-
Ever wanted an RSS feed of all your favorite gaming news sites? Go check out our new Gaming Headlines feed! Read more about it here.
Majority of 70 Blizzard Employees Strongly Support Company's Punishment Of HK Advocate & Player Censorship Of Sensitive Chinese Topics - Blind Survey
- Thread starter FliX
- Start date
OP
OP
This issue was one of the few that right and left wing folks agreed on, so calling Blind rightwing hardly means anything. That said, regardless of the reputation, there was 0 rigor in ensuring random selection, so a sample size of 70 does not fly here.
only 70 answered their poll. If you are directly involved in the company there could be more incentive to not answer the poll either way. You, as an individual person doesn't really know if there will never be a leak about your "anonymous" choice. So better, not answer.
But if you are, for example, working for Microsoft? And you get the question if Blizzard is evil? You mostly have no issues answering it.
I was originally on the side of it being a sufficient sample size, but after plugging it into a couple sample size calculators it needs to be be in the 200-300 range for it to have any sort of significant confidence level.
The algorithm scales really well with population size - a sample of a couple thousand people can accurately predict a population of millions, but at lower numbers the sample size % needs to be a lot higher to get an accurate picture of the population.
The algorithm scales really well with population size - a sample of a couple thousand people can accurately predict a population of millions, but at lower numbers the sample size % needs to be a lot higher to get an accurate picture of the population.
People certain that this isn't representative of Blizzard must have some insider information or something.
People being certain of anything from this poll don't understand how statistics work. This gives no reliable information. If it was 500+ employees then it would mean more. 70 is a paltry number, especially since 30-35% of them are on the right side of things. So only 45 people agree with Blizzard's actions, out of 70 who took the survey, out of nearly 5000 people in the company. Worthless data.
Sample size is too small to give us anything definitive. If you wanted to be technical more than 70 blizzard employees protested outside the building that first day. So there's a sample size equivalent for you.
in Brian Kibler's letter/response to the subject, didn't he also agree with the ban (at the time)? His reasoning, if I'm remembering correctly, was that it has always been Bizzard's policy that players never discuss politics. We all know why he was really banned, but if they had a policy in place well before this incident, then I can see it. Would they have banned him for discussing anything else political? Who knows. So in that aspect, he was OK with the ban. He was NOT ok with them stripping him of his earnings though, or anything else.
I wouldn't call 70 responses from a total company size of ~5k employees enough to draw this conclusion. Same goes for all the other companies -- 220 out of 60k for Facebook, for example, etc.
I'm not certain of anything. The poll just supports the interpretation that much of blizzard is fine with censorship on behalf of the chinese government. This isn't the only piece of evidence here.
It's incredibly unreliable. As someone said above, more people who work at Blizzard protested outside of Blizzard's HQ as a result of this fiasco than took this poll, never mind took the poll and agreed with their actions.
Edit: Again, 70 people answered the first question, only 45 agreed with Blizzard's actions. The title of the thread is wrong, and the data is, at best, unreliable based on sample size.
It's too small a sample size and, for obvious reasons, there's no regional or demographic information about the respondents. Disingenuous to say this is representative of even a slice of the company.
Nah? How about yeah?
Most polls for the elections have samples of 1 or 2 thousand people for populations of millions.
70 sample for a company of 5000 is more than enough.
The problem is not the quantity but the quality of that sample, that's why you can't extrapolate the results in this case
I didn't just read the title. Can you explain the math behind its being extremely unreliable?
Was gonna say the same thing. Now the title is objectively inaccurate, lol.
EDIT: It's now fixed.
Last edited:
70 employees respond to a random internet poll? Do we honestly feel this is a representative sample?
Likewise, I don't think I'd be communicating my opinion to 3rd parties, whether they stated you would be anoymous or not, about this topic if I valued my job and it went against the company lines.
If that's the way the company writ large actually feels? Then that's quite a departure from the rest of the tech industry in the US. (Though once again, even those stats are basically meaningless.)
