The ultimate Defrag Question has been finally solved !

[w]

!oldtop strafeged2
sheeeeet. low upmove again!
e: i wasnt being cereal when i demanded further proophication.
low upmove and replay is proof enuff, init?
i really dont get ur point.

[q3a]

Ok, here's what we got.

A record was set in the db back then (he was logged in and we have the demo).
I created a replay script based on that demo, and it shows constant lowest upmove value possible in all the jumps and crouches, which is CHEATING. You can't have constant low upmove using any of the official clients.
I recorded a demo using that replay script, and dumped frame-by-frame coordinates from both w3sp's original demo and my demo recorded by replay script. The dumped coordinates from these 2 demos are absolutely identical, which shows that the replay script is accurate, and w3sp indeed used low upmove hack.

[wubwub]

posting to get my name out there

[lith]

uk represent, wassup wubwub

[sn8k]

but was w3sp gay?

[wubwub]

uk represent, wassup wubwub

nationscup when?

[esc?nebuLa]

I think he clearly had the capability to set #1 times without these small-time cheats, but for some reason was compelled to use them years ago. At times, purposefully cheating to take back a #1 spot from a more blatant cheater.

What I'd be more concerned about is deceptive and obfuscated cheating. Upmove is sort of deceptive, but I'm talking about fully-automated human-like botting, where it's not immediately evident after the first few demo views. (I highly suspect Camping_Gaz had/has the ability and knowledge to pull this off.) *tinfoil hat*

On another note, I've noticed that in my experience, some apparently legit demos periodically look like they've been recorded in timescale 1.3; often giving a "too fast" effect when viewing in timescale 1. (I won't mention names.) This could be due to mouse and accel settings, background programs, hardware, etc. Imo, it's tough to tell with the naked eye and requires a demo dump like q3a did. Online demos can occasionally provide odd anomalies and missing frames though.

That being said... Upmove, yaw, -128, multiple actions on a key/macros, 0ups scripts, mousewheeling, wallbugging, etc. (While viewed as insignificant by others) is still cheating and damaging to a reputation.

Keep posting more analysis please.

[lith]

uk represent, wassup wubwub

nationscup when?

whenever u want dawg, no competition tho rite

[dqopb]

I'm talking about fully-automated human-like botting

In my opinion, replaying human input is more dangerous than that. Because it is harder to spot/prove. You see, fully-automated botting may be found after a few times, or after certain analysis. Whereas recorded human actions, well, they are performed by a human in the 1st place.

[wubwub]

uk represent, wassup wubwub

nationscup when?

whenever u want dawg, no competition tho rite

i dont want to get bullied by zoot though