No word has defined Ultimate’s online more than “lagtest” – as both players and TOs attempt to play in a competitive environment while dealing with the game’s infamous netcode, they are faced with constant connections problems that leave many wondering how to properly address it. Many methods have come and go, and while we have a much better grasp on things now than we used to, there’s still a lot unknown to the community, as well as a lot of misconceptions being thrown around as to what actually causes lag, how to fix it, or even diagnose it.
In this guide, I hope to teach you all about lagtests from a TO’s point of view, and how to properly do one when the situation comes around. Keep in mind that this is NOT an in-depth analysis of how the netcode itself works, or specific details over ISPs, networks, etc. This is simply a reference as to what steps to take if two players request a lagtest in your tournament.
Before I get into it, please remember that TOs are not network administrators – we do not have control over every single aspect of online, nor do we have some secret connection with Nintendo to be able to fix matches. We are playing the same game as everyone else, and have access to the same tools that anybody else does. We’re constantly finding out new methods and adapting to situations, so this’ll be updated as time goes on.
If you want more information on how Super Smash Bros. Ultimate’s online works, I recommend reading this page right here:
Without further ado, let’s dive right in!
What do I need to do a lagtest?
The best method to perform lagtests hands down is SableDetect – this is by far and wide the most efficient tool available to TOs to help diagnose issues, and is publicly available for free. It requires a little bit of extra setup on your computer, but is very easy to do, and in my opinion, should be required for all online tournament organizers.
As the website describes it: “Sable Detect is a windows based packet sniffer that is tuned to the SSBU protocol. It monitors incoming peer to peer traffic between 2 to 4 switches and keeps track of network flow for you. It checks for network instability, ping and more! The diagnostics it provides can help a TO or host determine who is the cause of the lag very quick and keep their event going smoothly!“.
While the program is obviously not perfect, it is by far the best method we have to monitor both players’ internet connection, and can help you quickly figure out what’s wrong with the match, and whether someone is at fault or not.
Here is also a more detailed guide as to how SableDetect works, so make sure to read up on that as well.
Many of you may remember that earlier on in the pandemic, speedtests were often used to determine lagtests. However, as you may have noticed for a long time now, it is not a recommended method anymore. The reason for that is that download/upload speeds are not nearly as relevant as people think for Ultimate specifically, and that packet loss/consistency is much more important to have a smooth connection. Having higher download/upload speeds does not guarantee or determine a more consistent, better connection with your opponent, so please do not use speedtests as an efficient method of lagtesting, nor should you bring up your results as justification for not being the source of the lag.
If you are unable to use SableDetect, or need an alternative as a backup, then I recommend this website:
This will allow players to perform a ping test via their web browser, which is much more indicative of their connection’s quality than speedtest or any other website. However, I still highly recommend that you use SableDetect no matter what.
What are the steps to performing a lagtest?
Before I continue, remember that you should only be performing a lagtest before or during a set – you should NEVER perform a lagtest after a set was played out, as the games are already completed. The players should be requesting a lagtest AS SOON AS THE LAG OCCURS, or in-between games, otherwise this can open up a can of worms, and potentially allow for players to abuse lagtests to slow down brackets or revert scores despite clearly playing out the set. Additionally, a successful lagtest should never result in replaying a game, unless the players stopped mid-match, in which case you can have them replay from where they left off. If they stopped early enough, you can have them replay the game, but if they stopped at the very last second, you do have the ability to deny a lagtest, as it would be far too late in the game at that point.
When it comes to actually doing the lagtest, there are only two very simple steps to follow that will get you the information you need:
1. Get proof of LAN from both players
As many of you are aware, having a LAN Adapter is vital to having an optimal connection to play online. No matter what anyone says, you should always be wired in order to play online.
When asking for proof of LAN, there are two acceptable proofs:
– Getting a picture of the in-game arena with the Switch sidebar open (first picture)
– Getting a picture of the Switch homescreen (second picture)
The reason why we require these is because there are two crucial items that are visible – the LAN logo and the time. The first is, obviously, to check whether or not you are playing on a wired connection, and the second is to make sure the picture was taken at a time that matches when you asked for said proof of LAN. If the Wi-Fi symbol is visible, or if the time does not match, then the player may be lying. Always make sure to get a FULL picture, that way you can see everything that’s visible.
