Comparison of NPC Crew Skill Level and their Gender... Male Bias perhaps?

Status
Thread Closed: Not open for further replies.

Susanna

Susanna
Hi Folks,

Firstly, [Big Caveat] this is on a sample of only 50 different Crew Lounge load screens over a couple of weeks (mostly taken while docked at Jameson Memorial) . I've tried to ensure that each of the 5 characters have loaded with a different character to ensure fairness that a new character has been loaded into each of the 5 possible NPC rank grades [from Harmless to Expert] to ensure that repeats are kept to a minimum but where repeats exist, this has been down to the game selection of those NPC crew slots.

Secondly, many of the female characters are so androgynous (dare I say 'blokish') that its pretty hard to tell whether they are even female at all (unless they have a certain chest characteristic which seems to be largely universal in the female character avatars)! Hence, a few of the assigned gender markers may be wrong where I may have attributed Male to some of them when they are in fact meant to be Female characters. But I have seriously tried to have a fair stab at getting this right. The linked Spreadsheet with the data has links to all of the images taken of the NPC Crews so you are welcome to challenge characters that I may have got wrong...

I know that FDev wish to have a 'gritty - space dock feel' to the game, but taking the female characters to such extremes seems a bit excessive to me personally... [You may like them, I generally don't]

Thirdly, I appreciate that more men play ED than women but that is no excuse for a gender bias being evident within a game IHMO... You may get different mileage.

Spreadsheet OneDrive Link: https://1drv.ms/x/s!AnKn07CCvLJ2rFk-evfOjFsXuMaA

Spreadsheet Dropbox Link: https://www.dropbox.com/s/09t905xnyklb2d2/NPC Crew Profiles.xlsx?dl=0

Image of Results Table & Graph of data:
rdEKnYy.jpg


As you see from the graph above, the more senior Expert NPC crew members appear to have a the higher the likelihood that they will be Male... (twice as likely). [tongue-in-cheek warning] In other words, I'm saddened that a clear gender bias is still taking place in the year 3404 :O [/tongue-in-cheek warning]

If you have Crew Roster images that you have screen shots of and wish to include, PM me and I'll give you my email addy for you to share them with me and I'll add them to the table.

I'd welcome comments about this finding... Do you think this should be a thing? That male NPCs appear to have an inherent higher rank(s) than their female counterparts? The Male Expert NPCs may have a higher ratio than the female ones but more data points are needed before any statistically meaningful conclusions can be even remotely considered... [tongue-in-cheek warning] In other words, this is only for fun - i.e. playing the Forums...[/tongue-in-cheek warning]

Personally (again), I find this an annoying... [mad] Can we have a fair distribution between the male and female NPCs please and remove this blatant gender bias which is just more subliminal programming that men don't need any more of!

Note-to-Self - don't use hyperbole or other 'amusing' wording as people take it literally all too often as they don't know your sense of humour and see only the text as it is ... Yeah - that is how it is. If you receive a message from an alien race - be very careful how you interpret it!

[video=youtube_share;AZPGJZen9v8]https://youtu.be/AZPGJZen9v8[/video]
If you haven't seen it - It's well worth it!
 
Last edited:

Bryan K

Bryan K
.................

You have a lot of time on your hands, don’t you? Repped for proper research.

I always assumed it was just RNG. Interesting to see such a clear bias.

Doesn’t bother me too much, though. What does bother me is how, as you say, “blokish” alot of the female pilots are. I only hire hot chicks ;).
 
Last edited:

ITZC0ATL

ITZC0ATL
Hmmm, very odd. If this is indeed the case and your statistics are even close to accurate, then someone would have had to specifically code this modifier to give male pilots higher skill. That's pretty inexcusable - there's just no good reason to do that in 2018, let alone 3304 :O
 

Susanna

Susanna
HAL- 9000 said:
Hmmm, very odd. If this is indeed the case and your statistics are even close to accurate, then someone would have had to specifically code this modifier to give male pilots higher skill. That's pretty inexcusable - there's just no good reason to do that in 2018, let alone 3304 :O
Click to expand...

