Statistics of Survival Games 2.0

[Zinc // Akash754]

Wow nice!

 

[Girls]

See everyone, MCSG isn't as bad as everyone makes it up to be!

 

[Illumiigel]

How can this be used as evidence at all? He can't observe every player at all times, and even if he could, how could he accurately determine whether each player was hacking or not?

To anyone who thinks this 'data' can actually be used as evidence in an argument, you need to learn to actually look at the methodology before looking at the results.

 

[SnoopSean]

How can this be used as evidence at all? He can't observe every player at all times, and even if he could, how could he accurately determine whether each player was hacking or not?

To anyone who thinks this 'data' can actually be used as evidence in an argument, you need to learn to actually look at the methodology before looking at the results.

I disagree

I believe that his methodology was pretty dang accurate.
I mean he must've been teleporting all over the place and it's not like you need to spectate millions of people at once, it's only 24 players and most die in the beginning chaos.
It's as accurate as you can get.

 

[Illumiigel]

I disagree

I believe that his methodology was pretty dang accurate.
I mean he must've been teleporting all over the place and it's not like you need to spectate millions of people at once, it's only 24 players and most die in the beginning chaos.
It's as accurate as you can get.

You ever done Overwatch cases on CS:GO? You watch a person playing an entire game and you might catch only catch one second of footage that distinguishes hackers from just being good players. It's a very similar story with MC. Hackers typically don't display consistent behaviour which shows beyond doubt that they're hacking, rather you can catch instances where they exhibit behaviour that wouldn't be possible for a human player. The more people you watch, the less hackers you're able to catch as a percentage of the whole.

You brought up the point of most people dying at the begin; that contributes to the reliability of the data. If half the game dies before you can determine if they're hackers or not, essentially half of your data is lost.

 

[Fireworks]

You ever done Overwatch cases on CS:GO? You watch a person playing an entire game and you might catch only catch one second of footage that distinguishes hackers from just being good players. It's a very similar story with MC. Hackers typically don't display consistent behaviour which shows beyond doubt that they're hacking, rather you can catch instances where they exhibit behaviour that wouldn't be possible for a human player. The more people you watch, the less hackers you're able to catch as a percentage of the whole.

You brought up the point of most people dying at the begin; that contributes to the reliability of the data. If half the game dies before you can determine if they're hackers or not, essentially half of your data is lost.

As I said in the main post, the data may not be 100% accurate, just from how difficult it is to keep track of just about everything. As SnoopSean earlier in regards to the team data, I was teleporting around constantly watching all the players. While some players do die immediately at corn, some of which could possibly be part of a team, weren't truly counted, for the team was never formed and never posed a threat to the solo players, thus not a team. The data represents those who had an formed team together, and posing that true threat a team gives, not theoretical teams.

As for hackers, it's a similar story. I have to rely on my ~3 years of experience playing to determine the difference between a hacker and a skilled player. I only marked down hackers in which I truly 100% believed that they were hacking, none that I was just suspicious about, which those counted weren't trying to hide their hacks at all if I recall correctly. An experienced mod could of gather more accurate data than myself, but this is the most accurate I could get to my abilities.

I really want to stress that this data is not 100% accurate, but pretty dang close to the true numbers.

 

[Illumiigel]

As I said in the main post, the data may not be 100% accurate, just from how difficult it is to keep track of just about everything. As SnoopSean earlier in regards to the team data, I was teleporting around constantly watching all the players. While some players do die immediately at corn, some of which could possibly be part of a team, weren't truly counted, for the team was never formed and never posed a threat to the solo players, thus not a team. The data represents those who had an formed team together, and posing that true threat a team gives, not theoretical teams.

As for hackers, it's a similar story. I have to rely on my ~3 years of experience playing to determine the difference between a hacker and a skilled player. I only marked down hackers in which I truly 100% believed that they were hacking, none that I was just suspicious about, which those counted weren't trying to hide their hacks at all if I recall correctly. An experienced mod could of gather more accurate data than myself, but this is the most accurate I could get to my abilities.

I really want to stress that this data is not 100% accurate, but pretty dang close to the true numbers.

It's not really to do with how you collected the data; any data collection methods that requires this much human observation and judgement will not produce incredibly reliable results. What I meant is that there's not much you could have done to increase the reliability of your findings; studying something like this probably wasn't the best idea when you're limited to only observing games.

However well you tried to monitor players, you won't be able to pick out every hacker. If you're trying to monitor 10-15 different people, you're not going to be able to observe the vast majority of any player's game. Sure, you could pick out the bulk of the people using obvious PvP hacks by only focusing on combat, but there's a huge plethora of functions most hack clients have that don't directly affect combat.

You state that your data is pretty close to the actual numbers. What actual numbers? Where have you got this reference data from? You can't make a statement that solid unless you have evidence to reinforce it.

 

[Fireworks]

It's not really to do with how you collected the data; any data collection methods that requires this much human observation and judgement will not produce incredibly reliable results. What I meant is that there's not much you could have done to increase the reliability of your findings; studying something like this probably wasn't the best idea when you're limited to only observing games.

However well you tried to monitor players, you won't be able to pick out every hacker. If you're trying to monitor 10-15 different people, you're not going to be able to observe the vast majority of any player's game. Sure, you could pick out the bulk of the people using obvious PvP hacks by only focusing on combat, but there's a huge plethora of functions most hack clients have that don't directly affect combat.

You state that your data is pretty close to the actual numbers. What actual numbers? Where have you got this reference data from? You can't make a statement that solid unless you have evidence to reinforce it.

I understand what you're saying, and from what I'm seeing it's completely valid as well. There's just so many factors going into this that just simply observing a game doesn't make it completely valid data. As that is as much as I'm able to do, there's not straight forward way to go about making it more accurate, just as you stated.

Same goes with hackers, for those observed and marked down were those using obvious combat hacks. It's entirely possible others were using hacks that are completely unnoticeable from a spectator's eye, making that data just as unreliable.

I guess the best way to describe this data is that this is 50 games that I observed with the recorded amounts of hackers and teams that I personally think I saw, and could be completely different for someone else who conducts the same study.

As for the last statement, you're completely right. There is no actual numbers to relate this data to. The only outside values I have to relate to are the claims that other players make of what they believe to see. In comparison to what they've claimed and what I've seen from these games don't match up, so I personally believe that there is some sort of difference in either our definition of a team, what were seeing, or some unlisted explanation to the difference.

 

[Vanicle | Vanessa]

Bookmarked. I commend you for this and I hope more people see it. As someone that plays quite often and spectates random games on a daily basis, I have to say this seems very accurate with how I believed the data would look.

 

[Zeff]

I wonder if a donor affects there being a hacker or not. I mean some new players think players with different colored names are mods.

 

[Fireworks]

Bookmarked. I commend you for this and I hope more people see it. As someone that plays quite often and spectates random games on a daily basis, I have to say this seems very accurate with how I believed the data would look.

Thank you. Glad my hard work is paying off in some way.

I wonder if a donor affects there being a hacker or not. I mean some new players think players with different colored names are mods.

Interesting point. I didn't really test for that, since hackers wasn't the primary research point, but that would something interesting to note down if a future study was conducted.

 

[CultHD]

shows how bad the teaming situation is