I’ve had a lot of people leave on me in ranked play and some I know for sure are rage quits because I came back from the brink of death and they get salty, but this is just a sign of bad things happening. I’ve seen these type of desyncs when I used to play emulators online, but this may be the first time on console I’ve seen it.
I’m not happy about the way the game handles rage quits. It seems like no one gets punished, and now on top of that, the desyncing issues seem to be a big problem. Hopefully they’ll all be sorted out soon.
I feel like this isn’t a discussion of, “can they patch it” but instead it’s a discussion of the direction behind KI3.
Tons of people were raving about the netcode, but now it’s very obvious that the netcode is not at all at the standard of fighting game netcode.
Instead it FEELS good and LOOKS good, and makes people feel good about the xbone jump, when in reality without an overhaul the online is basically nonexistent.
Now, I said nonexistent, but I should clarify, it truly depends on the frequency of these sync issues which is very difficult to pin down because so few people will upload their matches. I’m estimating that these sync issues are quiet high but that could be wrong, I just lack faith in an untested dev and M$.
Most things that are big now were things that people once lacked faith in. Capcom lacked faith in Ono and his attempt to make SF a relevant fighting game series, yet that turned out pretty well. People didn’t think Mortal Kombat could become a relevant, competitive fighting game again and it worked out. Which those same people then went to make Injustice as well.
I honestly have not had this happen to me yet that I can tell. If you watch the video, it should be obvious to tell when someone is desynching (I’ve had this happen to me in other games before) since the opponent should start stalling out and randomly moving very little to not at all. You can see in the video that the Sadira player in the bottom video is just not very active or almost stops moving altogether. The Sadira player in the top video is way more active and completing combos while the other Sadira looks lethargic and almost like a CPU.
This is not just about an Xbone jump or whatever. People want to believe “no there’s no way a game on the Xbox One could be good, it is better that I saved my money and didn’t buy in, I need a way to feel good about not buying something”. You can do that all day but in the end you can watch videos of Maximilian playing against other top players like Filthie Rich, CD JR and Xplicit Mind/EMP Riyo and see that they are having intense battles with no real complaints between anyone. Just like SFIV and those who didn’t want to believe that Ono could make SFIV relevant, this game will do the same and make people believe that an Xbox One sequel to what was more or less a kusoge fighting game will be VERY relevant in 2013/14 and beyond.
This is a rare case scenario, but of course youtube makes anything rare case look like it happens to everyone all the time.
Its a very very rare issue and it happens with rollback fight code you could even say its close to non-existent. You just seem like you want to be a hater of sorts? KI3 is headed in the right direction… ask anyone who has actually played it. The game is amazing. Best netcode in any fighting game ever.
It sounds to me like you are doing the same: covering up problems just to feel fine you spent 750 $ on a game that matches you with bots when the connection skips.
I also doubt it is a rare case scenario. The reason why I doubt this is that within the first week of the game being out there is actual video proof of this issue.
Consider how rare it is to have two players upload their replays of their casual match and both of them uploading them on youtube. That’s something I would call rare. I’m almost sure you cannot find one single case of this (two people uploading a replay of their low level match) in sf4 even though it has been out since 4 years and the number of players is infinitely larger.
You call it “close to non-existent” but that’s just because you cannot accept having spent 750 $ on something faulty
A DSP match being uploaded is not rare at all. He’s been doing it for years. And the first guy uploaded the ragequit of DSP for that reason - because DSP is a known person in the FGC. And since DSP is well known, I’m sure it didn’t take long for it to get back to him.
And most people who have watched DSP videos probably know that DSP does not ragequit. I’ve seen multiple people ragequit on him though while watching his videos. He’s pretty entertaining although a bit trollish.
For every video you see uploaded on youtube of this incident, there are thousands of matches that played out just fine. And yeah I’m making a total guestimate with that number, but you get my drift.
I’m not talking about KI videos of his, I haven’t watched any of his videos in over 2 years. Back when he was constantly uploading SF4 videos is what I’m talking about.
If this was a huge problem thered be multiple threads from different people and multiple youtube recordings of this. Again, its rather easy to tell when this is happening because the opponent stops playing like a human like in the video.
@musacci, I’m sure what you are talking about is a bit less rare than you may think. DSP is a regular on youtube so someone recording a match of his play isn’t farfetched. The rest I will agree to disagree with and am completely impressed with how little delays and skips there are in the netcode.
I can’t say because I don’t own an Xbone yet, but I feel like if the desyncing issues were a big problem, then I would’ve heard about them more, because Maximillian is just uploading everything and reviewing it and all he’s said is that is it’s good. As close to playing offline as online is going to get. If this were a rampant issue then I’m sure I would be hearing about it more from him, and acidglow and trueunderdawg, since they’re all playing it and uploading a lot.
Yeah RicoSuave is like rank 14 now and I never heard him saying anything about the netcode synching out while he was there. There would have been a considerable amount of reports of people supposedly “beating” him otherwise.
Sorry to burst your bubble but that’s actually one of the basics of game networking: because lag is just part of how net connections are and can’t be avoided, in any networking model that doesn’t outright freeze the game when lag is too big your opponent is handled by the program. Usually the “bot” will simply keep pressing the same buttons the other players were pressing when you last heard of them because that’s the simplest to program. In multiplayer FPS world this is called “dead reckoning”. But all the papers on the subject (including the one by the dev of GGPO) acknowledge more elaborate AIs could be desirable because the better your bot, the more lag you can afford, and as long as you eventually resync it doesn’t make a difference (because the resync overwrites any decision made by the bot).
Understand this: the very basis of game networking is not avoiding lag, because you can’t, lag is built in the very nature of playing over a huge network. It’s hiding lag. All the games that are lauded for good netplay over the net are doing exactly that, so all can exhibit this type of behavior (famously, the first ever network game of Warcraft ended by both players winning). Sync when you can and outside of those syncs your player is playing an approximation that’s slowly drifting away from reality, so it better be a good enough one to fool him!
So under that principle KI’s bot is actually a good idea and works as intended. What doesn’t work as intended is that no resync is coming and the game does nothing about it, so it lasts long enough that you stop being fooled, and that shouldn’t happen. GGPO for example will only allow that to go on for about 20 frames, after which it will call quits.
If this weren’t a rare case there would be tons of videos, or at least reports of matches deviating until the end. But there aren’t, so it’s not. Which also explains why it can be missed by QA.
A coverup would be stupid: if the issue was so common they’d already have enough data to have fixed it, which would have been much simpler.