
If you want to stay up to date on what I am
currently working on then you can follow me
on twitter or facebook, the links are also
in the description of this video.
And if you don’t want to miss the next one
then you might want to subscribe to my channel
and click on that bell icon below this video
to receive a notification when I upload the
So if you enjoyed this video then give it
a like, subscribe for more and I hope to see
you next time
until then, have a nice day and take care,
my name is chris and this was battlenonsense