Replays suggestion to combat hackers and save good moments

I’m gonna make it simple and easy for you to understand. Recordings are different than actually being there in the server. Recordings have to render everything on the map at once aside to rendering a specific portion when you’re playing.

And how is the software gonna know when the suspect did his fight?
Common sense, please.

Okay, let me explain this in a way you can understand.
CSGO’s matches have a software that automatically records the match from start to finish in case anyone was hacking so that the viewer can be able to see what the suspect did and all that. Keeping in mind that the maps are small, the match itself is short but still completely filled with shooting and interaction. Even those cases, some are difficult to detect, unless the suspect is spinbotting.

The average length of a fortnite game is around 15 minutes which is fine in regards to recording size compared to the average 30 minutes of a CSGO compared to a survival game that can last for days and weeks before getting shut down or restarted.

Why dont you ask all the hackers in CSGO (pre-F2P era) that know about Overwatch but still hack, keeping in mind that the game costed money. People have their own reasons to hack, even if it costed money, so nothing will stop hackers whether a pay wall or anything else. And not all of them will be blatant

Could be, but still doesn’t guarentee 100% loyalty and maturity of all accounted staff members

Replays don’t “”“render”"" anything. Don’t use the term render. Render suggests that you mean churning out graphical frames, but in this context I assume (And hope) you don’t mean that.

In regards to player-made structures, yes, it has to analyze the entire map, but in reality it doesn’t /have/ to. Once someone places something the replay knows its there and has no reason to do anything to it. Unless someone destroys something is when a new change (or calculation) must be done to the replay. Therefore you could build a gigantic mansion and it wouldn’t do much on CPU calculations or other nonsense

In regards to players, all the replay must do is record X,Y,Z (coordinates of player) and Yaw,Pitch,Roll (where player is facing) and then what item they’re holding. XYZ and YPR is done every tick or frame, while other things that don’t need to be recorded every second (like holding out a weapon) are not. The only updates to the replay or extra calculations done per tick is when a player does an extra movement, like pulling out a gun, aiming, firing, a combination of the two, etc.

So essentially we’re saving the following information per player at each tick -

X: 512.54
Y: 224.01
Z: 40.89
Yaw: 180*
Pitch: 45*
Roll: 0*

Meanwhile, it is the same for structures, except this doesn’t need to be calculated every tick, only each time a structure is updated (placed, shot at, or destroyed). Player stuff technically also needs no changing unless theres an update

X: 512.54
Y: 224.01
Z: 40.89
Yaw: 180*
Pitch: 45*
Roll: 0*

Calculation (CPU resources) is minuscule because the server already gets this information to run the server, the only challenge with replays is it can’t just send all these tiny files to the hard drive over and over, so first it needs to pack them all up into a block in the RAM, then it can be comfortably flushed to the hard drive. So it takes a little RAM storage.

Your internet does the literal exact same shit. It can’t just willy-nilly aimlessly send bytes of information out, so it packs them up into “packets” before sending, so it can be comfortably flushed out to it’s destination.

Besides that, it just needs a small amount of disk space. I’m sure most server machines run on HDD’s over SSD’s, and even then, these files are still incredibly small - and despite BEING incredibly small, you could compress them further as these files are most likely fully uncompressed and raw. Just how PNG files are way small than TIFF.

Just because a server needs to keep track of all players doesn’t mean it will destroy your CPU, if anything, it takes a little bit of the RAM. This could also scale extraordinarily easy, so larger servers will be affected in a linear fashion, not an exponential one. I’m sure servers won’t get any bigger than ~100 players, so it would be no problem. Speaking of 100 players, Fortnite does the same thing with Replays for manual moderation and they have millions of matches a month. It just goes to show how easy it is to create replays, and how small they are.

How did Observatory software know when the suspect was fighting?

Map size would have zero affect on the calculation or RAM required for replays.

Of course they’re difficult to detect sometimes. Sometimes it will be that way. :man_shrugging:

Refer to my stuff above. Fortnite is a very fast paced game, and the tick-rate for replays are pretty high - so a lot more information is being taken in per second. Also, those replays (as discussed earlier) are completely raw and uncompressed. 7z could easily demolish file sizes. I could spout out other ways to shorten the file, or how short the file could get, but you get it.

Maybe because CS:GO is such a popular, over-hyped, extremely competitive game?

What needs to be so loyal and mature of a staff member to review a replay? Are you referring to local staff teams, not observatory?

If then, are you referring to them abusing knowledge from a replay to better themselves in-game? I doubt it would happen, and theres plenty of other, easier ways to go about abusing staff privileges to gain knowledge.

If referring to Observatory I don’t think I saw a single case where an entire jury of ~10 people all decided to be abusive shits and all vote one completely anonymous guy as guiltily.

Also, again, reminding everyone -

Why dont you just have death scenes. It would show 5-3 minutes before the persons death. It would highlight the attacker in red, victim in green, and any bystanders in blue.

This would kinda cover things like aimbot or ESP, but wouldn’t be super useful for investigating other claims (like map exploiters).

1 Like

everyone would exploit it. you go up to a base wall, and use it, you can freecam kinda past the wall into the room, and look for storage items etc. Making it so you couldn’t do that would be a waste of time, though, so yeah.

alright, what you do is make it so the camera doesn’t pick up other players buildings.
you make it so the camera can only view builds you placed
also, you make it so the camera can only view your storages and not other peoples storages

ehh, that’d work. but it would have players hovering in mid air, and doing things you couldn’t recognise because you can’t see the buildables.

How about a button that records the last five mins+ five mins after.

So if I were watching my own replay of me sieging an enemy base it would just be full of floating players that I am (for some reason) not firing at? :thinking: that would also increase the difficulty of making replays, as the client-side needs to differentiate between group buildables and enemy buildables.


I mean we could do that. Players and servers could choose to only record 5 minutes worth of replay (Like Nvidia shadowplay - basically it records 24/7 but when you press a button it saves the last 10m) which would save on space - albeit, you don’t really need to save on space because these files are minuscule in size and afterall, old replays that are over a month or perhaps a week old could be auto-deleted to save space.

Thankyou for the common sense. <3

1 Like

idk your the genius server creator
suggest something that would work


too late. already did.

1 Like

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.


Wait what’s the point of the automatic closure then?

Dumb AuTOmAtiC cLOsInG system that is still around. I had a friend who wanted to come and talk more about the subject and or suggestion his own ideas, but he never came.

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.