It is recommended to get a picture with the Switch sidebar, as you can also see the player in the arena itself, as opposed to simply seeing the game as “Playing”. This is done by holding down the Home button on your Joycons or Pro Controller, as opposed to simply pressing the button. A picture of the players’ LAN adapter is NOT acceptable proof, as it can very easily be faked, and does not show that it has been properly setup. Additionally, a picture of their Nintendo Switch network test is NOT acceptable either, as it does not properly prove it’s the same player’s console. It is also not a good reference of someone’s internet connection, for reference.
2. Spectate a 60 second button check between both players/individual players
Once you’ve acquired proof of LAN, then you can head to the arena and actually perform the test itself. Make sure that the arena has a minimum of 3 spots available, and that there are no other players present in the arena. This is to ensure the best possible tick rate, and to avoid any potential disruptions during the test.
Hop into the spectator stands, and ask the players to do a button check for 60 seconds. While you can certainly test them for longer if you’d like, this is considered the sweetspot to determine whether or not both players’ connections are stable enough to compete. Doing it for a shorter period of time can cause you to miss potential lag spikes, and doing it for a longer period of time can further delay the bracket and not be entirely necessary.
During the 60 seconds, pay attention to SableDetect and the information it gives you. Here’s a quick summary of what you should be looking for:
F (Flags): This represents any major network interruptions that occurred during the test. Reaching 10 flags within 60s is considered an automatic DQ, however if the player reached a number close to it at the end of the test, you may have to consider a DQ as well. Having 1 or 2 flags does not necessarily mean they have a bad connection, as it could simply be a false flag and not mean anything.
L (Late Packets): This represents the amount of packets that came in late. Reaching 100 late packets within 60s is considered an automatic DQ, however the most important part to pay attention to is not the amount of late packets received, but moreso the frequency at which they are lost. If the player barely loses any, or is losing packets at a very slow rate, then it’s not egregious enough to DQ them for it. However, if they are rapidly, consistently getting late packets, or getting them in chunks, then you should be concerned.
O (Out of Order Packets): This represents the amount of packets that came out of order. Reaching 50 out of order packets within 60s is considered an automatic DQ, however as mentioned above, the frequency is generally more important to look at than the amount itself. It’s important to note that having more out of order packets is more problematic than having late packets, so keep that in mind when making a decision about a player. Also remember that even if you don’t reach the exact threshold for late/out of order packets, getting a number close to it could still be indicative of an unstable connection.
P (Ping): This represents your ping to that player. If you do not see a value for that, DO NOT PANIC – it simply means you’re not able to ping the player specifically, but does not mean that they have bad ping. Pay attention to your distance to each player, and the stability of the ping. The value will vary depending on distance, but as long as it stays consistent and doesn’t fluctuate greatly, you should be fine. This is more of a backup value in case you need more evidence, so please only refer to this data if you do not see anything else wrong, or if you properly understand correct ping values based on your location.
D (Data): This represents the amount of data/packet size being sent from that player. This does not indicate whether or not they are lagging, but is moreso a reference to make sure that they are properly sending information. You can generally ignore this value, but if you notice anything abnormal in this column while there is nothing shown in the columns described above, then there may be something wrong with the players’ internet connection. Again, just something to keep note of.
While performing the test, make sure to pay attention to the game as well. If there are major lagspikes or slowdown during the game, the program should be able to detect it and give you the correct information. If for some reason the program isn’t able to detect it and players mention connection issues, or if you don’t find the data conclusive enough, then you can perform individual tests as well. This avoids further discrepancies, and can potentially give you more accurate results as well, since you are alone with the opposing player. Take into consideration as well that due to the latest 11.1 update, SableDetect may have trouble detecting lag in certain situations, so relying on individual tests and getting a feel for the game itself, or using the backup method mentioned earlier, can potentially help you figure out the cause of the problem. You don’t need to do individual tests every time, but it is a common alternative that can be more effective than spectating both players at once, so use it when you think it is needed. Keep in mind however that this does take longer, so if you have multiple lagtests queued up, it will slow things down even more.
Some concerns have been brought up in the past as to the distance & location of the SableDetect user in relation to the players. Unless the distance is absurdly large (say over 8,000+ miles), you SableDetect should not have any issues in performance. That being said, you should always aim to have someone close to the players to perform the test if possible, but if absolutely necessary, you can perform a test from further away. In terms of the location of the tester, this becomes a bit more complicated. If you are in North America, you should not have any issues with this. However, if you are in South America or Europe, you may have a problem where the tester’s location can affect SableDetect’s results, due to the route that the ISP takes to connect to the players. This is extremely dependent on your location, and is something you will have to do research on to find out more. In most cases, you should not have to worry about this, but you should consider this in the event that you run into a lagtest situation where one or both players claim to not have experienced any lag before/with you, despite SableDetect showing otherwise.
