Support NeoGAF

[Negotiator]

Everyone keeps suggesting that a higher tickrate means a smoother and more accurate response. If I'm watching a player nearly right next to me suddenly teleport further down the track because they just used a mushroom, while I'm still experiencing no lag nor frame drops, is that not due to a low tickrate?

Teleporting has more to do with packet loss.

When your connection is losing packets (small chunks of data), the netcode doesn't know where you are in the game world.

Let's say that I keep moving my character for 10m and while I'm moving (it seems fine on my screen, frozen on others), my connection's upload is congested and therefore I'm not uploading data for a certain amount of time.

When I finally send the packets that report where I am, the game will have to render my current X, Y, Z position and other players will see that I teleported 10m away. Interpolation can only do so much!

That's why prolonged packet loss is penalized in several games (Uncharted white plug, Splatoon etc.) It's borderline cheating and that's why no one should play on cellular (3G/4G) connections (too much congestion).

 

[LCGeek]

Sure. In an ideal world, everybody would be hunting down and eliminating all sources of lag, as they sat down in front of their gaming monitors on their wired networks, killing all background processes on their PC, making sure that they had the latest driver updates installed for their new GPUs.

In the world that we live in, though, you have to make games work on what people have. And someone playing on a Switch tethered to their phone while they ride the A train to work is not necessarily going to benefit from a higher tick rate. Similarly, someone sitting at home on a nice clean wired network is not necessarily going to lose a match solely because there was a disagreement between machines about whether they jumped behind a wall in time or not.

That Nintendo, though ...

You don't know you don't need special hardware for anti bufferbloat tech. Most networking tech especially routers already use linux and using a better version a linux kernel along with enabling certain features and disabling offloading has been a huge help. Problem is most don't know to do it. This specific subject has taught me our own apathy, laziness and ignorance can create more problems in resolving things at times vs any technical issues.

I've already said it nintendo could have a better wireless driver, better engine, and better hosting yet they don't fix any of the issues related to sub optimal online experience in splatoon or any of their online platform games.. There's nothing idealized about that it's simply put them being lazy and knowing they can get away with it.

 

[Burnburn]

This is incorrect. The horrible eldtritch truth lurking below online video games is that everyone's television displays a slightly different version of events.

You understood me incorrectly. Of course I know that what your opponent perceives will be different from what you see. What I was saying is that if you were to set-up your own switch to 2 different TVs, your own system directly hooked into a high latency and low latency tv you would see the exact same thing on both screens. The only difference is that one TV shows it later than the other. What kind of TV someone is using has 0 effect on the things that your console is sending to a server and then receiving again

 

[Falchion]

Holy shit, that's pretty awful.

 

[Patch13]

You understood me incorrectly.

I did. Apologies.

Of course I know that what your opponent perceives will be different from what you see. What I was saying is that if you were to set-up your own switch to 2 different TVs, your own system directly hooked into a high latency and low latency tv you would see the exact same thing on both screens. The only difference is that one TV shows it later than the other. What kind of TV someone is using has 0 effect on the things that your console is sending to a server and then receiving again

This is true. Latency can affect your ability to react to a situation, however. If you roll out of sniper fire "just in time" and crouch behind a wall, then find that you have died anyway ... well, you may have noticed due to tick rate, but the latency in your TV might have been a more important factor in slowing your reaction times down enough so that you got hit.

 

[UncleSporky]

Maybe this would be a better way to illustrate...correct me if I'm wrong, but essentially this means that while Splatoon visually looks like the 60 FPS on the right, network-wise it's running like the 15 FPS on the left:

Note that this doesn't mean your actual game state as represented in memory on your console is running like the image on the left. It's not a 16 FPS engine that just looks smoother. Your own character and any actions you take, your shots, your bombs etc. are probably all running at 60 FPS on your end of things. And there is plenty of interpolation going on too, even though you only get an update as to your enemy's exact position every 4 frames, it's very possible to have a good idea of where they will be in each of those intermediate frames.

This is just one way to visualize it for people who keep thinking it means the game is running slower, or that it causes situations like the lag snipe in the OP.

 

[Maximo]

The console has also been on the market for only five months almost to the day. Assuming they manage to launch it on January 1st, which likely won't happen, we've got at least five months remaining until the online service launches. That's plenty of time for Nintendo to tighten up the eShop and other services.

