View Full Version : A top ten overall cheated.
about 20 minutes prior to this thread I was just kicked out of a battle that I spent at least a hour and a half on (all the matches included). I was waged in a deathmatch against ranked #3 "mike's awesome". I outnumbered him 2 to 1 with 1 HP each and his turn before the my final turn, he surrendered and instead of honoring my victory he disconnected right before the leaderboard upload and granting me my deserved promotion in rank. This touches on the point of the major flaw of the Wi-Fi capabilities is that anybody can be a coward and run away and not even be punished for it.
Linewalker
25 Sep 2007, 12:00
Mike's Awsome and I had a few battles last night, I think we played about 3 games in all. He won overall, but he didn't disconnect in the one I won. Probably best to give him the benefit of the doubt unless he does it again.
I actually D/C'd against the number 2 player last night, but it was a wireless problem. He beat me two games straight (I was having a bad night :( ) but the second game my wireless cut out on me.
Mong, if you're reading this, and if you can read English, I owe you a victory buddy.
It happens all the time, on my opponents final turn before mine they surrender and then they disconnect so their rank won't get reduced.
810Damage
25 Sep 2007, 12:48
was this on psp or ds
It can happen on both. I was facing the #7 Overall on the DS version and won the 1st round. On the 2nd round after he had his go I was thinking of what to do and suddenly his team surrendered(while it was MY turn remember). This could've been an accidental disconnection so i'll also give him the benefit of the doubt.
It happens all the time, on my opponents final turn before mine they surrender and then they disconnect so their rank won't get reduced.
i suspect very much that their rank decreases over a disconnect. The only question is if your rank improves, or decreases as well.
I also strongly believe that t17 would be able to work out which one has disconnected, since i'm pretty sure they host the servers, and the ds's are just peers. Meaning the server knows who DC'ed first.
I'm not entirly sure how their ranking code works with disconnects, but i can't see any reason why they can't penalize and reward as due.
The biggest problem is that the client does a hard abort when it knows the other opponent isn't playing, so you don't get to see your rank update, even if it already has.
Of course, if i'm wrong about the part about t17 being the server, and the ds are just peers ,then i'm compleatly wrong, but i'm pretty sure thats how t17 wrote it.
Moiz575
25 Sep 2007, 19:23
Linewalker, i had a match with u a couple of days ago and it was sweet. i think i added u to my rivals. my team name is moiz. wanna play sometime?
Man, that just freaked me out to see a Post by Linewalker. You see a name so much on the top 10 leaderboard, its like mythical. Good to see you posting!
Got any advice for us? Tips? Any really cool victories?
parsley
25 Sep 2007, 20:55
I also strongly believe that t17 would be able to work out which one has disconnected,
Not possible in a two player game.
since i'm pretty sure they host the servers
We don't, but that's beside the point. Who hosts the servers is irrelevant: that A and B can both communicate with C does not mean that A and B can communicate with each other. (A and B being the players, C being the server, in this case.)
and the ds's are just peers.
Both DS and PSP versions of WoW2 are peer to peer games. The server is only involved in matchmaking and ranking.
Meaning the server knows who DC'ed first.
Not possible.
I'm not entirely sure how their ranking code works with disconnects, but i can't see any reason why they can't penalize and reward as due.
We can't penalize/reward as due because it's not possible to detect what is due and what isn't.
The biggest problem is that the client does a hard abort when it knows the other opponent isn't playing, so you don't get to see your rank update, even if it already has.
It's a bummer. All you can do is play those who behave well and play well yourself, i.e., surrender when you wish to give up a game.
Of course, if i'm wrong about the part about t17 being the server, and the ds are just peers ,then i'm compleatly wrong, but i'm pretty sure thats how t17 wrote it.
DS = Two Tribes, PSP = Team 17.
DS = Quit unavailable (Nintendo requirement), PSP = Quit Available.
It's quite obvious that the DS folks are exercised by the lack of a quit option just as much as the PSP folks are exercised by the availability of a quit option...
Make what you will of that.
