Input Delay

Testing QoS

The aim.

The parameters I am changing are not in normal QoS setups. They are “under the bonnet” parameters. Quantum , burst and cburst. Some firmwares may have options to set these. I only apply these to the gaming UDP prioritisation.

I will explain what and why I am testing my QoS setup.
I have an ASUS AC88U router that runs Merlin´s firmware. I have flashed this firmware to replace the stock ASUS firmware. It gives me more options than the stock firmware, plus it also adds other features.

Now a few years ago it was discovered that ASUS QoS was not set up well. Credit to FreshJR on the SNBFourms. FreshJR went missing in action a year or so ago but after a while, another user decided to continue where FreshJR left the QoS project.

I do not need QoS, or so I thought. I have a 1 Gig symmetrical connection and never get close to maxing it. I would say the most I use is 50Mbs at one time. That is probably way more than what I use.
Occasionally I load the new script, FlexQoS, and have a play around with it. My latest “play around” came about because of my last tests with Wireshark, whilst testing the differences between my fiber and 4G connections. I noticed 4G had a lower burst rate. The gameplay was marginally better on the 4G connection. So naturally, I decided to look a bit further into this.

Early on in research, I read a few techies mentioning that if there is a high burst rate that it can affect the receiving end and how they can or can´t handle the packets. Too fast and they may struggle. This got me thinking. We do not know how ISP or EA´s network is configured to handle data rates. The way how I see QoS working is is if it is prioritizing packets it may burst them if they get queued. These packets would be sent at a higher rate.

So I want to test with the scenario of lower burst rate settings or settings related to this part of QoS.
These parameters are not usually open to change on home routers. Home routers are designed to be more user-friendly because most users do not understand or have the patience to sit down and learn the deeper settings. This, IMO, can be detrimental depending on settings outside of your network.
I aim to test with lower “under the bonnet” parameters and see what effect this has. I will use this blog page to update every day or every few days. I will rate gameplay based on after the script renews each day and then with the adjusted parameters.

Week 1 1/03/2021.

I changed the parameters on Saturday 27th of February so I will in clude Saturday and Sunday in this week.

Saturday: Whilst doing the Caputo objective. Gameplay before the parameter change was around 6-7/10. I would say more 6 because it felt like dribbling was not registering all the time. Changing direction quickly etc. defenders seemed reasonably awake though.

After the parameter change things felt more alive for me. Was able to get the objective done really quickly. The dribbling was more responsive. Skill moves seemed to be more in sync. 

Sunday: I didn´t put the parameter change in but did play 2 rivals games. won 1 in Div 5 5-1 against a weak looking side. The second game was a stronger but not quite the quality of my side. I felt a bit disadvantaged, certainly compared to the previous game. It was a draw. Gameplay 6/10.

Monday: Before the parameter change. I decided to continue the Fofana objectives. I played about 5 games in Managerial Masterpiece and had to quit every game. I had low control over my defence. I could see midfielders not stopping in time so trying to cut off passing lanes or tackling was quite hard. I noticed passing looked a little delayed. Dribbling was not responsive. I could build play up but tactics change didn´t help. Even bringing on loan Mbappe didn´t help. Gameplay was around 5/10.

After the parameter changes. The gameplay felt way more equal. Defenders awake. Cutting passing lanes was more effective. The ball would be collected rather than ignored. I was expecting the 1st game to be a one off so played another 2. Same. Much better gameplay. 9/10. Afternoon only two games and both in Managerial Masterpiece for Fofana. Felt a little bit worse than the morning although both players were clearly using high pressure. The first came a cropper in the second half. The second rubber banded after scoring with his Renier but noticed his centre backs were on zero stamina lol. Gameplay 7.5/10 which could have been the high pressure causing this. 

I only played one more Managerial Masterpiece game just before new SBCs dropped. Usually this can be the worst time to play.  It felt slower but still fair. This was with the new parameters as the script had not re set it self.

Tuesday: The script has reset over night so I am playing the first set of games under my normal QoS conditions. Straight away I notice my opponent faster , sharper etc. He scores I quit a bit later after bringing on Mbappe and noticing how he has no dribbling ability. Cutting passing lanes is a waste of time. The next game I played someone possibly just a few kilometres away. I take an early lead then bring on Mbappe. Gameplay felt slightly in my favour. He quits. The next game I play him again , but this time I am away. Players can´t move like the first game! Quit at 0-3. The next 2 games felt heavy and with one sided delay. Not really bad but noticeable. I tweet about it lol. Gameplay overall felt 5/10.

