Articles, Blog

PUBG vs Fortnite Netcode Updates


Hi my name is Chris and this is Battle(non)sense Since my last test of Player Unknowns Battlegrounds,
the developers did not only try to improve the performance of the game client and the
server by i.e. preventing players from seeing the inside of the airplane or reducing the
draw distance while in air. They also introduced ping based matchmaking,
which means that you should now mostly play with or against players who have about the
same ping to the server as you have. Which should result in a much more enjoyable online
experience. However, you must keep in mind that a player
might have a ping of less than 30ms when the match starts, and then later his ping spikes
to more than 300ms. Which will then again cause issues like receiving damage far behind cover as you can see here. But when you die far behind cover then that
is not necessarily the fault of the player who shot you. It could also be your connection
to the server that suddenly got worse. Sadly, PUBG still does not tell you anything about
your connection to the game server, like your ping, or that you have packet loss. An issue that can cause extreme ping spikes and packet loss in any online multiplayer
game is buffer bloat. In the description down below, you can find a link to a video where I explain what that is and how you can fix
or prevent it. Now if you haven’t seen one of my other
netcode analysis videos yet, then you should watch my Netcode 101 video where I explain some basic things about 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. The card overlay, as well as the link in the
description down below will both take you directly to that video. So, while PUBG still does not show any kind
of network info, Fortnite provides Network Stats for quite some time now, which you can enable here. However please be aware that this option only shows up in the menu after you joined a match, and you must enable the network overlay every time you join a new one. I really hope that the developers will change this in the future. Before the 3.1.0 patch you might have noticed
that the ping inside the network overlay, was much higher than the ping displayed for
the selected region. This was caused by a previous update, which made the ping value include the processing delays of the game. This change then caused
some confusion as the overlay suddenly showed a ping of 80 to 100ms, while previously it
showed a ping of around 25ms. As a result, players suspected that the matchmaker was
broken, and that they got placed on servers outside their selected region. This was changed again in patch 3.1.0, However as you can see here the ping still shows way too high values in the early stages of the match, and towards the end of the match the
ping inside the overlay even gets lower than the ICMP ping to the datacentre which doesn’t make any sense at all. So, you should not trust this value too much right now. In my last video I showed you how PUBG’s
servers struggled quite a lot throughout the match, while Fortnite’s servers managed
to reach a farily stable 20Hz later in the match. That said, some people misunderstood what
I said about that 20Hz limit in my last video. The Unreal Engine is not limited to 20Hz. It was the decision of Fortnite’s development team to limit the server to 20Hz, as their goal is to achieve a stable 20Hz for the entire match before they try to increase the rate further. The PUBG developers set the limit to 30Hz But the Unreal Engine itself is not limit to 20Hz nor 30Hz. The 100 players are what is causing the performance issues for both PUBG and Fortnite, while games like Battalion
1944 can run at a stable 60Hz on the same engine. So how did these performance optimizations
affect the PUBG servers, and more importantly, how does the player count affect the servers of both Fortnite and PUBG. This image here shows how many updates per
second I received from the PUBG server during a match. When the first player jumped out of the aircraft the server’s send rate was at 23Hz, which then dropped down to 8Hz when the last player
left the aircraft. Then as more and more players got killed,
the servers send rate started to increase, until it hit its maximum send rate of 30Hz
when there were only 42 players left. So, while the servers performance is better
than 2 months ago, there are still some quite big drops in the servers send rate throughout the match, which have a negative impact on the delay. When we then take a look at the results of
two other matches, then we see the same drops in the servers send rate, and that the server
performance is affect by the player count
in the same way as it only reaches its maximum
send rate of 30Hz when there are less than 45 players left. However sometimes it takes
a bit longer until that many players died, which might create the impression that the
servers performance takes longer to improve, while its directly linked to the amount of
players who are still alive. The situation is quite similar in Fortnite. The server sends 20 updates per second when
the first player jumps out of the bus, and the send rate then drops down to 10Hz once
all 100 players left the bus. Then as the player count drops, which happens a lot faster in Fortnite due to the design of this game, the servers send rate starts to increase until it reaches its maximum send rate of 20Hz when there are less than 50 players left. What you will notice here is that while the
servers send rate is lower in Fortnite than in PUBG, it is also a lot more stable as it
does not suffer from those massive drops. When we then compare the results of Fortnite
and PUBG, then we can again see how the player count affects the server’s performance, as both games only reach their maximum send rate when there are less than 50 players left. That said it seems like PUBG manages to achieve higher server send rates at the same player count, However the PUBG servers also suffer from quite big drops in the servers send rate throughout the match. So, the question is how that affects the actual network delay between 2 players. Before I show you the results of my network
delay tests I want to ask you once again to watch Netcode 101 video first to find out
how I do these network delay tests, as that is required to understand what these results
mean for your online experience. 2 months ago Fortnite showed much lower network
delays than PUBG, which suffered from some pretty bad server performance issues at that time. Now, after the recent optimizations and the
introduction of the ping-based matchmaker, the network delay in PUBG got a lot better, however when you look at the longest delays, then you can see what kind of impact those
drops in the servers send rate have on the player to player delay. Fortnite did not really get any major network
optimizations and so the results are very similar to those from my last test. Sadly the developers decided to disable friendly fire and so I can no longer do the damage
delay test, as it’s simply impossible to get my 2 test players onto the same server when they are not in a group. So far, I always showed you results from very
early in the match where you have the lowest server send rate and thus the longest network delay. But today I also want to show you the delay
that you get when the game servers operate at their maximum send rates. So, the best
case scenario if you will. When PUBG’s game server sends 30 updates per second then this reduces the average network delay by about 12ms, however you can still see the impact that these drops in the servers send rate have on the player to player delay. When there are less than 40 players left in
Fortnite, then the average delay does not only decrease by 10ms, the network delay is also lower than in PUBG. Which is quite surprising as its game servers only run at 20Hz, while PUBG’s run at 30Hz. This shows us once again that higher update rates do not necessarily result in lower network delays. And that the EPIC still has a lot
of work to do until the Unreal Engine can maintain 20 or 30Hz in games with more than
50 players. So that’s all that I have for you today. I will keep you updated on the progress that the developers make, and if you enjoyed this updated netcode analysis of Fortnite and PUBG, then it would be great if you could support
me on patreon as YouTubes ad revenue is sadly not enough anymore to run a niche channel like mine. Without the awesome support that I get from my patrons, Battle(non)sense would
not exists anymore. You can find a link to my patreon in the description down below, where you will also find links to my social accounts in case that you want
to stay up to date on the videos that I am working on. So, if you enjoyed this video then please
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 Battle(non)sense.

