255 réponses à ce sujet

[ShepardTheHopeful]

SRobson wrote...

ShepardTheHopeful wrote...

Yeah but to play devil's advocate people complain online all the time. Everyone complains about everything this movie was bad etc. But this game created enough hate on the internet to actually revoke the ending in it's original quality. In order for that kind of response this must've been past normal internet trolling and hate. This must've hit a very high number of disatisfied customers.

If you're putting art in a museum and someone says "it sucks" big deal but if an entire community goes into the museum and looks at and a good majority of the people leave that's a need to change your stance or you won't be able to sell more art later on.

It's basic consumerism if there is no demand there's no supply. If they weren't having serious problems with PR then the game wouldn't have changed at all. But this is world wide hate towards a lot of people and it's hard to ignore. 

So there's gotta be someone wrong to a large number of people otherwise this would just be seen as average complaints not world wide news and demanded response. 

I do understand your point, honestly. And debating is cool n' all, but I'm terrible at it, and thats not the reason I jumped in this thread. All I wanted to do was add one teeny voice of support for Bioware, because they're getting ragged on a lot these past few weeks, and that must be pretty crappy. A thread titled 'THANK YOU, Bioware' seemed like the most appropriate =)

Oh no i'm not trying to deabte you or tear you apart Devil's advocate is my way of healing lol. But I'm just keeping hope but with very low expectations after the past month i've seen some very unfavorable colors in reaction to all of this from fans, and from EA and  Bioware. Too much ego and pride is involved it should just be about making a good game everyone can enjoy. If thousands of people couldn't look past 5 minutes of an otherwise fantastic game it's only fair to question those last 5 minutes. But it seems they don't even want to try taking a look it seems they just want to stick to this artistic integrity line and that's why i'm not.

I'm not mad about the ending i'm mad about their response. 

[Grunt420]

i not saying thank you till after i see the extended ending

[SLana]

Grunt420 wrote...i not saying thank you till after i see the extended ending

I think I can say "THANK YOU!!!" to people who made Tuchanka or Rannoch or other pieces of the game unforgettable, but to people who are responsibe for this, or the last 5-10 minutes of the game? Are they really so proud of all this? "General, you jeopardized your mission and your people, get the hell off my ship!" No matter how many really talented people are in BW, they all suffer from rage now along with those who decided to show us what they call "artistic integrity"...

[GuyIncognito]

Thank you very much Bioware.

[Bcoolpro]

kbct wrote...

Imperium Alpha wrote...

There is a 100% margin of error. I'm always right so now listen to me.

It's about a 20% margin of error for small sample, self-selecting polls.

Okay I'm not trying to open the same can of worms again, but I would appreciate some clarification.  From what I remember of my stats class a 4%+/- margin of error is at the upper end of what can be considered statistically significant.  If your 1 in 50 poll had a 20% margin of error then it is not a viable result.  Could you explain how it still has any predictive power when it falls so far outside the acceptable margin of error? 

Modifié par Bcoolpro, 11 avril 2012 - 04:42 .

[SupR G]

How about we save the thank you's until they actually make something? How many thank you's did you give them when they announced ME3 and all those "different endings", only to be disappointed? Point is, let's not act like little children, be patient and wait and see if it's worth a thank you when it comes out.

Modifié par SupR G, 11 avril 2012 - 04:47 .

[Treemarker]

Bcoolpro wrote...
Okay I'm not trying to open the same can of worms again, but I would appreciate some clarification.  From what I remember of my stats class a 4%+/- margin of error is at the upper end of what can be considered statistically significant.  If your 1 in 50 poll had a 20% margin of error then it is not a viable result.  Could you explain how it still has any predictive power when it falls so far outside the acceptable margin of error? 