Again, it’s important to remember that SableDetect is a diagnostic tool – it helps you understand the players’ connection better. Simply having a number higher than 0 in the flag, late, or out of order sections does not mean you are automatically lagging, which is why it’s important to be able to understand the data that’s given to you and be able to read the information the program gives you. A lot of decisions you’ll make with SableDetect is based on experience, so the more you use the tool, the better you’ll become at it. I recommend using it while playing Quickplay just to get a better idea of how it works and being able to better understand the program. As great as this tool is, it is not perfect by any means, and will not always be able to pinpoint the exact problem in every situation. This is why you should use it as a resource to help you make better decisions, and to use your best judgement/experience when dealing with certain situations.
What do I do if the players have a lot of delay, but no visible lag?
Delay is mainly caused by the distance between both players. Using SableDetect, you can see the approximate distance between them, which can help figure out why there is so much delay. Having 2000+ miles is an easy way to have a lot of delay, but there can also be delay within a 300-500 miles connection. While there are scenarios where a players’ connection can cause additional delay, there is generally not much you can do about it.
The best solutions to combat delay is to have the other player host the arena instead, or to host it yourself and test out the delay that way. Sometimes it won’t change anything, sometimes it’ll make it better, sometimes it’ll make it worse. It’s highly dependent on everyone’s location, and can very easily change, so it’s something you just have to figure out on the fly/with experience. But again, there isn’t much you can most of the time to really fix the issue.
What do I do if one or both players disconnect from the arena?
In the event that a disconnect occurs during a lagtest, SableDetect should be able to pick up the cause of the problem in time, so make sure to pay close attention to the program while doing so, as well as who is still in the arena afterwards. However, you may need still need to perform individual tests, as SableDetect may not be able to detect the correct data in time while spectating, or if the DC happens early enough, it may not have time to catch anything at all.
Unfortunately, similar to what I described earlier, the latest 11.1 update has caused accidental disconnects to occur, meaning it is likely that it may not be either players’ fault at all. Again, make sure to perform individual tests in this happens, and use your best judgement in this situation to determine whether or not a DQ is necessary.
What do I do if two players cannot join each other’s arena?
This is a rare, but infamous problem that has caused a lot of issues during events, and one you surely have seen memes about on Twitter. There are steps to potentially fix this issue, but remember that in 99% of cases, there is NOTHING YOU CAN DO ABOUT THIS. Like I mentioned at the beginning of this guide, TOs are not network administrators, nor do we have access to every players’ ISP or Nintendo themselves, so this is something out of our control. There are a few solutions you can try however:
1. Have the players add each other as friends, and host a friends-only arena
Not only will this check whether or not one of the players has the other one blocked, but for some reason, friends-only arenas use higher port values, which can make it more likely for two players to connect to each other. This is generally the best option available, but it does not work every time, so don’t assume this is a guaranteed solution.
2. Have someone else host the arena
As mentioned in the section about delay, sometimes hosting the arena for the players can circumvent this problem as well, and should be something to try out in case they cannot connect to each other. If you’re hosting an event in a region with multiple countries competing (such as Europe or South America), it is a good idea to get someone from a country that is not one of the two players having issues, as it’s the best way to potentially fix this problem. But again, this is not a guaranteed solution, just an option.
3. Resetting the router
Sometimes, the good old “turn it off and on” method can solve the biggest of problems, so it’s worth trying this out as well. This can also help fix connection issues in general, not just the inability to join someone’s arena. However, keep in mind that some routers take a long time to properly restart, and not everyone has the ability to do so during an event, as there may be other people in the players’ home currently using the internet connection, so this should be seen as an absolute last resort as opposed to a necessity. It’s not likely that this option will work if you’ve already tried the two others, but you never know.
A common misconception is that your NAT Type may be the cause of this issue. That is not true – having a NAT Type of A or B does NOT determine your connection’s quality whatsoever, and is simply determined by the port window for port direct. If the players have a NAT Type of C or below, they would run into connection issues beyond just joining someone’s specific arena, and may not be able to play Ultimate online at all, or use any of Nintendo’s features. You can still check both players’ NAT Types if you’d like, but I can almost guarantee you it will not help resolve the issue.