Linewalker
25 Sep 2007, 21:00
Lol never been described as mythical before...
Moiz, I'll play anytime, but I'm normally on ranked. I am on most nights though, so I'll look out for you.
Ren, story's about great wins errrrrr.....actually after last night I can probably tell you about some great losses! Was a good night for the top 10 last night, we all kept swapping places. Think I dropped to 4th after a few bad losses, than fought my way back up to 1st. I reckon mikes awsome was probably pretty p***** off, he whipped me good then still ended up 4th.
As for tips, the only thing I can say is don't just stick to one style, just take each turn as it comes. And try and think about what you'll be doing in your turn while your opponent takes his. Oh yeah, and don't D/C if you can help it!!!
Actually, there are probably a lot of people more qualified than me to talk tactics, so I'll leave it to them.
Good luck guys, see you in the top 10.
well, my team "GEN X" is only ranked about 130, but I had a really funny win on the weekend. On the space level with low gravity, my opponent teleported his last worm to the far right lip of the map, fully expecting me to not be able to get near him. (we have no jets or ropes left).
On my very next turn, with the strong right hand wind, i parachuted the ENTIRE level, he was almost a full screen above me, I double jumped to do a backflip - getting awesome height due to low grav, fire punched which got double height with low grav, and somehow it hit twice for 60 damage! I think it was something to do with the low grav, but i hit him at nearly the top of the punch,, and he fell into the punch AGAIN. he was on about 50 life and died.
Wish i had a reply =)
I'm sure others have had some funy stuff happen, but so far, the low grav on the space level makes for some really funny deaths.
quakerworm
26 Sep 2007, 02:10
Not possible in a two player game.
i asked before, but i didn't get an answer. would it really be too much overhead to require that both players stay tcp connected to the name server during the actual match.
by name server, in this context, i mean the server that is responsible for tracking ip addresses of players who are trying to host or join a game before the game actually starts.
parsley
26 Sep 2007, 10:12
i asked before, but i didn't get an answer. would it really be too much overhead to require that both players stay tcp connected to the name server during the actual match.
by name server, in this context, i mean the server that is responsible for tracking ip addresses of players who are trying to host or join a game before the game actually starts.
No overhead at all, but we can't require it: how can we force the players to do anything? Further more, it doesn't achieve very much at all.
We don't, but that's beside the point. Who hosts the servers is irrelevant: that A and B can both communicate with C does not mean that A and B can communicate with each other. (A and B being the players, C being the server, in this case.)
Both DS and PSP versions of WoW2 are peer to peer games. The server is only involved in matchmaking and ranking.
Heres what doesn't quite add up for me then. Perhaps you can enlighten me parsley?
For a peer to peer connection to occur, one of the peers would need to have an active connection. (requiring an inbound pin-holed port or use of the DMZ). If neither have such a port active, communication between the devices is Impossible
Since the DS's only method of network access is via an access point, all DS connections will inherently be passive, unless the user manually forces the wifi access point to direct traffic to the DS via pin-holing, or via use of the DMZ.
I'm willing to bet that 90%+ of all DS owners don't bother doing such a technical feat.
I've reread through the Nintendo WiFi manual and the worms ds manual, and there is no mention of inbound port mapping. A requirement for peer to peer networking. Google returned no results either for details on how to pinhole your DS.
I can only assume from those details that the DS was never designed to allow inbound network access from the internet, which is a smart move from Nintendo, considering how frustrating and technical it is to pinhole a device.
So given that fact, i can only assume your system is one of the below:
1. A peer to peer system, that somehow defies my knowledge of the limits of TCP/UDP networking.
2. There are users out there who have bothered to make their DS connect active, and there the only ones that i (and the population at large) can connect to.
3. The DS is a client online, and there is a server out there. This wouldn't surprise me since i would expect 1 server could easily host a couple hundred concurrent games from the DS, a thousand more if it doesn't actually bother to verify that the data is legitimate, and just passes the data betwee clients.[/quote]
4. You have a gateway system, which is technically also a server, but your not calling it a server. It would have the potential of working out who disconnected first.