100 thoughts on “PUBG vs Fortnite Netcode Updates

  1. The pubg movement delay and the delay in when you Open a door is freaking me out i Hope they will fix this sooner than later

  2. Could you do this on a game called APB:Reloaded? The game is not big / known but I would really love to see it.

    Good video btw 🙂

  3. im saying it since all this bullshit began with the desync. pubg should release smaller maps with 50 players.

    less desync
    less delay
    less boring midgame

  4. So annoying that the update rate is so bad at the start of the round in Pubg, makes going to highly populated areas stressful

  5. This vidoe just shows how gamers standards have dropped. 20hz servers for 100 players, are you joking. no wonder they give it away. yet I bet everyone who plays it has spent atleast £10 – £20 for it. so really its not free. Gaming has become a joke now as it caters for shitty connections cheap servers. no wonder gaming is doing the whole loot box thng.

  6. Are you planing on doing an update for Quake Champions Netcode? It's been 6 months now, and it feels like the game is way more consistent now

  7. +Battle(non)sense
    Mavericks: Proving Grounds is around the corner, more info on GDC 2018.
    Doing netcode annalistic video for that game will be next level.

  8. Hi Chris! Could you possibly do a video on the different models that exist for hit detection (fully server side, fully client side, client side with server validation and so on, including explanations for terms such as backwards reconciliation, lag compensation and so on), similarly to your netcode 101? I think that would be quite insightful for a lot of people since comprehensive articles on this topic are very scarce.