If you frequently run into this issue with players, you can take a more extreme route to fixing it and turn off the DMZ in the port settings of your router. This is much more complicated to do and can be vastly different depending on the type of router you have, so you would need to do research on your own in order to do so. Lastly, this can be a very long process if you’re not familiar with it, so I would not recommend doing this during an event, as it slows down the entire bracket.
If both players are unable to connect despite everything you’ve tried, then it is up to you to figure out how to choose whichever player moves on. I would personally not recommend using a speedtest or ping test as an indicator, as if both players have completely reasonable connections, there is no reason to choose one player over the other. I personally have both players do Rock Paper Scissors, as there is a clear winner from the outcome that you can base your decision on. I completely understand that this may not be fair to the losing player, but please remember that the entire situation is neither the players’ fault, nor is it within the TOs’ control to fix it. It can simply be due to Ultimate’s netcode, or the route that is taken by one of the players’ ISP to connect to the other player, which again, no parties can control during a tournament. Always remember to be honest with both players, and to be understanding of any salt or emotions they may have when this happens.
What are the best methods to avoid lag?
If you’re a player looking to avoid as many connection problems as possible, here are a few things you can do to ensure the best possible connection possible:
1. Get a LAN Adapter
This should be REQUIRED in order to compete in online events, as it ensures both the best possible connections and consistency when playing online. Even if your “Wi-Fi is good enough”, you should still get a wired connection to maintain stability during gameplay. There’s a huge amount of information out there about wired vs wireless connections, so I will let you do your own research on that.
In terms of what LAN adapter to get, I recommend this one:
This is considered by most to be the most efficient LAN Adapter for your Nintendo Switch, is extremely cheap, and is also compatible with multiple other devices. It also connects via USB 3.0, so you can use the one port for that on your console to ensure the best possible performance from it. If you don’t end up getting this exact LAN adapter, I recommend getting one with USB 3.0 to ensure the best performance possible.
In terms of an ethernet cable, I recommend you have a CAT 6 cable or above, as it is the recommended category for most connections. Having a CAT 5 or below cable can be an issue, so it’s worth upgrading as soon as possible. To check what type of cable you have, you can look near the actual plug of the ethernet cable, and you should be able to find markings that tell you what type of cable you have.
Connecting your Nintendo Switch directly to your router is the best way to ensure a smooth connection, but if your router is too far away, then I would highly recommend looking into a powerline adapter – this allows you to have a wired connection to your router via the electrical circuit in your home. It may not be as effective as a direct connection, but it is still a very strong alternative if needed. I personally use one at my parents’ home, and have never had an issue with it.
HIGH PERFORMANCE OPTION: https://www.amazon.com/TP-LINK-Powerline-Pass-Through-TL-PA9020P-KIT/dp/B01H74VKZU
2. Turn off other devices/close other applications
This applies to more than just Smash Ultimate, but if you want to avoid any other bottlenecks, I suggest making sure that there is nothing else running while playing online. If your home connection cannot handle it, having something like Netflix, YouTube, Steam, etc. can affect your connection and potentially cause slowdowns/lag, so make sure to get rid of anything that could potentially be taking away bandwidth. If you are streaming, turning off your stream can also be a potential fix to your connection, so make sure to try that as well if all else fails.
3. Open up ports on your router
If you really want to get to the core of your home network, you can open up ports to facilitate your Nintendo Switch’s connection to the internet, and potentially solve any connection issues you may have. As described earlier, the methods to do so are highly dependent on the type of router you have, so you will need to do research on your own for this. Feel free to use this website as a reference of what ports to open, and how to do so:
Turning off the DMZ, as mentioned before, can also be a potential solution if needed, but should not be the first option you go to.
All in all, keep in mind that a lagtest is simply a diagnostic to determine whether or not there is an issue coming from one of the players’ connection. It is in no way shape or form meant to be a way to permanently fix the connection quality between two players. Ultimate’s netcode is notoriously bad, and there is honestly nothing you can do in most situations that will fix the issue. It is very likely that there’s nothing wrong during a lagtest, but then lag occurs during the match anyways, that’s just the nature of the game. Please use your best judgement when performing them, and remember to be honest, transparent, and civil with the players with whatever decision you make. I hope this guide has helped you understand them better, and if you have any other questions, feel free to message me on Twitter via @Cagt3000.
Thank you for reading!
Cyrus “Cagt” Gharakhanian