The only other possibility, is the use of UPnP to establish open ports dynamically, but i've seen no mention of its use from nintendowifi
I would be interested to know how worms manages its networking...
Wormetti
26 Sep 2007, 17:22
NAT traversal magic (http://en.wikipedia.org/wiki/UDP_hole_punching). It's not perfect but it works in more cases than UPnP.
NAT traversal magic (http://en.wikipedia.org/wiki/UDP_hole_punching). It's not perfect but it works in more cases than UPnP.
Yup, guess thats how it works.
Shame really because if that is the case, which it appears, there really isn't a thing the can to do prove a disconnect.
It would be nice if they implemented a punish feature, where those that have abnormal dc rates, are assumed to have lost when they DC
quakerworm
27 Sep 2007, 02:14
No overhead at all, but we can't require it: how can we force the players to do anything? Further more, it doesn't achieve very much at all.
ah, but it does achieve a lot. sure, when a true disconnect occurs, you cannot do anything about it. but say one of the players quits (or physically shuts off wifi, or resets the system, etc). that will disconnect the said player from the game and from the name server (or at least, you could code it to do exactly that when you hit quit, requiring a name server re-connect to go back to the lobby.) the player who did not quit remains connected to the name server, and only loses connection to the other player. now your name server knows which of the two players actually quit. hence, only that player loses game completion percentage, cheating is now punished (if not severely), and you get a lot of happy wormers.
Eyedunno
27 Sep 2007, 08:16
Yup, guess thats how it works.
Shame really because if that is the case, which it appears, there really isn't a thing the can to do prove a disconnect.
It would be nice if they implemented a punish feature, where those that have abnormal dc rates, are assumed to have lost when they DC
It seems to me that a better way to do it might be to have a report option on disconnect that reports the combined health of both teams.
If the discrepancy between the two teams is above a certain ratio and in favor of the person who reported the disconnection, then it counts against the other person, as compared to their total disconnections number, i.e., if it looks like they're disconnecting a lot more when they're losing, then they get some kind of punishment, either a harmless message declaring them a spineless coward, or eventually an assumption that they've lost every time they have a reported disconnection where they were losing.
Sorry for how longwinded that idea is, but it came about simply because of how often I get pathetic cowards.
parsley
27 Sep 2007, 12:09
ah, but it does achieve a lot. sure, when a true disconnect occurs, you cannot do anything about it. but say one of the players quits (or physically shuts off wifi, or resets the system, etc). that will disconnect the said player from the game and from the name server (or at least, you could code it to do exactly that when you hit quit, requiring a name server re-connect to go back to the lobby.) the player who did not quit remains connected to the name server, and only loses connection to the other player. now your name server knows which of the two players actually quit. hence, only that player loses game completion percentage, cheating is now punished (if not severely), and you get a lot of happy wormers.
All it does is forces users to deploy a different (and more efficient) attack vector.
spaghettyO
27 Sep 2007, 14:46
I have had a fair few disconnects. It says ive had about 20 on my stats but I can only remember 3 times where I actually did (game crashed twice and wifi dropped once)
You do drop rank when it disconnects though, both of you. People do it way too much on the last shot. Its really annoying. Especially when its been an epic battle.
It happens to me a fair bit after a couple of turns too and they realise I can actually play. Make a couple of decent shots and they disconnect. Bah.
My rank drops more form disconnects than losses, which is annoying.
spaghettyO
27 Sep 2007, 14:57
and it just happened again! Bah!
i kno how tht is spaghettyo >_<
frosty
spaghettyO
27 Sep 2007, 16:18
And again :(
Won first game, one shot off winning second and they disconnected and my rank went from 8th overall to 12th. Dropped out the top ten cos of someone cutting off :(
right next to me now, but i got bumped out cuz of ark...curse you ark, i have yet to beat you
parsley
27 Sep 2007, 18:04
It seems to me that a better way to do it might be to have a report option on disconnect that reports the combined health of both teams.