Margin of error has more to do with predictive power than signficance testing.  Though it is related. More variation (error + true score) means you need to observe larger effects for significance.  Most polls don't report margin per item.  They report an estimated margin of error based on sample size and assumptions about the distribution of underlying opinions.  To take an extreem example, assume that the population is split 90/10 on some item.  I need a very small sample to detect significant differences with a 4% margin of error.  If the population is split 50/50, then I need a *MUCH* larger sample to achieve the same margin of error and detect significant differences.  In essense, its more about the confidence of a result than about finding a significant one.  Wikipedia has a pretty good write-up.  The point I'm making can be seen clearist in the section labeled 'Different Confidence Intervals' where the formula doesn't require the proportions, just a z-score associated with your level of confidence (99, 95, 90) and the size of your random sample.  The bottom of the page shows an actual test for significance of propotions under 'Comparing Percentages'.

[Treemarker]

hoorayforicecream wrote...There's been plenty of evidence in the past that game forums are not representative of the entire fanbase. The group that does post to forums is both small and self-selecting. There's plenty of evidence of this. It's been noted in MMOG forums especially, which are typically even worse than the BSN.

Frankly, you're both wrong. 

The link to http://www.slideshar...s/paying-to-win has one meaningful slide (slide 51) showing that about 22% of players of a free-to-play game actually ever visited the forum, and 2% were posters.  I'm not certain generalizes to anything else.  The article about MMOG forums has zero statistitical information but does represent, in a very readable way, someone's opinion on it. 

What we can say is that forum polls may generalize to forums.  Trying to go beyond that is shoveling crap.

That said, EA, Bioware, and such probably have a much better idea of just how representative the poll was than anyone participating on this forum.  After all, they can tie my forum ID directly to my game.  They KNOW who is posting, who actually owns the game, who actually plays online, and probably have a fairly solid guestimate of who finished the game, broken down by age, gender, and anything else you filled out on your profile. 

Thanks for those links though, both were interesting and the slideshare link gave me a whole new perspective on game marketing and business.  I had never though of conversion rates in that manner but it makes quite a bit of sense.

[The Ole Ultra Violence]

"You did good, son, you did good. I'm proud of you"

[Tazzmission]

earthonline wrote...

Just read here:

Pax East 2012 Bioware and Mass Effect 3 Panel Part 2: Extended Cut DLC

Producer Mike Gamble: "
We wanted to clarify a number of things. That we actually answer the questions. We are all on Twitter, Facebook. We have the community team feed us the questions that you guys have. We want to give more closure, answer questions, fill in some of the unknowns, give the players a sense of personalization with the endings, many people mentioned that some of the choices that were within the game are not necessarily reflected in the ending scenes, we're definitely going to focus on things like that, we want to make sure that when you see the ending of Mass Effect you now have the information in the context to feel satisfied. " 

THANKS BIOWARE for listening to us!!! We love you!!!

Please post here to support Bioware!!! <3<3<3

i have to give bioware credit because they have really thick skin when it comes to what alot ( not all) fans have been saying lately

[earthonline]

GuyIncognito wrote...

Thank you very much Bioware.

+1

[Xewaka]

SRobson wrote...
Changing an ending because most people didn't like it is equally ridiculous. No one demands this from movies or books.

So Ser Arthur Conan Doyle resurrected Sherlock Holmes because he felt like it, instead of extreme fan pressure?

Modifié par Xewaka, 11 avril 2012 - 08:33 .

[Icinix]

A clarification and context extension seems a little....overkill. They could get the same result from a press release...

..I think it will be bigger than we think or that they've implied.

Then again. Epic Community Event.

[Bcoolpro]

Treemarker wrote...

Bcoolpro wrote...
Okay I'm not trying to open the same can of worms again, but I would appreciate some clarification.  From what I remember of my stats class a 4%+/- margin of error is at the upper end of what can be considered statistically significant.  If your 1 in 50 poll had a 20% margin of error then it is not a viable result.  Could you explain how it still has any predictive power when it falls so far outside the acceptable margin of error? 