Come on the system has come out people shouldn't have to wait and see *if* Nintendo fixes their online services they already should have.

 

[LordOcidax]

Tickrate cannot contribute to lag? Or latency?

Actually no, the tickrate is the same for everybody... Around ~63ms (coorect me if i am wrong) refresh rate from the server. Assuming that everybody have a perfect pin and 1ms monitor (And a bunch of others factors), nobody is going to lag because the tickrate is low, so you are not going to lag compared to others players.

 

[Negotiator]

Maybe this would be a better way to illustrate...correct me if I'm wrong, but essentially this means that while Splatoon visually looks like the 60 FPS on the right, network-wise it's running like the 15 FPS on the left:

Note that this doesn't mean your actual game state as represented in memory on your console is running like the image on the left. It's not a 16 FPS engine that just looks smoother. Your own character and any actions you take, your shots, your bombs etc. are probably all running at 60 FPS on your end of things. And there is plenty of interpolation going on too, even though you only get an update as to your enemy's exact position every 4 frames, it's very possible to have a good idea of where they will be in each of those intermediate frames.

This is just one way to visualize it for people who keep thinking it means the game is running slower, or that it causes situations like the lag snipe in the OP.

Yeap, missing frames are being interpolated and true 60fps applies only to your character and only on your screen.

With 16 Hz sometimes an opponent will see a former version of your character being in front of cover, while on your screen will you see that you are behind cover and therefore you will die, even though it wasn't necessarily your fault. It's that simple, not rocket science.

It would actually be very simple to demonstrate this in action if someone had 2 Switches in the same house (tabletop/portable mode <-- you cannot blame TV sets there and connect them via WiFi @ 5 GHz (802.11ac), which basically is interference-free (everyone should avoid 2.4 GHz). You should have 2-3ms of network latency in that case, with almost no ping spikes.

Any takers?

 

[nynt9]

Some math, which may or may not be interesting:

At 60 frames per second ...

* 16Hz is ~4 frames. That gives you a worst case of 8 frames of latency, due to tick, if two consoles push out updates with just the wrong timing.

* A 100ms ping adds 6 frames of latency. (Ping is round trip, so we don't have to add your ping to your opponent's ping; we'll assume that you both have roughly the same ping, or that the average of your pings is 100.)

* We might have another frame of latency due to buffering in the graphics pipeline.

* Our Lady of Google informs me that average human reaction time to a visual stimulus is 0.25 seconds, which is 15 frames.

So given a decent ping, a tick rate of 16Hz updates the game about as quickly as you can react to it -- 8 + 6 + 1 = 15 frames of latency in the hardware matching the 15 frames of latency in the wetware in your brain. You see a thing, and do a thing, and that's, in the worst case, when you see the next thing.

"Pro" players are going to have faster reflexes, and might be able to shave a few frames off of that timing. They probably also live on faster networks with lower ping, though, so the game isn't necessarily updating outside of their reaction window (and remember that the 8 frames of latency due to tick is out worst case scenario -- most of the time, the lag due to tick is going to be somewhere between 4 and 8. And each source of latency doesn't necessarily straightforwardly add to the others).

Basically, my back of the envelope math suggests that 16Hz is okay, for a tick rate, though it doesn't leave much room for latency in the network, video pipeline,or for lack of latency in the wetware of the squishy human playing the game. A higher tick rate is safer, leaving for more slop in the ping, and is friendlier to really fast humans. I imagine that the lower tick rate is part of the reason why the team at Nintendo went with regional matchmaking, mitigating some of the worst case scenarios with ping. I suppose time will tell whether the squishy humans feel that the tradeoffs that the dev team made, in terms of battery life, mobile data friendless, and kindness to the cpu in the machine hosting the match,were worthwhile. I know that I, with the reflexes of someone who is now pushing 40, am not the best judge for how this feels to a strapping young person in their 20s. But I don't think that the situation is quite as dire as you might think, reading through the rest of the thread.

That's not how this works. Low tick rate means you can move behind cover and get shot. It's not just an artifact of reaction time. It means that, if a person is shooting at you, and you're running away, they'll have several more frames where they can hit you. They can even hit you without ever appearing on your screen, no matter how good your reaction time is. See:

https://www.youtube.com/watch?v=sN1jy2DpIxA

The person who gets shot never even sees the assailant, even though the assailant's perspective shows that they should have. It's not about reaction time. This is when the game had a 30hz tickrate. Splatoon is twice as bad as this.