Likewise, I don't think I'd be communicating my opinion to 3rd parties, whether they stated you would be anoymous or not, about this topic if I valued my job and it went against the company lines.
If that's the way the company writ large actually feels? Then that's quite a departure from the rest of the tech industry in the US. (Though once again, even those stats are basically meaningless.)
The sample size is not the issue here. The biased sampling is disqualifying in itself. Even if was 700 people it would not be a valid poll.
70 people is actually probably enough to have a decently representitive sample in terms of size.
But this poll is not even remotely scientific; it's biased to hell and back, and so, it lacks the rigour and quality needed to draw any reasonable conclusions.
But this poll is not even remotely scientific; it's biased to hell and back, and so, it lacks the rigour and quality needed to draw any reasonable conclusions.
Kinda justifies the full company boycott.
This isn't just a top-down problem, the whole damn company is rotten.
Free Hong Kong.
This isn't just a top-down problem, the whole damn company is rotten.
Free Hong Kong.
And all Pepsi employees love Pepsi over Coke, and they would never be seen drinking Coke. The company that employ them have no influence with their decisions at all. Also they really like their new key chains.
Last edited:
There's multiple things. So, 70 out of 5000 is 1.4%. That's simply the number of people who answered the first question (I'm sticking with that result simply because it had the most number of respondents). I'm rounding the numbers, but that means only 45 of those respondents agreed with Blizzard's actions, or less than 1% of Blizzard's active work force. Making broad statements about all or even most Blizzard employees based on an anonymous survey that only 1.4% of the company participated in (and 0.9% agreed with their actions) is tenuous. It may pass some tests of adequate sample size, and fail others.
There are also other factors in play. How closely scrutinized is this website at making sure a person actually works at said company? Where were these respondents from? Those from China who are anti-HK protests will of course agree with Blizzard's actions, and with such a low sample size could quite easily skew polling results. Are these high level executives who have financial conflicts with China, and thus would have their opinion skewed? Are there actually employees of Blizzard outside of China who agree with Blizzard's actions, either through personal conviction or through fear of retaliation if their opinion goes against the company? That would also skew results.
So we have a sample size of 70 people, 45 of which side with Blizzard, which represents 0.9% of the company's workforce. Then you add in all of the variables which can very easily and widely skew any results. This gives a rather large amount of uncertainty as to whether the results are actually reliable. I'm not saying the results are definitely wrong, but I'm also not saying the results are definitely correct. I'm saying the results shouldn't be considered at all (or at least be taken with a giant grain of salt) due to the large amount of variables and relatively small sample size given. We have information that Blizzard's own employees had a larger gathering than respondents to this poll who were protesting outside of their headquarters due to this decision.
Well, at the very least we can now stop saying it's only one or two bad apples at ActiBlizzard making these decisions. This marginal survey alone confirms it can potentially be pretty widespread.
yeah this
people continue to not know how stats work
the sample size isnt the problem
It's an interesting question. Blizzard had a clear and straight forward policy already in place. They followed it. They were justified in following it. I think the original punishment was way too severe, I am glad they backtracked on it a bit. But they have to enforce their own rules.
I firmly believe that even if the political message were the exact opposite, proclaiming how great mainland China is, etc... Blizzard would still have had the same reaction.
They're simply protecting their IP and attempting to keep the long term conversation and focus about the game and the tournament.
As a gamer and as an esports follower, this is an interesting time. Professional gamers, who are kids (or young adults) mostly, with little to no experience in the outside world, who spend the vast majority of their days grinding away on their game of choice, have a very strong platform to cast out anything they want to say. There is absolutely a ripple effect on those words. In this case, thankfully, the message was an objectively good one. There are unimaginable atrocities going on in Hong Kong right now. However, there are tons of examples in the past where pro gamers are just shooting themselves in the foot, even to the point of ruining their own careers. The esports industry as a whole is simply not mature enough to be well equipped to handle this sort of volatility, and being in the public eye, as pro-gamers, talent, staff, streamers, etc... are, is a trained skillset... it takes coaching.