Indeed... it appears to be a deliberately coded bias towards the male characters having a higher likelihood of being presented and in the case of Expert pilots, the ratio of 2:1 is hard to fudge with a sample of 50 ... If you flip a coin 50 times you may get the same 2:1 results but the odds are improbable that it will be anything other than close to 50:50 split over a large enough sample... I'm sure someone can quote me the number of standard deviations required to get such a 2:1 ratio.

Hence, this looks deliberate to me. I'll continue to monitor and see how the trend continues...
 
Last edited:

Darty

Darty
Maybe there is still an elusive male conspiracy in 3304 to keep women down through ever increasingly subtle social engineering.
 

Jukelo

Jukelo
I understand the sentiment but at the same time...

HAL- 9000 said:
there's just no good reason to do that in 2018, let alone 3304 :O
Click to expand...

... I disagree with the principle of that sentence. I have no problem with human society in a fictitious world displaying gross misoginy, racism, and all the baser instincts known to man. It's not very original, but it makes for a more interesting world (ie one where conflict exists)

However I'm not aware that human society in 3304 still has those traits (then again humanity is not monolithic, so it's entirely possible that the various instances of powerful women in the ED universe isn't representative of the whole).

On the other hand, let's also look at the experiment itself: you have a sample size of 250, which is not exactly an awful lot. 4 of your subsets are probably trending towards an average of 50%, and only one outlier shows a gross disparity. If indeed this result is more than just sheer luck, I'd expect the repartition of expert pilots to be a bug of some sort rather than a conscious decision considering the other ranks don't appear to be showing such a disparity.
 
Last edited:

Kaocraft

Kaocraft
Susanna said:
Hi Folks,

Firstly, [Big Caveat] this is on a sample of only 50 different Crew Lounge load screens over a couple of weeks (mostly taken while docked at Jameson Memorial) . I've tried to ensure that each of the 5 characters have loaded with a different character to ensure fairness that a new character has been loaded into each of the 5 possible NPC rank grades [from Harmless to Expert] to ensure that repeats are kept to a minimum but where repeats exist, this has been down to the game selection of those NPC crew slots.

Secondly, many of the female characters are so androgynous (dare I say 'blokish') that its pretty hard to tell whether they are even female at all (unless they have a certain chest characteristic which seems to be largely universal in the female character avatars)! Hence, a few of the assigned gender markers may be wrong where I may have attributed Male to some of them when they are in fact meant to be Female characters. But I have seriously tried to have a fair stab at getting this right. The linked Spreadsheet with the data has links to all of the images taken of the NPC Crews so you are welcome to challenge characters that I may have got wrong...

I know that FDev wish to have a 'gritty - space dock feel' to the game, but taking the female characters to such extremes seems a bit excessive to me personally... [You may like them, I generally don't]

Thirdly, I appreciate that more men play ED than women but that is no excuse for a gender bias being evident within a game IHMO... You may get different mileage.

Spreadsheet OneDrive Link: https://1drv.ms/x/s!AnKn07CCvLJ2rFk-evfOjFsXuMaA

Spreadsheet Dropbox Link: https://www.dropbox.com/s/09t905xnyklb2d2/NPC Crew Profiles.xlsx?dl=0

Image of Results Table & Graph of data:
https://i.imgur.com/rdEKnYy.jpg

As you see from the graph above, the more senior the NPC crew member [Expert] the high the likelihood that they will be Male... (twice as likely). In other words, I'm saddened that a clear gender bias is still taking place in the year 3404 :O

If you have Crew Roster images that you have screen shots of and wish to include, PM me and I'll give you my email addy for you to share them with me and I'll add them to the table.

I'd welcome comments about this finding... Do you think this should be a thing? That male NPCs appear to have an inherent higher rank(s) than their female counterparts?

Personally (again), I find this an annoying... [mad] Can we have a fair distribution between the male and female NPCs please and remove this blatant gender bias which is just more subliminal programming that men don't need any more of!
Click to expand...

Why you gotta use pink to represent the female numbers and blue to represent males? Huh? Seems like clear gender bias at play here.

Seriously, though. I support this work. Keep going! You need waaaaay more samples before you can reasonably draw any conclusions here. Personally I suspect that once the sample size gets large enough, that the ratios are going to converge to 50/50. Why? Because it would take extra work to build a non-egalitarian gender-skill distribution into the game, and I just find it implausible that Frontier would bother to do that.
 

Susanna

Susanna
Yeap - it certainly could just be a bug!