I had half an hour break and then set the parameters in QoS. Only played 3 but instantly a noticeable difference. Not the best gameplay (but still better than earlier) until I switched formation in game and had gone from drop back possession 4-4-2 to press on possession loss 3-5-2. The game seemed to sway regarding momentum. gameplay was only about 6-7.5/10 depending on the in game changes. Gameplay was a lot fairer than the earlier session. Cutting passing lanes did win the ball occasionally. Tackles would also win the ball. Dibbling was a bit better.

In the next session things changed for the worse. It was like I had normal QoS parameters. Several times I re input them but to no avail. I checked Pingplotter to the server I last played on and noticed ICMP packet loss and jumpy pings in the tier network before the iD3 network. I gave up for the day.

Wednesday: I set the parameters and went on to quit at least 5 games because of the same issue yesterday. Instead of checking Pingplotter this time , I decided to refresh my dynamic IP (That is the public IP address your ISP gives you). This can sometimes help. But disaster struck after setting the parameters again. I was locked out of my router. The WiFi and Ethernet wasn´t working either. I hard reset (turned the router and ONT off and pulled the plugs , left for 5 minutes) the modem and router and then tried again. Once again bad gameplay. Ah , QoS was set to traditional QoS which would therefore not be running the Flexqos script. So I re-installed Flexqos , put my settings in again and then set the burst , cburst and quantum parameters in.

In doing the reset I also got another new dynamic IP.

When I fired up I was a bit worried that nothing would have changed , but how wrong was I?

Bang. Mbappe now has fluidity and can turn players inside out (Managerial Masterpiece). I win 7-1 . Opponent rage quit on the hour mark.

I then broke for lunch. My mood changed after that game :).

About 5.30 CET I fired up again. Rivals. Once again the gameplay had become bad. Not as bas as the morning but still noticeable. defending and cutting passing lanes in particular. I noticed nothing spectacular on Pingplotter apart from a few jumpy pings in the ISP network. After several rage/Skynet quits I decided to block the Spanish servers and try the Paris ones.

Only played 2 but the difference was once again like night and day compared to Madrid. Players would intercept balls right next to them instead of letting it go by. I could build meaningful attacks , dribble better , shoot with gusto etc

So servers double the distance are playing better for me at this particular time. I played again a few hours later , for the silver objective IF player. The gameplay was ok maybe 7.5/10. Not quite as good as rivals had been slightly earlier but still way better than the morning.

I will go back to the Madrid servers on Thursday. 

Not a good day for Madrid. Another user confirms their servers are playing badly for them too. Our opponents are getting great gameplay though. That should not be the case.

Thursday: A quiet day onlinewise. Just the one game played on the silver TOTW objective. No adjusted parameters , but no issue with that game.

Friday: Played a couple of Managerial Masterpiece games in the afternoon for Fofana without the parameters set (Spanish servers). Bad gameplay. 5/10. Players stuck in the mud. No noticeable improvement after setting the parameters. Blocked the Spanish servers and despite trying to match me again in Spain I got matched in Paris after a couple of ready ups.

The gameplay was superb in comparison. 8.5/10. Players were fluid. AI was also awake. Got two rage quit wins. Ranted on Twitter. Played a few later on and although not quite as good still fair enjoyable gameplay.

Saturday: Bck on the Spanish servers and with my 3 new recruits. What If Lucas Leiva , Flashback Tony Kroos and the new IF De Jong card. Changed formation to fit in LL21 but can only manage 7 chem. But in the gameplay this morning it didn´t matter. He was good like the gameplay. Early rage quit in WL and a one sided 0-0 in rivals. Had 8 shots to his 1 although he had a weaker side and had no complaints that he had no option but to play drop back. At half time he had 12% possession. We go again later.

The afternoon session turned from good to bad. Right from the word go. Ended up rage quitting 3 WL games and then reverted to testing on the Fofana objective. That was also a real struggle. Things improved after changing the parameters to really small values , losing a game badly , then setting them to slightly higher than previous. I got 3 or 4 games in and quite enjoyed it.

Sunday: The plan was to get about half the WL games done but after testing on the Fofana objective. I ended up not playing any WL  games in the end because the game was horrendous nearly every game except usually on the first after parameter changes. This was on both the Spanish and French servers.

Week 2 8/03/2021.

Monday: Tested without parameter changes around midday on Division Rivals. Not great. Clear one sided delay worse than anything last week. A few times I reset QoS and then entered the parameters and it seemed to play fairer for a half or a game and then reverted back . 

I hard reset my ONT and router and , whilst waiting I fired up the mobile hotspot. It played exactly the same. Bad.

After the IP refresh it did play very well but this only lasted for 2 games. Then back to not being able to dribble , AI not bothering to do much etc.

Before the new SBCS came out ,it was even worse. One game was so badly out of sync , the controller nearly went flying.

Luckily for most of the rivals games, I had Skynet on hand so didn´t use up too many of my attempts. With gameplay leaving me at a serious disadvantage, why should I miss out on some rewards?