 

[LordOcidax]

That's not how this works. Low tick rate means you can move behind cover and get shot. It's not just an artifact of reaction time. It means that, if a person is shooting at you, and you're running away, they'll have several more frames where they can hit you. They can even hit you without ever appearing on your screen, no matter how good your reaction time is. See:

https://www.youtube.com/watch?v=sN1jy2DpIxA

The person who gets shot never even sees the assailant, even though the assailant's perspective shows that they should have. It's not about reaction time. This is when the game had a 30hz tickrate. Splatoon is twice as bad as this.

Thats not a problem caused by low tickrate because everybody have the same tickrate in the match... Thats a netcode problem aka Lag.

 

[LCGeek]

Yeap, missing frames are being interpolated and true 60fps applies only to your character and only on your screen.

With 16 Hz sometimes an opponent will see a former version of your character being in front of cover, while on your screen will you see that you are behind cover and therefore you will die, even though it wasn't necessarily your fault. It's that simple, not rocket science.

It would actually be very simple to demonstrate this in action if someone had 2 Switches in the same house (tabletop/portable mode <-- you cannot blame TV sets there and connect them via WiFi @ 5 GHz (802.11ac), which basically is interference-free (everyone should avoid 2.4 GHz). You should have 2-3ms of network latency in that case, with almost no ping spikes.

Any takers?

You're using wireless and everytime wireless switches it speed you have a huge spike that can range from a few ms to whole seconds.

Wireless for gaming sucks. It has hidden problems that have been studied and short of a gamer having a proper router to sort this out you're putting yourself at the mercy of the moment and the envinroment your wireless devices are in.

 

[Negotiator]

Thats not a problem caused by low tickrate because everybody have the same tickrate in the match... Thats a netcode problem aka Lag.

What kind of "logic" is this? You probably also think traded kills are "fair".

Two wrongs don't make one right.

 

[Negotiator]

You're using wireless and everytime wireless switches it speed you have a huge spike that can range from a few ms to whole seconds.

Does this apply to 2.4 GHz, 5 GHz or both? What about Ethernet/USB drivers?

The experiment can also be conducted in docked mode with 2 decent TV sets (game mode on, <20ms*) and USB Ethernet adaptors.

* https://displaylag.com/samsung-4k-tvs-tested-input-lag/

 

[UncleSporky]

With 16 Hz sometimes an opponent will see a former version of your character being in front of cover, while on your screen will you see that you are behind cover and therefore you will die, even though it wasn't necessarily your fault. It's that simple, not rocket science.

While technically true, people will misinterpret this statement. Opponents will "see a former version of your character" for about 4 frames. That's a shockingly low amount of time, too fast for them to be able to react to it. Actually, probably even a lot less than 4 frames due to the interpolation/predicted movement. It is not a situation like the tweeted clip in the OP at all.

When this kind of thing comes up in a meaningful way it will nearly always be an accident. i.e. someone was shooting vaguely in your direction and the game thought they clipped you when from your point of view it barely missed you by 1/16th of a second. No one can really react that fast as discussed above, but you might be able to say you saw it took place.

 

[LordOcidax]

What kind of "logic" is this? You probably also think traded kills are "fair".

Two wrongs don't make one right.

Cant wait to hear your explanation then... Low Tickrate doesn't give any advantage to one player... because is global... Some variation is related to the Pin, or a bad netcode that cant handle well the data, and others factors.

 

[Koren]

Cant wait to hear your explanation then... Low Tickrate doesn't give any advantage to one player... because is global...

They can give an advantage to some weapons... Granted, anyone can use the weapons favored by the lower tick, but that's still annoying.

 

[Negotiator]

Cant wait to hear your explanation then... Low Tickrate doesn't give any advantage to one player... because is global... Some variation is related to the Pin, or a bad netcode that cant handle well the data, and others factors.

My explanation of what exactly?

That's like saying that communism is "OK", because everyone is equally poor, regardless of their abilities. Sounds fair, right?

Spoiler

/s

I don't want to be poor, especially if my connection can afford a higher tick rate.