  9. Since they do not allow friendly fire for your more comprehensive tests, perhaps look into doing "decal latency tests" instead, where you shoot a wall the other player is looking at and then measure how long it takes for the bullet hole decals to appear on the wall for the other player. Just a thought. Perhaps this would not be a very effective replacement for actual damage, as bullet hole decals may take a lower priority than player hits and damage, I dunno.

  10. Can you think of a reason why the ping on my PS4 is always double than it is on my Xbox One? They are both wired hooked up the same way yet my PS4 is always over 80ms but my Xbox One is always under 50ms. Thanks man. I really appreciate these videos. 👍🏼✌🏼️

  11. Keep in mind that ICMP RTT can and mostly will differ in comparison to TCP/UDP RTT. ICMP echo(8) is deprioritized by a lot of transient routers on the internet.

  12. Tic rate variability is why I jump out last and float in the sky into the middle of the storm while waiting for other players to die.

  13. i have 100-120 fps, but PUBG is almost unplayable for me after last update…it feels like playing in some jelly / liquid, hit registration is absolute gamble mess sometimes in my favor, sometimes in opponent's. Turning v-sync on kinda makes it better that my eyes dont hurt, but feels even more unresponsive…

  14. I'm actually quite impressed how much pubg netcode has improved over the past several months. Of course it's still bad but it is heading into the right direction.

  15. You said "the developers have introduced ping-based matchmaking", but as far as I've read, it was in testing phase (feb 20)? Or have you done your tests on the test server?

  16. Is there a small tutorial(written would be fine too) on how to capture the packets of xyz.exe and calculate the tickrate? I would really love to be able to do that for any game, nothing detailed or super professional like you, just a "short" test of one game or half an hour of DMing. 🙂
    Cheers

  17. Hello there, cool videos, really like them. Still there is something I'd like to share with you: At 3:09 you show a graph of the update rate of a Fortnite server, connecting the data points linearly. All the values are integers, so the representation you chose is not ideal. A way to fix this would be to use stepfunctions with the datapoints you recorded as the respective step hights or even ditch the graphs altogether. I would go for the stepfunctions though, because they are easier to catch for the eye than "loose" points. Apart from that: How do you do your plots? If you didn't "automatise" the process yet, I could offer you some sort of automatisation for plotting from textfiles. Cheers!

  18. you can't just unfriend the guy and then just click "Play" at the exact same time and meet up at an obscure part of the map to get the Damage times for fornite?

  19. Anyone who hasn't subbed give this guy a hand, this channel supplies very useful info, totally deserves more subs !

  20. Sad that they can't make the game run with 60hz servers.
    Would be amazing.
    And if they only could improve the responsiveness of the movement, that would be great.
    Always having that jelly-movement, it's super annoying and gets me killed a lot of times.

  21. they had fixed it. CLEARLY! it ran very smoothly and i had no issues for 2-3 weeks. Now everything is back to same. I cant pick up items. I get rubberbanded out of the building. Superbullets ect…. that means after they optimized it they upped the VM's per server to lower the cost. So we are at the same place we started.

  22. I love how this comment section is pretty much fill with Fortnite fanboys… what they are saying is true, but seriously… how old are these guys? 12?… if u don't like it say that you don't and done. Not fuking "this game is trash and i'm glad that it's dying…."

  23. Both games going in the right direction, pubg seems way more stable since last test but it still has a way to go. the day both games can run 60 hz servers through an entire match now that would just be the cream on top of already awesome games

  24. To be honest, my experience with PUBG has only improved, i dont know stats but for me it has improved.

  25. At 6:36 you explain how hard it is to get 2 players into the same game, however I have tried starting at the same time with my friend and we get into about 3/5 games. You can check whether you are in the same game by looking at the last few digits in the long string in the top left.

  26. I can't wait for some billion game company pick this BR mode from ashed and make a masterpiece. Still playing Pubg but the frustration is boiling my blood from stupid delays, desync & lags.

