forum

Doubt about dragonhuman/sayonara-bye

posted
Total Posts
22
Topic Starter
Kanye West
Reddit user commityoursudoku ran graphical analyses on this user's replays and has some interesting things to say about sayonara-bye/DH. Here's what he posted:

Hey everyone, ever since I've seen dh's skill skyrocketing nonstop, I can't help but look into his plays. This is just to raise a little awareness of the game's "anti-cheat" and how it needs to be changed.

I'm not going to throw accusations without showing any reason why, so here's what I've looked into.

First I compared dh's DTHD play with SnowWhite's play on Luminous. http://puu.sh/a0vt5/23c15abd40.jpg

Small translation first: Using dh's side -
Notes Hit Early: 298 Notes Hit Late: 378 Miss Reason: Hit too early: 0 Hit too late: 0 Aim: 0 Slider Break: 0 Key hit stats: Left-click: 638 Right-click: 88

SnowWhite only missed 4 times near the end, but other than that it was a perfect play. On the other hand dragonhuman played a bit too perfectly and I think I'm right about it, he hits his notes later despite it being DT.

SnowWhite's ms skips are not as tight as dragonhuman's, and SnowWhite played perfectly, though it may look more human on SW's side, I'm still a little skeptical about it.

This was played by a bot with 5 intentional misses and set with higher ms offsets and less frequent/ more gradual ms shifts (in the 20ms range), though this is not reliable because it was used alongside osu's auto pilot: http://puu.sh/a0wvg/141d707301.png

Before I go any further, the anti-cheat is essentially a data packet sent over to the osu! server after a score submits, and all it does is show what processes you were running. You can essentially bypass this by renaming your process and putting it somewhere where another program would normally run from and use a file pumper to add empty data to said process for it to take more memory, thus making it look like a legit program rather than some cheat. Kriers talked about this on his ask.fm, though his account is deactivated, hopefully someone can vouch here.

Another suspicious play is dh's DTHD play on this map on Extra diff: https://osu.ppy.sh/b/313621
http://puu.sh/a0tZR/b6aff00b3d.png

His offsets are very far apart and his 100s are made randomly (ei. in the middle of a stream there'd be a 100 but then followed by a stream of 300s).
I've noticed a pattern in his other replays until his more recent plays, most of his note hits are hit late.

Here are some more examples, this time using one of Cookiezi's DT score on Bloody Night and dh's play on the same map https://osu.ppy.sh/b/143989&m=0:
Cookiezi: http://puu.sh/a0yoT/b2ab7d7f29.png dh: http://puu.sh/a0yz7/691d18c3b2.png

As you can see, dh's ms jumps are bigger than Cookiezi's making his unstable rate believable. His hits here are also made late, but you can see the early/late almost equalizing here.

This is another replay proving my point about his hits being late. http://puu.sh/a0AJw/3e78b2ed81.png https://osu.ppy.sh/s/90419

Though this pattern was broken when I tested his more recent ranks where his hits are done earlier. There's no late/early variation in between replays, that much I can say. But his ms jumps and quirks are still very questionable:
https://osu.ppy.sh/b/132608?m=0 http://puu.sh/a0BEU/4de0399e44.png
https://osu.ppy.sh/b/163812?m=0 http://puu.sh/a0BP6/45a195749a.png

Moderators, please do not delete this post. This is only for the sake of discussion. I've stated my opinion and have shown the facts above.
If dh is indeed legit, then I apologize.
I wanted to post here to get more discussion going and also to get commityoursudoku's work out here as well. Keep it civil and discuss the facts he presented please.

Link to his original post: http://www.reddit.com/r/osugame/comment ... not_legit/
Piine
Mother of analysis.....
Never thought the reddit osu! community was active ._.
buny
it's tom69 all over again

what with all these squiggly graphs and all.

This will get locked though because discussing hackers and questioning peppy's anticheat is not allowed.
Vuelo Eluko
if you could take Japan and Taiwan out too that'd be great

#USA2014OWC
#HoroWasRight

there's something eerie about his graphs though.
still he's probably legit. comparing to snowwhite is kind of silly, and also, reddit is awful.

and why does he keep going on about hits being late? of course they're going to be late more often than early at those speeds, and also i dont think he understands how hit windows work on streams at those kinds of BPM's if he thinks one 100 in a full stream of 300's is suspect.
Lach

Sulker wrote:

Mother of analysis.....
Never thought the reddit osu! community was active ._.
It is, sadly.
Pettanko
i honestly have no idea what the fuck those pictures mean, and even if they proved something they could easily be edited and we would be none the wiser
Tanzklaue
posting to say that i would use a cheatprogram to draw a penis or something of similiar immaturity with the squiggly lines.


people wouldn't even be mad.



also wondering if squiggly lines will go 3-0 against osu players.
xasuma
I find nothing abnormal on any of his 'suspicious' plays.
He probably has no idea what he is talking about tbh..
Rewben2

buny wrote:

it's tom69 all over again

what with all these squiggly graphs and all.

This will get locked though because discussing hackers and questioning peppy's anticheat is not allowed.
Mhm. Incoming lock.