 

[thepotatoman]

Some math, which may or may not be interesting:

At 60 frames per second ...

* 16Hz is ~4 frames. That gives you a worst case of 8 frames of latency, due to tick, if two consoles push out updates with just the wrong timing.

* A 100ms ping adds 6 frames of latency. (Ping is round trip, so we don't have to add your ping to your opponent's ping; we'll assume that you both have roughly the same ping, or that the average of your pings is 100.)

* We might have another frame of latency due to buffering in the graphics pipeline.

* Our Lady of Google informs me that average human reaction time to a visual stimulus is 0.25 seconds, which is 15 frames.

So given a decent ping, a tick rate of 16Hz updates the game about as quickly as you can react to it -- 8 + 6 + 1 = 15 frames of latency in the hardware matching the 15 frames of latency in the wetware in your brain. You see a thing, and do a thing, and that's, in the worst case, when you see the next thing.

"Pro" players are going to have faster reflexes, and might be able to shave a few frames off of that timing. They probably also live on faster networks with lower ping, though, so the game isn't necessarily updating outside of their reaction window (and remember that the 8 frames of latency due to tick is out worst case scenario -- most of the time, the lag due to tick is going to be somewhere between 4 and 8. And each source of latency doesn't necessarily straightforwardly add to the others).

Basically, my back of the envelope math suggests that 16Hz is okay, for a tick rate, though it doesn't leave much room for latency in the network, video pipeline,or for lack of latency in the wetware of the squishy human playing the game. A higher tick rate is safer, leaving for more slop in the ping, and is friendlier to really fast humans. I imagine that the lower tick rate is part of the reason why the team at Nintendo went with regional matchmaking, mitigating some of the worst case scenarios with ping. I suppose time will tell whether the squishy humans feel that the tradeoffs that the dev team made, in terms of battery life, mobile data friendless, and kindness to the cpu in the machine hosting the match,were worthwhile. I know that I, with the reflexes of someone who is now pushing 40, am not the best judge for how this feels to a strapping young person in their 20s. But I don't think that the situation is quite as dire as you might think, reading through the rest of the thread.

This doesn't make sense at all because reaction time is added on top of latency, not during it. Just accounting for the difference between 16hz and 60hz, if an attacker turns a corner it takes up to 100 more milliseconds before the defender can even see it. Then it takes up to 50 more milliseconds after that for either player to register a hit on the server. So a defender with an average reaction time of 250ms vs a player with a reaction time of 400ms will end in a tie.

That's pretty huge, and can be very noticable in game, especially when you're adding it on top of typical latency. People may not know about netcode, but they will know when outcomes of battles just don't seem right, and it's enough to really shape how the game is played and what tactics or play styles can or can't work. I think it's less noticeable the slower your reaction times because it's less of a difference as a percent, but you don't have to be a pro to notice 16hz vs 60hz. It's also worth noting that reaction times to audio cues is faster than visual cues, which also can play a role in this game, like with grenades.

I don't think it's very different from everyone with 200 ping playing on 60hz servers, and everyone with 100 ping playing on 16hz servers, but there's only so much you can do about ping while you can do something about the tick rate. That means region specific queues with dedicated servers is probably more important than higher tick rate, but both should be addressed to make it more in line with what people expect from a modern day online shooter.

 

[UncleSporky]

What kind of "logic" is this? You probably also think traded kills are "fair".

Two wrongs don't make one right.

Traded kills are fair. This is not a hitscan game. If it had a 128 hz tickrate you would still see a ton of traded kills. Ink blobs are visibly flying through the air toward each player via physics, sometimes 3 or 4 at a time. Of course people are going to end up both hitting each other a lot. A faster tickrate would barely affect this.

 

[LordOcidax]

They can give an advantage to some weapons... Granted, anyone can use the weapons favored by the lower tick, but that's still annoying.

On a perfect world with a perfect pin, 1 ms monitor and very fast input controller... Yes. But thats Pro PC not a console...

 

[Koren]

On a perfect world with a perfect pin, 1 ms monitor and very fast input controller... Yes. But thats Pro PC not a console...

Well, I disagree, because I noticed the difference between Splatoon 1 and 2 before even knowing it was a thing, so...