It may also be a case that there were a lot of male backers who had their names included in the game as NPCs... Could this be what has caused the bias to occur?
 
Last edited:

CMDR Corlas

CMDR Corlas
Interesting data. I wonder if it's just a statistical anomaly with your data sample, rather than a hard-coded bias.

Although, to be fair, my only logic in that is it would be much more a pain to hard-code a gender bias than it would to just implement generic pseudo-randomness into NPC pilot creation, and coders tend to be lazy (it's one of the virtues of a programmer, after all). I guess I shouldn't underestimate the pettiness of some men's patriarchal leanings though. :/

I wonder if I can replicate your results there, that would be interesting.

I suppose it might be an example of an exhibited unconscious bias - for example, if possible NPC names were stored by starting rank, and the person in charge of populating that list thought of more "male" names than "female" or something?

Either way, interesting data. Would be interested to see an official response.
 

Susanna

Susanna
Kaocraft said:
Why you gotta use pink to represent the female numbers and blue to represent males? Huh? Seems like clear gender bias at play here.

Seriously, though. I support this work. Keep going! You need waaaaay more samples before you can reasonably draw any conclusions here. Personally I suspect that once the sample size gets large enough, that the ratios are going to converge to 50/50. Why? Because it would take extra work to build a non-egalitarian gender-skill distribution into the game, and I just find it implausible that Frontier would bother to do that.
Click to expand...

I agree - why would they bother! But I was watching it for a while and it started to look odd to me so I started to track it. It may just be RNG, it may be a bug, but it appears to be persistent so far - at least for the Expert rank. The others have trended towards 50:50 as I've gathered more data points.
 

ElleShaped

ElleShaped
What if the bias isn't because they set it up to explicitly make more male pilots than female pilots, and instead is because it picks a pilot randomly from a pool of pre-designed 'expert skill pilots' and -unconscious- bias by the design team caused that pool to be overpopulated by males?
 

Adhock

A
You've never worn body armor have you?

In order for body armor to do it's job properly, it needs to have a specific outer shape. If this shape is not maintained, the armor will do a improper job and the individual in question is more likely to die. Hence the armor in question will appear to 'flatten' out the more 'curvy' feminine features we associate with the female gender as it is scientifically proven that angular armor is more effective and we tend to associate angular features with the male appearance.

The reason I bring up the whole point of body armor is that our flight suits double as such. It makes a certain sense given we are in vehicles with angular insides that tend to explode. As someone who's worked with and inside armored vehicles, even a flak vest was welcome protection when I was involuntarily flung into something or something was flung into me.

Therefore it stands to reason that in order for women to appear busty, they would indeed need busts of unusual size in order to affect the shape of their armor/flightsuit. An average woman would flatten out.

Detail number two, Newton's First Law.

Specifically, how it applies to hair.

Long hair is often a trait associated with being 'feminine'. It is also a death sentence in high speed situations. How? Think about how much hair you have in front of your face. Think about how much hair would shift around from behind you to in front of you if you were to suddenly and violently change direction.

If your hair is going to blind you, you can't see and therefore are more likely to collide with something or get shot because you couldn't evade incoming fire due to being blinded.

Smart females would cut their hair short or even off.

Thus with short or no hair being associated with the masculine, women would appear more androgynous.

So already you have a double jeopardy of feminine traits being muted in favor of survivablity and life preservation. This would in turn convince me your results could very easily be contaminated with several cases of gender confusion without other details confirming their gender. You should have added a third column of "unknown" in order to better present your case.

Also the 50:50 coin split experiment was done already by a children's show. In order to achieve something close to a 50:50 split, the coin flip was completed in excess of 2,500 times.

Your sample size is WAY too small to even account for a fraction of that. When it was done up to 4 times it was shown to be a 3:! split.

Come back with a sample size of 3k and a third column to account for those you can't tell rather then just assuming them to be male. After all, assuming someone's gender is a crime these days right?
 

Turbosnowy

T
Well this certainly seems topical given the mass hysteria going on at the moment. All I would say if you can probably tell the gender based on their blurb 'I used to be a space trucker but then something happened yadda yadda'. This may well have a give away as to whether they are male or female.
 
Status
Thread Closed: Not open for further replies.
Back
Top Bottom