  27. Reminds me the COD age when shitty games like PUBG get popular. Just wait until a real studio make a good BR with a real netcode and good servers and we won't hear those pigs defend their shit game anymore LUL.

  28. Netcode has fucking terrible netcode. I have many many clips of shooting people in the heads and spraying corners with semi, accurate, and fast fire and they still kill me instantly

  29. I have a edgerouter and i can't hear anyone voice chatting in pubg,fortnite,and rainbow six siege. no one can hear me and i cant hear them and i can't see the voice chat icon. So far i tried doing CTRL+T and changed the voice channel and also disabled the "alg sip" on the edgerouter. Any ideas?

  30. I get very bad statturing in BF1 ,
    I play at max settings @ 1080p 120fps lock , vsync off , my monitor doesn't have g sync , pc specs ( i7 7700k overclocked at 4.5ghz @ 1.2v , G.skill 16gb overclocked at 3000mhz , gtx 1080ti , msi z270 m5 , hyperx ssd 120gb , Samsung evo 850 250 ssd , firecuda 2tb sshd, Why I keep getting statturing and sudden fps drops, please help me.

  31. Can you please make an another video on Escape from Tarkov about the netcode, considering they've made a first update about that recently. Thank you for your work.

  32. I remember BF4, it has and have if the server doesn't change, has only 13hz… for 16 players… 🙁
    You could be 10m from a wall and still be killed…

  33. Fortnite players: I like fortnite because it’s fun
    Pub g players: Now I’ll tell you why you’re wrong…

  34. one of the hottest point of match 20guys land in Pocinki or Pecado and you try! 8hz TRY! 8hz aaaand you're ded ;___;

  35. Could you test the netcode of Ironsight?
    The game got a lot of attention recently and I think that the netcode needs some work.
    It would be nice if could test it in order to know if thats actually the case.

  36. Do aprecciate all your hard work man , but can you explain why in Planetside2 we have 300 players facing each other with tanks and airplains and other shcabamms and my shots allways hit the mark with no lag ? they can do 1000players online and this servers cannot handle 100 players ? why ?

  37. Hey, perhaps you could do a netcode analysis for Sea of Thieves? I get so bad ping in this game while getting good pings in quake or so.

  38. I wonder how DICE managed to get 64 player servers running on a stable 60-120 Hz. I hope BlueHole finds out some day.

  39. Hi! great vids and great work!! When You gonna do new PUBG update 7 test ? For me its worse then ever :/
    Any advice how to test the stability of mine connection? Download and upload is fine but Im wondering about stability.( ping, packet loss etc.) Take care ^^

  40. Heya, I don't know if you remember but you said you'd be doing a netcode analysis on Escape from Tarkov. They're releasing a patch today that will be working on Server Optimizations. Please give it a check. Thanks.

  41. So it's getting better on PUBG, but still, it's not really good. I would have loved to see videos like that at the realy state of the game because I really feel that since the 1.0, the game is getting worse and they're just trying to prevent it for going worser. I love this game, but I can't play it anymore unless they're doing something to it.

  42. Could you do this for the rates on Xbox One for both games? That would be amazing to see how far they are coming with both games for us plebs

  43. The movement delay in pubg literally makes me rage quit the game. It basically means that I have no time to react to enemy peeking out of a corner in first person, even when I was looking at the same exact corner he was peeking at.

  44. Hey! Could you make a video about PUBG pump shotgun input lag delay. It feels the slowest gun to fire when ever you use it for some reason. It makes the gun hard to use since you have to keep aiming at the guy for a good half a second or so before it fires.

  45. I think it would be really interesting to see a comparison of PUBG and Fortnite with their recently released mobile clients.

  46. Myself and a friend did some testing of pubg ping using this site: https://pubg.disquse.ru/ It clearly shows huge variation in ping, from 30ms up to as high as 1100ms with 300ms being a regular event..I really find it insulting that Bluehole keep releasing their emotes and skins patches when the game play is so frustratingly bad even a year later…… Ive given up hope on pubg becoming a game I can take seriously by now…

Leave a Reply

Your email address will not be published. Required fields are marked *