 

[MrCunningham]

Maybe this would be a better way to illustrate...correct me if I'm wrong, but essentially this means that while Splatoon visually looks like the 60 FPS on the right, network-wise it's running like the 15 FPS on the left:

Note that this doesn't mean your actual game state as represented in memory on your console is running like the image on the left. It's not a 16 FPS engine that just looks smoother. Your own character and any actions you take, your shots, your bombs etc. are probably all running at 60 FPS on your end of things. And there is plenty of interpolation going on too, even though you only get an update as to your enemy's exact position every 4 frames, it's very possible to have a good idea of where they will be in each of those intermediate frames.

This is just one way to visualize it for people who keep thinking it means the game is running slower, or that it causes situations like the lag snipe in the OP.

Basically. The servers are only refreshing every player position at 15frames per second, while the players are seeing everything at 60fps. That 15fps character on the left will still be moving at 60fps to the player. But the server is only capturing 1/4th of that movement. Which means the player will miss a lot more frequently, and will be prone to a lot more phantom deaths.

 

[Negotiator]

Does a higher tickrate necessitate a better connection from players and if so, by how much (i.e., would an increase from 15Hz to 30 Hz also double the required minimal bandwidth)? In other words, are lower tickrates being used to allow people with slower connections to play the game in question, or are there other reasons?

Of course it does, but 30 Hz wouldn't be out of reach for the majority of ADSL users.

80 Kbps upload usage is laughably low. 30 Hz would make it 160 Kbps and it would still be too low. UC4 MP at 15 Hz is ~300 Kbps for 10-player matches and another 200 Kbps if you also use Party Chat.

1 Mbps upload (typical for ADSL) is 1024 Kbps. 802.11n WiFi is like 30 Mbps at least. There's plenty of room to spare as you can see.

Regarding the train part, I thought Japan had uber-fast 4G connections, so it shouldn't even be an issue for them. Then again, there are plenty of great games (like Zelda or MK8) to play while commuting to work, not just Splatoon. It's a mind-boggling decision to nerf everyone's connection "just in case".

Look, even if you don't notice or know about it, what's the harm in increasing the tick rate?

If it leads to an even smoother experience, you should be in support of it.

Some people may have potato-powered connections and/or are afraid of data caps. Everyone should endure a universal nerf because of them.

How are they going to justify paying for a subscription that offers a 16 Hz p2p experience?

Is there a reason we have three threads for this by the way?

Three? I've only seen two of them. Where's the 3rd one?

Playing on wireless without debloated drivers, which is most people is insanely stupid if you care about anything real time and networking.

What's pathetic is none of the big 3 companies have addressed this on their consoles when the solution literally is a driver fix, no need for new hardware.

AFAIK, WiFi drivers are closed-source binary blobs and it's not easy to modify them without having access to the source code. I saw your link about Linux, but I'm not even sure if it also applies to FreeBSD (PS4/Switch) or Windows (XBOX ONE).

Btw, this may also be of interest to you: https://arstechnica.com/information...illion-phones-to-a-wi-fi-hopping-worm-attack/

Switch uses Broadcom's BCM4356 WiFi chipset: https://www.ifixit.com/Teardown/Nintendo+Switch+Teardown/78263

God knows if/when they're gonna patch it!

Also, remember that the majority of people uses tvs to play with their consoles, so the input lag is normally really really high. Maybe if you use a gaming monitor with a input lag below 1ms and a really good pin you can notice something.

Actually no, the tickrate is the same for everybody... Around ~63ms (coorect me if i am wrong) refresh rate from the server. Assuming that everybody have a perfect pin and 1ms monitor (And a bunch of others factors), nobody is going to lag because the tickrate is low, so you are not going to lag compared to others players.

Well, technically with a normal tv (not a pc gamer monitor) + and a average pin is almost impossible to notice something... Can be Better thats for sure, but is not going to solve the "being shoot behind a wall" problem that a lot of people think that is caused because the low tickrate.

You don't even know how to spell ping correctly (seriously, I was scratching my head trying to understand what "pin" meant) and we're supposed to take your non-factual opinion seriously because...?

Listen: it's fine if you enjoy this game, despite the ridiculously low tickrate. I believe you. Millions of people also enjoy McDonalds, despite the added chemicals/preservatives. I believe them as well. Some people have higher quality standards though. Is that OK with you or should they suck it up?