It's a slippery slope to not try to maintain some sort of control over what might go on when one has all the cameras pointed at them. This is not the same as taking a knee during the national anthem. It's not the same as tweeting out your own opinion on your own free time. I don't know what the right answer is for this situation, but I do think Blizzard as a whole (maybe not Blizzard China) is merely setting an example to try and keep the game tournament about the game.
I firmly believe that even if the political message were the exact opposite, proclaiming how great mainland China is, etc... Blizzard would still have had the same reaction.
They're simply protecting their IP and attempting to keep the long term conversation and focus about the game and the tournament.
As a gamer and as an esports follower, this is an interesting time. Professional gamers, who are kids (or young adults) mostly, with little to no experience in the outside world, who spend the vast majority of their days grinding away on their game of choice, have a very strong platform to cast out anything they want to say. There is absolutely a ripple effect on those words. In this case, thankfully, the message was an objectively good one. There are unimaginable atrocities going on in Hong Kong right now. However, there are tons of examples in the past where pro gamers are just shooting themselves in the foot, even to the point of ruining their own careers. The esports industry as a whole is simply not mature enough to be well equipped to handle this sort of volatility, and being in the public eye, as pro-gamers, talent, staff, streamers, etc... are, is a trained skillset... it takes coaching.
It's a slippery slope to not try to maintain some sort of control over what might go on when one has all the cameras pointed at them. This is not the same as taking a knee during the national anthem. It's not the same as tweeting out your own opinion on your own free time. I don't know what the right answer is for this situation, but I do think Blizzard as a whole (maybe not Blizzard China) is merely setting an example to try and keep the game tournament about the game.
The sample size isn't that significant problem. If this was a representative random sample, the MoE would be a bit over 11% (95% confidence level), which is high but not high enough to make it completely worthless. There are other factors in play that likely make this unreliable, however.
Last edited:
Thanks for the post explaining why this whole thing is rediculous.
I can't believe this thread is still open.
Looks like you can get a margin of error of 8% with 64 people out of 4,700 and a confidence level of 80% (which I don't know how you'd calculate that here, but it was the lowest I could select on the calculator). The biggest problem I see is the fact that the sample was self-selecting.
Last edited:
Also, since there is no way to know if they actually ARE blizz employees...
Ah like I said, the sample size part is debatable, I suppose. Others here feel it is adequate, I'm a little more on the fence (and certainly there are people here who are far greater experts than me!). However, what cannot be ignored is the overall quality of who was sampled. By that, I mean where they're from, their background, position within the company, etc. Those factors *in addition to* a lower number of respondents make it fairly unreliable. Certainly proclaiming it represents all of the employees at the company is a huge leap.
To those saying a poll that represents 1% of the population is insufficient to represent the population itself, let's take U.S. political polling. There are about 330 million people in the United States. By that logic, political polls with less than 3.3 million people surveyed would not be valid.
Country-wide political polls tend to have a little smaller sample size (anything more than n = 100,000 is very unusual), but you don't see statisticians rioting. Having a representative, unbiased sample is substantially more important.
Country-wide political polls tend to have a little smaller sample size (anything more than n = 100,000 is very unusual), but you don't see statisticians rioting. Having a representative, unbiased sample is substantially more important.
I don't care about this issue, but I need to point out, as someone with a background in social research, that 70/4500 is a sample large enough to make estimations, contrary to what most posters seem to believe.
Why is it so different from other companies? Probably because those 70 people are constantly exposed to communications from their employer telling a version of the story that their employer is happy with.
Why is it so different from other companies? Probably because those 70 people are constantly exposed to communications from their employer telling a version of the story that their employer is happy with.
I don't see any reason to believe that that is what we have here.
Technically, it's only like 40-50 considering it was only 60% of them.
You're gonna get shit for this post but in my conversations with friends on the HS team, they've told me the same.
Of course not, any blizzard employee with half a brain wouldn't be honest about this for fear of losing their job.
