[Poll]After the hitbox....

Things seem pretty good overall, but there are a couple of players in Miller who have been blatantly cheating for months now without a ban. TR's most notorious cheater has still been able to accurately headshot targets over a long distance with the MCG(which I have witnessed first hand...) while also being able to quickly shoot down even ESFs with it. It's that blatant and obvious, but he has been rampant for months.
I know nameshaming is not allowed in here, but I trust that almost all active members of Miller know who I'm talking about. However, I think that DGC has done progress with hitbox exploiting on a larger scale while there's still a few stinky cracks left to fix

Not going to bother quoting posts or I'd end up quoting half this thread, but...

Although this thread probably could've waited a few more days, if it continues getting attention for at least the next month it could really reveal how significant the hitbox fix was.

Right now? Anyone saying "I think I'm doing better therefore I know with certainty cheaters were a huge problem" has no good reason to believe that. There are too many factors involved with one person. That's why this thread is nice; it's asking for the stats of other players as well.

But just comparing a lot of objective stats (OP does encourage using a stat tracker, after all) is not enough for such a short time period. Everyone's stats improve significantly all of a sudden? Doesn't necessarily mean cheaters were keeping it down; maybe recent server maintenance had an impact as well or something else, you know? That's why it's important to see if legit players' performance has changed significantly over a long period of time - where servers have gone a long time without maintenance but the hitboxes are still fixed (for that one example of server maintenance.)

The data provided in this thread so far all points to one conclusion: something has affected players' performance. Maybe something didn't change; there's a chance a handful of people have just performed significantly better recently and they're the only ones talking about it, but that doesn't magically make it "confirmation bias" and the likelihood of that drops every time someone adds their input.

In fact, since the goal of this thread is to get more data (and, I hope, data over a long period of time), jumping straight to saying it's a thread about confirmation bias is laughable. As far as I can tell, the entire point of this thread is to avoid that, and I wonder if it would be "confirmation bias" if nobody noticed any difference in performance.

Cobalt's NC got a kick in the balls 4 sure.
Dem polaks now dont even pose a threat.

VS is good overall (lots of Russians there with +2000 hrs)

I think that is the largest problem facing this thread - it has a specific goal but it's collecting general, inspecific data that it desires to attribute as proof. What we are looking for is not reachable with any certainty what you want to start with. It would require a long series of transforms to get to any end-goal. A change in individual experience does not directly point to the bugfix solving a problem nor does a lack of change in experience point to the original lack of a problem.

If you want something to look for: find good players whose known (notorious) characters have been banned. We've been told that "people who were exploiting were banned." If there was a wave of bans recently, it should be observable.

I have doubts that DBG was even capable of detecting any players that might have modified the flat file. (Okay, that might not be true. If the asset validation routine logged what files which user had to redownload, then it might have been able to detect incorrect parameters, though that depends whether the changed file is marked as a core file and whether general corruptions would trigger an automatic asset repair on Launchpad startup.) I don't believe the swiftness of their solution helps their story that they were on top of this. They might have been vaguely aware of it being possible, either due to their own knowledge of the engine or by user feedback, but, if they never banned anyone for it, they might have not even known that it was outside the realm of possibility (and was actualyl happening). It'd be easy to catch such low hanging fruit and offer them up as proof of "see? we're actually doing stuff" without having to explain the "how they were cheating" to cover against the difficult cheating they can only observe statistically and require further investigation.

Additionally, I don't have much confidence in the goal. It desires to claim "look at how serious the hitbox problem was" but there is nothing but self-serving applications. We already know DBG's lackadaisical attitude to minor bug fixes and exploitation is endemic so it is not as if this kind of conclusion can be forged into a blade to dangle over them. If we do pursue this line of inquiry, what have we to gain from the revelation?

Yes, that is exactly why I made this thread.
I noticed that my "Death-by-headshot" ratio went down significantly and basically wondered: "Coincidence or....?"

My own datasample only amounts to about 100-150 deaths after removing explosive and vehicle deaths, so thats not really a statistically relevant number. So I need to get more.... even better if the data is not from a single server anymore.

"Certainty" can never be reached without actual proof. So unless cheaters (if there are any ) come here and confess....
But I'm trying to get the next best thing: "Evidence".

But seriously I don't think looking for banned accounts will indicate anything. I'd rather have people tell me how their performance changed since the event in question than rely on anything SOE/DGB provides.
Their track record shows they will tell us what they think we want to hear, the prime example for me still being the so-called "lethality nerf".
My trust in that direction is so far gone, if they claimed that their employees have trouble breathing in the void of space I wouldn't believe it without any means to check for myself.
....
Maybe not THAT far, but pretty darn close.