If the discrepancy between the two teams is above a certain ratio and in favor of the person who reported the disconnection, then it counts against the other person, as compared to their total disconnections number, i.e., if it looks like they're disconnecting a lot more when they're losing, then they get some kind of punishment, either a harmless message declaring them a spineless coward, or eventually an assumption that they've lost every time they have a reported disconnection where they were losing.
Sorry for how longwinded that idea is, but it came about simply because of how often I get pathetic cowards.
Okay, then every time I'm in a winning position, I'm going to activate my firewall and get my machine to report you as a spineless coward.
spaghettyO
27 Sep 2007, 19:21
I just played 3 ranked games, ended up against a team named 'Boss team" every time. And everytime they disconnected on me :(
quakerworm
27 Sep 2007, 19:31
All it does is forces users to deploy a different (and more efficient) attack vector.
true. however, compare the number of people who can flip a wifi switch on their system to the number of people who can selectively block traffic through their wireless router. i can guarantee that most people who cheat now do not belong to the later category. i agree, the problem does not go away completely, but it dramatically diminishes in magnitude.
right now, enough people cheat this way to make percentages useless. if only one in a thousand will cheat this way, the completion percentage will become a good way to determine who stays through the game. together, using completion percentage and name server connection check solves the problem almost entirely.
parsley
27 Sep 2007, 21:02
true. however, compare the number of people who can flip a wifi switch on their system to the number of people who can selectively block traffic through their wireless router.
This is the traditional security-through-obscurity argument: mistaking the number of people who know how to do it (now) for the number of people who can do it (ever).
So, let us compare....
Deep thinks....
Having compared them, the number of people who can flip a wifi switch on their system is roughly equal to the number of people who can selectively block traffic through their wireless router.
Why? Someone posted a script on t'intertube.
i agree, the problem does not go away completely, but it dramatically diminishes in magnitude.
I will never implement a system (unless ordered to, that is) as you've described because it doesn't improve the situation, it makes it worse by injecting an appalling, trivial cheat vector.
It is a bad suggestion.
right now, enough people cheat this way to make percentages useless. if only one in a thousand will cheat this way, the completion percentage will become a good way to determine who stays through the game.
This is, yet again, Fact-by-Fiat. What happens when there's a non-zero preference for playing players with a completion rate of greater than n? For every non-zero n, bad behaviour gets shunned. It's only how fast and how bad that changes.
together, using completion percentage and name server connection check solves the problem almost entirely.
NO! It injects an appalling, trivial cheat vector.
Current solution:
I'm losing and force you and I to be marked as a quitter.
Your situation:
Either
....I'm losing so I force you to be marked as a quitter and I get away Scot free.
Or
....Either
........There's no way to decide what's happened so I get away Scot free.
....Or
........You and I are marked as a quitter.
This is not an improvement.
-------------------------------
Fundamentally, this is a psychology problem, not a technology problem. Certainly, we can apply technology to gain psychological leverage (completion rates, for example), but we cannot solve it with technology.
parsley
27 Sep 2007, 21:57
For a peer to peer connection to occur, one of the peers would need to have an active connection... [deletia]
You're not connected directly to the other machine. You're connected to your router which is connected to your ISP which is connected through zero or more backbone routers that are connected to their ISP which is connected to their router which is connected to their machine. Whew!
There can be tens or even hundreds of 'jumps' between your machine and your opponent's machine, regardless of P2P/C~S.
It is possible to tell (vaguely) which link failed in this sequence but it's entirely impossible to assign blame. (This is made up, but, say, the AT&T router X715MDA[Michegan] failed... is that your fault or mine?)
Therefore, it's not possible to decide, in any reasonable way which party is at fault.
quakerworm
28 Sep 2007, 02:09
NO! It injects an appalling, trivial cheat vector.
you already have a trivial way to cheat, so i'm not sure how you make it worse by making cheating harder to do. at very least, this will require the potential cheater to look up the instructions on blocking ports on a router, which is different for every router.
Current solution:
I'm losing and force you and I to be marked as a quitter.
Your situation:
Either
....I'm losing so I force you to be marked as a quitter and I get away Scot free.