 

[PotionBleue]

It's interesting that he's talking about cheating - this is IMO a far bigger problem that I don't see being discussed much especially on NeoGAF. Online Nintendo games have been plagued by cheaters since the DS/Wii with Nintendo doing little against them.

It's hard to justify paying for Switch online service if this point isn't improved significantly.

 

[Nekketsu Kõha]

It's interesting that he's talking about cheating - this is IMO a far bigger problem that I don't see being discussed much especially on NeoGAF. Online Nintendo games have been plagued by cheaters since the DS/Wii with Nintendo doing little against them.

It's hard to justify paying for Switch online service if this point isn't improved significantly.

I saw one earlier actually in MKD. Some American who had named himself troll...

But cheating isn't a Nintendo problem. Most cheaters I ever run into is on Steam on their VAC servers. Also saw plenty on XBL all the time and I'm sure PSN has its fair share too.

 

[schlynch]

well... since it's more like a casual hop-in-and-go game without any real competetive substance... it may be an non-issue as some has mentioned.

 

[Horp]

Man the defense force posts in this thread :/
Like holy shit, did someone shit in your cereal?

On topic, 16Hz is unacceptable for anything but the most casual of experiences. If they want to go esports this has to be changed.

 

[Waji]

It explains some of my deaths...
It's most certainly something they should upgrade.

 

[#1 defender]

Of course it does, but 30 Hz wouldn't be out of reach for the majority of ADSL users.

80 Kbps upload usage is laughably low. 30 Hz would make it 160 Kbps and it would still be too low. UC4 MP at 15 Hz is ~300 Kbps for 10-player matches and another 200 Kbps if you also use Party Chat.

1 Mbps upload (typical for ADSL) is 1024 Kbps. 802.11n WiFi is like 30 Mbps at least. There's plenty of room to spare as you can see.

I'm pretty ignorant here, so bear with me. If the bandwidth requirements at 15Hz are that low, then why is the recommended minimal upload speed for U4 mp 1Mbps? One would think that they could have a 30Hz tickrate at that speed then.

 

[Huge Succeeded]

On a perfect world with a perfect pin, 1 ms monitor and very fast input controller... Yes. But thats Pro PC not a console...

okay why do you keep acting like you need a computer monitor for any sensation of increased response time to be noticible

because even if that weren't wrong to begin with

it's a silly ass point to keep trying to reiterate in a thread about the flagship shooter for the Switch

the platform whose appeal
whose marketing
whose unique use cases

are all centered around the fact that the Switch console has a screen

 

[notaskwid]

The defense force to this is a riot, keep at it.

 

[faridmon]

While it wouldn't matter to casuals who just want to hop in and out, Its seriously funny that Nintendo is trying to market this as an Esport title.

Yeah, ain't happening

The defense force to this is a riot, keep at it.

You only read what you want to read

Keep at it

 

[Paz]

Maybe this would be a better way to illustrate...correct me if I'm wrong, but essentially this means that while Splatoon visually looks like the 60 FPS on the right, network-wise it's running like the 15 FPS on the left:

Note that this doesn't mean your actual game state as represented in memory on your console is running like the image on the left. It's not a 16 FPS engine that just looks smoother. Your own character and any actions you take, your shots, your bombs etc. are probably all running at 60 FPS on your end of things. And there is plenty of interpolation going on too, even though you only get an update as to your enemy's exact position every 4 frames, it's very possible to have a good idea of where they will be in each of those intermediate frames.

This is just one way to visualize it for people who keep thinking it means the game is running slower, or that it causes situations like the lag snipe in the OP.

Unrelated to what you're trying to explain but FYI gifs are limited to 50fps in browsers so all these framerate comparison gifs end up not being accurate.

On topic - for the kind of game Splatoon is I have no problem with a low tick rate but everyone has different goals I suppose.

 

[Chao]

Does tickrate matter when your teammates always neglect inking your base and makes you lose because of it though

 

[OrbitalBeard]

Does tickrate matter when your teammates always neglect inking your base and makes you lose because of it though

haha

When I'm the only one inking my base I'm like "yeah, this isn't going to end well"

 

[faridmon]

Does tickrate matter when your teammates always neglect inking your base and makes you lose because of it though

Lol

 

[pfr]

Of course it does, but 30 Hz wouldn't be out of reach for the majority of ADSL users.