Pettanko wrote:

i honestly have no idea what the fuck those pictures mean, and even if they proved something they could easily be edited and we would be none the wiser
I don't know much about the whole squiggly lines thing... But apparently it indicates that he cheated because the offsets (when he gets 100's) are seemingly random, not realistic. Like he gets 100's randomly in streams and then does the rest fine.

I understand what he's on about but I really don't think it's enough to call him a cheater over.
winber1
depends on how off the 100's in the middle of the 300's is. If it's like 50ms off from the other 300's, that's a little suspicious, but it seems we were given no information on that.

The way his cursor moves is also really awkward, but it reminds me of silynn. They both have really awkward shaky movement sometimes, but still hit streams and other notes nonetheless. Not really an indicator of anything, but thing is I don't remember dh always having this sort of movement... but I may be wrong and I'm too lazy to find other replays.

I've heard rumors of korean account sharing though
Rewben2

winber1 wrote:

depends on how off the 100's in the middle of the 300's is. If it's like 50ms off from the other 300's, that's a little suspicious, but it seems we were given no information on that.

The way his cursor moves is also really awkward, but it reminds me of silynn. They both have really awkward shaky movement sometimes, but still hit streams and other notes nonetheless. Not really an indicator of anything, but thing is I don't remember dh always having this sort of movement... but I may be wrong and I'm too lazy to find other replays.

I've heard rumors of korean account sharing though
Yeah, if the Squiggly Lines™ had axis and we could actually see the offset it would be... helpful.

As far as the cursor movement goes I think it's quite normal. WWW and Elysion sometimes have really shaky/weird aim but still hit everything, even more shaky than sayo i'd say.

Definitely some account sharing going on with koreans. Cookienia said on his ask.fm that he had scores set by reimu, I've seen shizuru log onto someone elses account as the end of a stream... Yeah. Probably more unknown cases of it too.
ballplay
only 1 thing...

Starz0r
holy shit just enjoy the game, no need for public accusations

I do understand everyone's doubts about dragonhuman but please just use the report button if it bothers you that much.
uzzi
My only question is: why bother? So what if he's a cheater? Not like it's going to change anything.
Purple
I have to ask, why do people get bitchy when hacking accusations take place? If anything, it should be taken as a compliment. It's not a "witch hunt", as these idiotic redditers like to call it, it's just expressing doubt about someone's abilities to get certain scores. If the accusation has merits, then people will believe it, and if it doesn't, then why care.

If the player cares so much about his reputation, he/she can go ahead and try to prove they're legit. If he/she doesn't, they can go ahead and continue playing because peppy wont ban people on popular demand. If the player cares about their reputation and decides to cheat and ban him/herself as a response, then I can only imagine such player is either a cheater or has some sort of brain disease.

To me, it seems impossible to prevent cheating in this game. I view these "hackusations" as a way to make up for the anti-cheat system's lack of robustness. So I view them in a good light, provided it's a decent discussion and people don't get moronic and overly salty about it.
Rewben2

- [ U z z I ] - wrote:

My only question is: why bother? So what if he's a cheater? Not like it's going to change anything.
I don't know if the op made the reddit thread with the intention of having something done about it. It's clearly not enough/good enough evidence. I guess it's just to spark some discussion about it, no real value/point to it. Can't say that everything that's talked about has a point to it.
haha5957
You all can suspect this but prior to make anything big ask him to upload his realtime play video once
Purple

- [ U z z I ] - wrote:

My only question is: why bother? So what if he's a cheater? Not like it's going to change anything.
In a competitive environment, people feel better about being 2nd place when they think 1st place is a cheater, and they feel even better if people share his/her belief
buny

- [ U z z I ] - wrote:

My only question is: why bother? So what if he's a cheater? Not like it's going to change anything.
So what if top players of other games are cheaters? Not like it's going to change anything.
So what if players cheat versus you in a game with a competitive background? Not like it's going to change anything.

People care because people play the game competitively. By not weeding out cheaters reflects an inaccurate ranking of said player. Hence why people do get angry that (proclaimed) cheaters aren't punished, and causing an outrage.

Keep in mind that such posts that I'm assuming "annoys" you, replying with such a similar mannered post merely adds to the fire and just starts an endless debate.
Secretpipe

Starz0r wrote:

holy shit just enjoy the game, no need for public accusations

I do understand everyone's doubts about dragonhuman but please just use the report button if it bothers you that much.
xsrsbsns
Subscribing to relevant thread
Ephemeral
None of this is founded in any objective fact and is just baseless shit-flinging and accusation. Investigating discrepancies in a person's play by measuring when they hit 100's compared to other players who hit 100's in similar spots is such an arbitrary criteria and subject to potentially hundreds of factors that I literally feel embarrassed for the person who suggested it may be a viable method of determining incidents of foul play. This literally goes down to factors as basic as muscle twitches or idiosyncracies during play - it is influenced by so many things that it is impossible discern anything solid from it.

Posts like the OP are the precise reason why we do not, and never will support the osu! subreddit community, as they are almost exclusively dedicated to creating baseless posts for the sake of controversy and satisfying their own desire to feel important by demeaning the efforts of other players.

Do not make threads of this nature in the future. Any suspicious play should be raised via a support ticket (http://osu.zendesk.com) or by sending an email to me (ephemeral@ppy.sh) regarding the issue.
Please sign in to reply.

New reply