Margin of error has more to do with predictive power than signficance testing.  Though it is related. More variation (error + true score) means you need to observe larger effects for significance.  Most polls don't report margin per item.  They report an estimated margin of error based on sample size and assumptions about the distribution of underlying opinions.  To take an extreem example, assume that the population is split 90/10 on some item.  I need a very small sample to detect significant differences with a 4% margin of error.  If the population is split 50/50, then I need a *MUCH* larger sample to achieve the same margin of error and detect significant differences.  In essense, its more about the confidence of a result than about finding a significant one.  Wikipedia has a pretty good write-up.  The point I'm making can be seen clearist in the section labeled 'Different Confidence Intervals' where the formula doesn't require the proportions, just a z-score associated with your level of confidence (99, 95, 90) and the size of your random sample.  The bottom of the page shows an actual test for significance of propotions under 'Comparing Percentages'.

That is why these studies have to be conducted so carefully.  I would also posit that simply asking ones "like" or "dislike" of the ending isn't sufficient evidence for arguing whether or not it should be changed.  While a lot of people may not like it doesn't mean it spoiled the whole game for them.  On meta critic the positive and negative reviews are only separated by about 100 right now with a little over a hundred mixed.  That means even some who didn't like the end still gave an over all positive review.  So we may need to allow more distance from the flash point to allow for an accurate read.  

[Bcoolpro]

Xewaka wrote...

SRobson wrote...
Changing an ending because most people didn't like it is equally ridiculous. No one demands this from movies or books.

So Ser Arthur Conan Doyle resurrected Sherlock Holmes because he felt like it, instead of extreme fan pressure?

Doesn't count, Doyle didn't change the ending to "The Final Problem" and rerelease it... he simply found a way to have Holmes survive his alleged death. 

[kbct]

Bcoolpro wrote...

kbct wrote...

Imperium Alpha wrote...

There is a 100% margin of error. I'm always right so now listen to me.

It's about a 20% margin of error for small sample, self-selecting polls.

Okay I'm not trying to open the same can of worms again, but I would appreciate some clarification.  From what I remember of my stats class a 4%+/- margin of error is at the upper end of what can be considered statistically significant.  If your 1 in 50 poll had a 20% margin of error then it is not a viable result.  Could you explain how it still has any predictive power when it falls so far outside the acceptable margin of error? 

A random sample of only about 750 people can predict the a population of 10 million with a margin of error of 3-5%.

We know the BSN poll of 71,500 people shows 2% liked the ending as-is.

Let's say BSN poll is a random sample (we know it's not). A sample of 71,500 would have a margin of error of less than 1%, but let's call it 1% for now.

Under a normal distribution, the entire population people that liked the ending as-is would be 2% +/- 1%. 68% of the time it would be between 1% and 3%, 95% of the time it would be between 0% and 4%, and 99% of the time it would be between 0% and 5%.

A self-selecting sample has predictive power if the sample is large enough and the results are lopsided enough. Both criteria exist in the BSN poll. It is almost 100 times larger than what is required for a random sample. It is so skewed that it really can't be any more skewed. I mean, what is lower than 2%???

So, even if we use a 20% margin of error, 66% of the time, the percentage would be between 0% and 22% and 95% of the time, the percentage would be between 0% and 44%. That's still not a majority.

The thing is with a sample size so large, the margin of error won't be 20%. It will be less. So we have even higher confidence that the people that liked the ending as-is is not a majority.

Nothing is certain unless we ask every man, woman, and child what they thought of the ending, but we can state with a high degree of confidence that the majority didn't like the ending as-is.

Beyond that, there is more confirming evidence to suggest the majority didn't like the ending.

Modifié par kbct, 11 avril 2012 - 02:27 .

[kbct]

L00p wrote...

kbct wrote...
A random sample of only about 750 people can predict the a population of 10 million with a margin of error of 3-5%.

We know the BSN poll of 71,500 people shows 2% liked the ending as-is.

Let's say BSN poll is a random sample (we know it's not). A sample of 71,500 would have a margin of error of less than 1%, but let's call it 1% for now.