PS: Something like "server maintenance" can imo not be a cause for drastically changed performance of players.
Even if a maintenance influences the performance of the game and thus of the players it should do so for all players equally.

I was playing the day after the patch and dare I say it felt like I was doing really well for some reason.
Normally it would be:
1) take out one guy
2) his buddy turns and 'tink tink tink' *I'm dead*

But it was more like crouching behind crates or getting back around a corner and I was still alive. Maybe I was imagining things but damn if it felt like I was on fire because I wasn't getting worked over constantly.

I wonder how many peeps think I was part of the ban list, haven't had time to log in and test out a cheater-free environment on account of finals. Will get back on this thread tomorrow when I get on that

Good thing they did this change think about how many people were cheating and how long they were doing it

Confirmation bias is not pseudoscience in any sense of the word. It has been found in many peer-reviewed studies. Just because a widely accepted scientific concept doesn't fit with your worldview doesn't make it pseudoscience.

Furthermore, there could be confirmation bias here if people are posting 'my stats went up because the hitbox bug was squashed!' but actually their stats are the same as usual (which could even be the continuation of an upward trend). You have not shown how you know this to not be the case.

There is also very likely to be another bias at work: selection bias.

People who believe that there were lots of hitbox abusers and then find their stats have gone up are likely to post in this thread, because they get to brag about their performance and about the 'confirmation' of their pet theory. But people who believe that there were lots of hitbox abusers and then find their stats have gone down/stayed the same are less likely to post, because it is embarrassing to post that you're doing badly, especially when this offers evidence against your pet theory.

Hence, your 'wider datasample to either confirm or invalidate a hypothesis' is likely to be skewed in favour of the hypothesis due to confirmation bias and selection bias. Of course, this shouldn't surprise anybody - this is a thread in a video game forum, not a controlled experiment.

Personally I've noticed a steep drop in being 4 shot head shot from 200+m away by and LMG. So I would say it made at least SOME differance.

I'm just going to say that the Oracle of Death tracks headshots both as a percent of all shots and as total kills.

If you want actual data, take a look at that. Though it might also be that the hackers were too small of a population to account for much.

Or that the patch didn't work.

Why you guys all have to say how good you are?

I'm depressed now

I'm not sure this exploit was entirely that wide spread... it took someone 3 years to get around to mentioning it. If it were really wide spread, some other good samaritan would have noticed and shared the knowledge at an earlier time.

I'm thinking this thread's posters are experiencing a bit of a placebo effect. Yeah, there are fewer questionable head shot-only deaths by someone moving like an inflatable noodle guy seen at car sale lots at 50m+, but ultimately, either few people were actually messing with the file for an advantage hence why no one posting here seems to have known about it, or very few were banned for it thus there may be people here faking they didn't know about it. Forum users are a very small sampling, but we're not an isolated sampling. It's going to be hard for me to believe that somehow the majority of forum regulars were oblivious to this exploit while somehow many players knew about it.

You do realize you are doing that yourself right now? You might want to look up actual statistics like Chingles suggested instead of relying on probably skewed personal opinion.

"Actual Statistics" is nothing more then a large quantity of "Skewed" personal data that eventually may show a trend that "hey, a lot of people's data is skewed in this direction, maybe there's something to this".

So no.

Well unless someone is actively messing with the oracle of death it will give you actual, as in not skewed, data. Here anybody can simply make up whatever he wants and theres no proof if this data is actually correct. The only thing this poll shows is popular opinion.

so yes.

That requires the assumption that people are generally assumptive, stupid and dishonest, which generally is a sentiment that I prefer to avoid.

That being said, it is true that the Oracle will provide raw statistical data, which is less likely to have been altered.

That also being said, the Oracle will only provide "raw" statistical data, which could also lead to incorrect assumptions. The individual's ability to filter out deaths irrelevant to the poll is a valuable asset.


So in the event that people provide honest assessments of their performance, I'd consider their input to be of greater value then the charts on the Oracle. But sadly, as much as I dislike grouping the entire human race together as idiots, they generally are, so that also has to be taken with a grain of salt.


Perhaps it'll all eventually simply boil down to the individual. Perhaps some people will see significant performance improvements, even if only through a placebo effect.

I was totally unaware the Oracle tracked headshot quotas.
At least in the dozen weapons I checked there is nohing unusual in the headshot percentage.

Although a couple people, including me, definitely noticed a change in their stats its doesn't seem to be enough to indicate anything worse than a temporary deviance from the norm.
Of course it doesn't PROVE anything, but at least it indicates that the problem can't have been that widespread or we would notice more and bigger changes. Also there would propably be more people noticig this and posting their experiences.

I noticed that NC population dropped to 26% on prime time. And 16% on alert continents. But that might have other reasons =\
Getting killed by multiple headshots. No observable change.