Or
....Either
........There's no way to decide what's happened so I get away Scot free.
....Or
........You and I are marked as a quitter.
again, that's only if the guilty party blocks the port. in that case both players remain connected to the name server, and since the name server cannot decide who quit, it records a quit for both players, which is exactly the same as it is now. it does not make cheating more beneficial. it does make it harder. ergo, fewer people will cheat. ergo, that system is better than current system.
parsley
28 Sep 2007, 10:37
It does not make it harder, it makes it easier AND potentially cheatable.
It is not a good suggestion. Give it up, please.
Eyedunno
28 Sep 2007, 11:08
Okay, then every time I'm in a winning position, I'm going to activate my firewall and get my machine to report you as a spineless coward.
Hey, I didn't say it was perfect, but I did suggest that reported disconnections be judged against total disconnections (based on some arbitrary ratio), and I also didn't suggest automatically giving a win to the person who did the reporting, so it would seem to require deliberate, focused maliciousness for one person to do this to any one other person. And suspicious reporting practices could also be monitored.
Or heck, how about a simple, client-side username ban feature? That alone would help a lot, at least in dealing with repeat offenders.
parsley
28 Sep 2007, 11:17
Or heck, how about a simple, client-side username ban feature? That alone would help a lot, at least in dealing with repeat offenders.
Bit confused by this: do you mean the block list?
just make the game detect if wifi is turned on while your playing online... if it gets turned off then it will (next time your online) update your ranks with a loss... for the person that DIDNT turn off wifi... he would be rewarded with a win.
if you use the quit or surrender option then it automaticaly contacts the other player before quitting and tells them its quitting. thus giving them the win.
Wormetti
28 Sep 2007, 11:44
Bit confused by this: do you mean the block list?
Eyedunno was referring to the DS version, it doesn't have a block list.
parsley
28 Sep 2007, 11:49
Thanks. Got a bit confusticated there, I did.
quakerworm
28 Sep 2007, 14:46
It does not make it harder, it makes it easier AND potentially cheatable.
i'm not sure if you are simply not thinking it through or just trying really hard to ignore it.
how is blocking a particular port on the router easier than flipping a wifi switch or hitting a quit option in the game? it is not easier. it is harder. by how much may be arguable, but it is harder without a doubt.
and there is no more potential for cheating. if you do find a way to block the particular port, the worst case scenario is that everything works the same way it works now. you can cheat now by quitting, and both sides take a hit on quit percentage. you will be able to cheat then by causing a disconnect, and both sides will take a hit to quit percentage. there is no scenario when it is worse.
let me recap. you need to perform a more complicated action to "quit" without being detected as a quitter, and if you do manage it, in the worst case scenario, you get what happens in the status quo.
now, if you are simply refusing to listen on principle, just say so. don't give me any of, "it will be easier to cheat" with nothing to actually support that. this isn't quantum gravity. if you think there is a new way to cheat, you should be able to explain it.
parsley
28 Sep 2007, 17:46
This is the last I'm going to post on the subject.
Action 1:
Flip WLAN switch off and on
Reconnect to Wormnet
...30 seconds later...
Continue playing
Action 2:
Press 'return' on your PC.
...15 seconds later...
Continue playing
Which is easier?
quakerworm
29 Sep 2007, 21:32
you can do that now. people chose to flip the switch. why? because that's what is easier to do.
but even to get to that point you first need to figure out how your router works. not everyone out there even knows the password to log into their router. i've tried to help people to set up their port forwarding, and it's a pain in the rear to explain it to some people. then there are a lot of people who connect through someone else's wifi in the first place. it might not be any more difficult for some people to cheat this way, but it will be a hell of a lot more difficult for a lot of people out there.
even if you eliminate some percentage of people, it is a better system. i'm willing to bet that it would eliminate most of the cheaters, but that's arguable. the fact that it will not make it any worse is a certainty.
feel free not to reply, but if you don't understand something like this, i really have no more questions left.
vBulletin® v3.8.6, Copyright ©2000-2013, Jelsoft Enterprises Ltd.