Remember this is a mobile game console. So this game is going to be played over 3g and 4g networks that are then shared over wifi when on the go. I'm sure this is the biggest reason they went with a low tick rate.

 

[Chauzu]

The defense force to this is a riot, keep at it.

Fantastic contribution to the thread right there.

 

[FoxhoundNL]

Lol. I knew there were some funny things going on in S2. Far too often I see and shoot someone first, to only still be killed by the same guy. I blamed it on the Wi-Fi, but is seems to be netcode then... Now these sort of deaths are going to be even more annoying though

 

[LCGeek]

AFAIK, WiFi drivers are closed-source binary blobs and it's not easy to modify them without having access to the source code. I saw your link about Linux, but I'm not even sure if it also applies to FreeBSD (PS4/Switch) or Windows (XBOX ONE).

Btw, this may also be of interest to you: https://arstechnica.com/information...illion-phones-to-a-wi-fi-hopping-worm-attack/

Switch uses Broadcom's BCM4356 WiFi chipset: https://www.ifixit.com/Teardown/Nintendo+Switch+Teardown/78263

God knows if/when they're gonna patch it.

Only atheros and Marvell based stuff has airtime fixes. There is Broadcom tech that is anti bufferbloat it just not in the switch. Bufferbloat is mainly a cable modem router issue so you fix on those devices not things that connect to the network. BSD like Linux has fixes for it while the windows stack is still shit at addressing various network issues related to bufferbloat.


The site I linked too, lede project, openwrt and dd-wrt all have software to help you with the issues, though having a proper router to modify is more of a pain than anything. Pm if you have anymore questions about the subject been stuck in POE a lot and haven't been reading forums.

Due to various problems on routers gamers suffer from buffer bloat and packet mishaping due to offloading, saturated fifo queues or packet aggregation. Any of these things warp and delay your packet generation which causes even more sync issues internally and eventually to anything you interact with.

Remember this is a mobile game console. So this game is going to be played over 3g and 4g networks that are then shared over wifi when on the go. I'm sure this is the biggest reason they went with a low tick rate.

They could easily have a self adapting engine like overwatch and make everyone happy including themselves consider asymmetrical and adapting would reserve on bandwidth on links unable to handle them. They took a sloppy solution like the rest of their online choices the last 3 gens. One size fits all is incredibly stupid for fps games these days not all links are the same.

 

[killatopak]

as a game wanting to enter e-sports this is bad.

 

[Toxi]

When your tick rate is the same as a N64 game's frame rate.

 

[LCGeek]

When your tick rate is the same as a N64 game's frame rate.

When your tickrste isn't even half of doom from 1995......

 

[Ardenyal]

Using such a low tickrate has latency just based on the huge gap between updates.

Low tickrate also means they have to use interpolation instead of predicting movement and doing minor corrections. Interpolating means they need at least two data points to interpolate between. Even ignoring network lag the game will have tons of inherent lag, add your mobile tethered wifi in a mountain cave lag and it's pretty bad for gameplay.

Pretty weird that Nintendo puts such little importance on gameplay as they have traditionally favored 60fps and snappy response.

 

[Wowfunhappy]

Question: People often talk about low tickrates as a cost-saving measure on the part of the developer, but if Splatoon is p2p, how does Nintendo "benefit" from using a lower tickrate?

 

[Negotiator]

I'm pretty ignorant here, so bear with me. If the bandwidth requirements at 15Hz are that low, then why is the recommended minimal upload speed for U4 mp 1Mbps? One would think that they could have a 30Hz tickrate at that speed then.

1 Mbps is merely a guideline and still many people play with sub-1 Mbps speeds (like 500-600 Kbps ADSL upload). 30 Hz would push them out of the player pool. You get a warning message from ND if their speedtest algorithm detects <1 Mbps, but they still allow you to play, so it's not a strictly enforced rule. They're very lenient IMHO.

8-player Plunder had 20% less bandwidth usage and as I also said, voice chat consumes upload bandwidth too. Many people don't know that though.

Remember this is a mobile game console. So this game is going to be played over 3g and 4g networks that are then shared over wifi when on the go. I'm sure this is the biggest reason they went with a low tick rate.