Under a normal distribution, the entire population people that liked the ending as-is would be 2% +/- 1%. 68% of the time it would be between 1% and 3%, 95% of the time it would be between 0% and 4%, and 99% of the time it would be between 0% and 5%.

A self-selecting sample has predictive power if the sample is large enough and the results are lopsided enough. Both criteria exist in the BSN poll. It is almost 100 times larger than what is required for a random sample. It is so skewed that it really can't be any more skewed. I mean, what is lower than 2%???

So, even if we use a 20% margin of error, 66% of the time, the percentage would be between 0% and 22% and 95% of the time, the percentage would be between 0% and 44%. That's still not a majority.

The thing is with a sample size so large, the margin of error won't be 20%. It will be less. So we have even higher confidence that the people that liked the ending as-is is not a majority.

Nothing is certain unless we ask every man, woman, and child what they thought of the ending, but we can state with a high degree of confidence that the majority didn't like the ending as-is.

Beyond that, there is more confirming evidence to suggest the majority didn't like the ending.

And kbct drives the ball out of the park - NASA reporting some foreign object is headed straight for the moon.

Are you reading this, you bastards? Are you?

Oh, and kbct? You're wrong because...well...because *I* liked the ending, and BioWare and EA did, and what you're saying is, just like, your opinion, man.
/sarcasm

Thank you, but I want to be clear that I'm only talking about the people that have experienced the ending since you can't have an opinion about the ending unless you experience it via the game, YouTube, or other source.

Then again, people that don't finish the game aren't exactly cash cows for EA/BioWare. Maybe they thought the game sucked or discovered the genre wasn't their cup of tea. I know when I don't finish a game, I usually don't like it for one reason or another. I wouldn't expect those people to pre-order, purchase DLC, or purchase game-related merchandise. And I wouldn't expect them to purchase the next game in the series or even be brand loyal.

[dfstone]

earthonline wrote...

Just read here:

Pax East 2012 Bioware and Mass Effect 3 Panel Part 2: Extended Cut DLC

Producer Mike Gamble: "
We wanted to clarify a number of things. That we actually answer the questions. We are all on Twitter, Facebook. We have the community team feed us the questions that you guys have. We want to give more closure, answer questions, fill in some of the unknowns, give the players a sense of personalization with the endings, many people mentioned that some of the choices that were within the game are not necessarily reflected in the ending scenes, we're definitely going to focus on things like that, we want to make sure that when you see the ending of Mass Effect you now have the information in the context to feel satisfied. " 

THANKS BIOWARE for listening to us!!! We love you!!!

Please post here to support Bioware!!! <3<3<3

Good for Bioware.  Thats exactly what I was hoping they would do.

[Treemarker]

kbct wrote...
A self-selecting sample has predictive power if the sample is large enough and the results are lopsided enough. Both criteria exist in the BSN poll. It is almost 100 times larger than what is required for a random sample. It is so skewed that it really can't be any more skewed. I mean, what is lower than 2%???

You were fine to here.  The math behind margin of error, and more importantly, the logic, assumes a random distribution of error and bias.  Its why the error rate is specified as a +/- since it is equally probable that random error is skewed positive for a sample or skewed negative.  This is not true for non-random samples. 

We don't know what biases may have affected the poll.  If I remember, nerd rage at the time the poll was conducted was fairly high on the forums.  Perhaps the people who liked the ending decided not to participate in what they saw as a hostile environment.  Perhaps the thread title chased people off.  Perhaps the question wording affected response (the equvilant of a push-poll in politics).  I can go on...

Seventy-one thousand is a pretty good number.  It should be used to suggest things, not argue they are proved.  In trying to come up with another example, I'm going to resort to Cosmopolitan.  When I was an undergraduate they performed their first poll of their reader's sexual behavior.  Totally unscientific, self-selected sample.  But it was *HUGE*.  So social scientists discussed it.  Questioned it. Replicated some of the items.  And discovered it didn't generalize to the generalize to the population (no surprise), but it did seem to describe the behavior of Cosmo readers fairly well.  Same situation here, only worse.  I'm not aware of any demographics associated with the poll. We can't tell if it was 71,000 kids, adults, or some mix thereof. 

Its a poll. People who answered the poll didn't like the ending. An overwhelming majority of people who took the poll didn't like the ending.  That's a pretty strong statement itself.  Going beyond that is 'speculation'.

[Solmanian]

slyguy200 wrote...

hoorayforicecream wrote...

kbct wrote...

I've seen you post that before, but that's a different subject. We don't know if it differs by company, brand, genre, platform, region, etc.

Got any BioWare specific data?

Roughly 50% of players actually finished Mass Effect 2.

What of those waiting for a fix for ME3's ending before continuing.

I'd be shocked if they comprised even 1% of the player base. Why would you wait for a fix something you can't even tell that is broken? Or perhaps you're suggesting they watched the ending to ME3 on youtube, because everybody who watches/play a trilogy of movies/games first check out the last five minutes of the serious to make sure it doesn't suck..

[Solmanian]

hoorayforicecream wrote...

Most players don't finish video games. It's highly unlikely that most players would want a new ending if they haven't even seen the old one.

Sad, but true. I have 200 games in my steam library (damn bundle specials...), and I finished maybe 10% of them. I'm glad to know I'm not alone. The ME series are one of the few games that have a good rythem and allow me to finish them without burning out of either boredom or exhustion. And ME2 is unique in that it's the only game (realy short games like CoD don't count) that I did a complete playthrough from start to finish without sleep, on launch day no less. Finished ME3 in a week, also an impressive time for me considering the months it took me to finish AC2 ...

[kbct]

Treemarker wrote...

kbct wrote...
A self-selecting sample has predictive power if the sample is large enough and the results are lopsided enough. Both criteria exist in the BSN poll. It is almost 100 times larger than what is required for a random sample. It is so skewed that it really can't be any more skewed. I mean, what is lower than 2%???

You were fine to here.  The math behind margin of error, and more importantly, the logic, assumes a random distribution of error and bias.  Its why the error rate is specified as a +/- since it is equally probable that random error is skewed positive for a sample or skewed negative.  This is not true for non-random samples. 

We don't know what biases may have affected the poll.  If I remember, nerd rage at the time the poll was conducted was fairly high on the forums.  Perhaps the people who liked the ending decided not to participate in what they saw as a hostile environment.  Perhaps the thread title chased people off.  Perhaps the question wording affected response (the equvilant of a push-poll in politics).  I can go on...

Seventy-one thousand is a pretty good number.  It should be used to suggest things, not argue they are proved.  In trying to come up with another example, I'm going to resort to Cosmopolitan.  When I was an undergraduate they performed their first poll of their reader's sexual behavior.  Totally unscientific, self-selected sample.  But it was *HUGE*.  So social scientists discussed it.  Questioned it. Replicated some of the items.  And discovered it didn't generalize to the generalize to the population (no surprise), but it did seem to describe the behavior of Cosmo readers fairly well.  Same situation here, only worse.  I'm not aware of any demographics associated with the poll. We can't tell if it was 71,000 kids, adults, or some mix thereof. 

Its a poll. People who answered the poll didn't like the ending. An overwhelming majority of people who took the poll didn't like the ending.  That's a pretty strong statement itself.  Going beyond that is 'speculation'.

Do you have a link for the Cosmopolitan poll? I want to make sure you're comparing apples to apple here.

Remember, I'm using a poll of 71,500 to infer information about a population of about 3 million ME3 game owners. To be comparable, the Cosmpolitan poll would need about 7 million votes to infer information about the population of the United States.

Or, like you said, It would have been more comparable if they inferred information about the population of the Cosmopolitan readers. After all, we are inferring information about population of ME3 owners, not the United States.

Anyway, I'm not suggesting 2% is the correct number for the population of ME3 owners that experienced the ending. It could be 10% or 30%, but it is almost certainly not 51%.