This game shouldn't be played over 3G/4G networks, not just because of lag, but because of NAT Type incompatibilies too (cell phone providers rarely offer a routable IP address).

Why not play Zelda on the train instead of Splatoon? Why do they have to ruin the experience for everyone else? I don't get it. :|

Only atheros and Marvell based stuff has airtime fixes. There is Broadcom tech that is anti bufferbloat it just not in the switch. Bufferbloat is mainly a cable modem router issue so you fix on those devices not things that connect to the network. BSD like Linux has fixes for it while the windows stack is still shit at addressing various network issues related to bufferbloat.

The site I linked too, lede project, openwrt and dd-wrt all have software to help you with the issues, though having a proper router to modify is more of a pain than anything. Pm if you have anymore questions about the subject been stuck in POE a lot and haven't been reading forums.

Due to various problems on routers gamers suffer from buffer bloat and packet mishaping due to offloading, saturated fifo queues or packet aggregation. Any of these things warp and delay your packet generation which causes even more sync issues internally and eventually to anything you interact with.

I have an ASUS VDSL2 modem/router and I haven't noticed any issues. I don't know if there's something I can modify on its firmware (there are binary blobs in there as I said).

Ethernet ping is always <1ms and 802.11ac ping is 2-3ms with no ping spikes.

They could easily have a self adapting engine like overwatch and make everyone happy including themselves consider asymmetrical and adapting would reserve on bandwidth on links unable to handle them. They took a sloppy solution like the rest of their online choices the last 3 gens. One size fits all is incredibly stupid for fps games these days not all links are the same.

Agreed. Dynamic/adaptive tickrate would do wonders for nearly every shooter game out there.

It can be done if they want: https://www.reddit.com/r/CoDCompetitive/comments/6ki0rz/wwii_will_have_a_60hz_server_tick_rate_when/

Question: People often talk about low tickrates as a cost-saving measure on the part of the developer, but if Splatoon is p2p, how does Nintendo "benefit" from using a lower tickrate?

Nintendo "benefits" by having a supposedly "huge" player/customer base consisted of 3G/4G WiFi tethering users. :\

Quantity != quality

When your tick rate is the same as a N64 game's frame rate.

This and the fact that console p2p games rarely display ping millseconds (those vague bars...)

They really treat us like console peasants and it's the only thing I cannot disagree with PCMR elitists when they mock us.

 

[2+2=5]

Maybe this would be a better way to illustrate...correct me if I'm wrong, but essentially this means that while Splatoon visually looks like the 60 FPS on the right, network-wise it's running like the 15 FPS on the left:

Note that this doesn't mean your actual game state as represented in memory on your console is running like the image on the left. It's not a 16 FPS engine that just looks smoother. Your own character and any actions you take, your shots, your bombs etc. are probably all running at 60 FPS on your end of things. And there is plenty of interpolation going on too, even though you only get an update as to your enemy's exact position every 4 frames, it's very possible to have a good idea of where they will be in each of those intermediate frames.

This is just one way to visualize it for people who keep thinking it means the game is running slower, or that it causes situations like the lag snipe in the OP.

More or less yes, it means that almost 3 frames over 4 are wrong in a way or another because the local game draws them "blindly" without new server data, meaning that the real game and what is visualized could be very different, the fifth frame(made with server data) could be visually very different from the fourth(made without the server data).

Example:
in the first frame that you got from the server you see an enemy going straight, during the second frame the enemy turned or jumped or shoot but since you won't get updates from the server for other 3 frames you will see the enemy going straight for other 3 frames, when you will finally get the fifth frame from the server you will see the enemy warping from the calculated position to its real position, this applies to everything, projectiles included,

 

[SmokedMeat]

Traded kills are fair. This is not a hitscan game. If it had a 128 hz tickrate you would still see a ton of traded kills. Ink blobs are visibly flying through the air toward each player via physics, sometimes 3 or 4 at a time. Of course people are going to end up both hitting each other a lot. A faster tickrate would barely affect this.

A faster tickrate would mean fewer trades kills, because information would be moving at a faster rate. The person with the more powerful weapon would still be standing after the exchange, instead of dying, because the weaker weapon's damage was doubled.

 

[Kurdel]

Splatoon 2’s online experience is considerably worse than the original game in quite a number of categories.

Omg how low is the tickrate? All those categories, how can it be so considerably worse?