Wednesday 20 December 2017

Cloak of Invisibility

If you've read the title of the post, you might wonder whether I'm going to be constructing some elaborate metaphor about marginalized people or whether I'm posting about video games again.

It's video games.

I've been playing a lot of Hearthstone dungeon runs. It's very fun single player content which is what I want out of computer card games. I find Hearthstone fun to play, but when I play against other people, those people use up an
entire minute
deciding on their mulligan and then another entire
minute
before passing the turn without doing anything on turn one. I open up video games to play video games, not to fantasize about other humans beings choking.

I've written about hearthstone single player content previously. Sky at Bright Cape Gamer had written about the challenge in single player content and I had a very different take where I largely disparaged the idea of challenge in single player content. This time I am also launching off of something Sky wrote but instead of woe I am writing with incredulity. I agree with Sky's assessment - I like the dungeon run, there is challenge but it doesn't feel like just rolling the dice over and over, it's fun to keep doing even after you win. He's also right about
Potion of Vitality
. But he raises something I find super weird:
I have found it super interesting that people have wildly differing ideas of the power level of various items.  Some are obvious, such as the Captured Flag which gives your minions +1/+1.  It is excellent, one of the best for every class and strategy.  However, there is one in particular, the Cloak of Invisibility, that seems to have some serious disagreement on its strength.
Okay, so I guess I'm not surprised that Hearthstone players are not sold on the power of Cloak of Invisibility. Apparently people weren't sold on Dr. Boom when he came out. I heard people disparaging Darkshire Councilman when it was first available. The Hearthstone community is not good at evaluating how good cards are.

As above, Sky is right. It's good. Partly because it allows you to have good trades in combat, and partly because it has the potential to break the game against some encounters, particularly the Darkness, leaving them stranded with a full hand and unable to do anything for the rest of the game.

But what struck me is that anyone could even debate whether it's good or not in a general sense. I'm sure no one is debating whether doubling your battlecries is "good". There are a couple of decks that effect is very good in, but for most decks it's very close to useless.

Cloak of invisibility gives all your units stealth permanently. That doesn't obviously interact with cards the way double battlecries interacts with battlecry cards or sceptre of summoning interacts with cards that cost 8, 9 and 10.

Hearthstone is full of minions that have devastating effects while they are on the battlefield. It's
pretty obvious
that if your opponent can't remove your KelThuzad you win, but there are plenty of other cards that are very problematic if they stick around. Pirate and murloc decks that have cheap minions that buff other minions become very difficult to beat. Cards with powerful inpires like Thunder Bluff Valiant and Nexus Champion-Saraad are brutally overpowered when they can't be attacked. Frothing Berserker sometimes seems outrageously unfair when you can attack it, but when it can't be attacked it's easy to trade other minions while you beat them dead with a 15/4.

Cloak of Invisibility is the most broken effect of any of the passive treasures. Breaking the game is good, but you have to make sure it breaks in your favour.

It would be silly to try to say whether Robe of the Magi or Ring of the Justicar are good without thinking about what class you are playing. Khadgar's Scrying Orb is sometimes good and it sometimes isn't that good, but you can't evaluate it the same for
a warrior and a shaman
.

Active treasures lend themselves a little better to a strict ranking list where some are just plain a lot better than others. But aside from
a few
that are just all-around good and
one
that is all-around bad, the value of all passive treasures is "it depends."

Monday 18 December 2017

Sex-Blindness

I was reading an article from the Atlantic about the American Republican party's troubles holding onto supporters who are female. I'm not exactly recommending it, since it could be summarized pretty easily by saying that the Republicans have long thought that they have a problem in the way the communicate with women, and they are being rudely awoken to the fact that they have a problem with their actual policies.

But there was one part I found a very interesting insight into how patriarchal policies work:
“This is one of the main differences between the left and the right: We don’t see every issue as being a ‘man’s issue,’ or a ‘woman’s issue.’ It’s not a men-against-women, us-against-them mentality,” said Sue Zoldak, the head of communications for RightNOW Women PAC. “I don’t understand the idea that something is a ‘women’s issue.’ I don’t comprehend that as a statement.”
This is an interesting way of casting right wing thinkers as the righteous ones on sex equality. Well, I say "interesting". It's not that interesting because it's exactly what they do with race as well.

The subtitle of The New Jim Crow, which I do recommend, is "Mass Incarceration in the Age of Colorblindness". The book explains at great length how "colorblind" language is used to justify policies that as a matter of fact disadvantage people of colour, but that never overtly target people based on colour.

The penalties for drug possession don't actually say that black people will be put in prison for longer than white people, but the laws against drug possession have clearly been used as a tool to incarcerate black people.

It's undeniable there there is a difference in the wages people get paid based on their sex. If you were to choose a sex based purely on maximizing salary, you'd choose male. While there are fools who deny it, some of this gap is based on straight up inherent bias - the employer simply offers a person who looks like a woman to them less money than a person who looks like a man. However, there's lots of reason to think that kind of bias is only a fraction of the gap.

This Vox article goes through a bunch of data and puts together a few patterns:
  • jobs where certain hours are of higher importance and jobs where working long hours are valued tend to be jobs where the gap is higher
  • the gap increases when people are 30 and 40 and then goes down when they are 50 and older
  • women are disproportionately doing the work of raising children
You put all that together and it looks an awful lot like women's duties to their families prevent them from achieving the same salary as men in jobs where work hours are very inflexible and/or long. None of this is all that new to anyone who's watched this discussion. It's actually a very common explanation of the wage gap from people who want to downplay it's importance - that the wage gap results from real world differences between how men and women work. If a man gets paid more because he was willing to work longer hours, then that wage gap is justified, they say.

The Vox article doesn't go this way, but instead suggests that something we could do to help close the wage gap is change our attitudes about work. Some jobs have inflexible hours for very legitimate reasons, others have them just because that's how it's always been.

What it stops short of saying is that financially rewarding people for their ability to adhere to inflexible work hours are a discriminatory practice hidden behind a veil of sex-blind language.

We like to think there is a clear ordering of things, a clear way to say what causes what: A person who is male has their wife take on more of the after work childcare duties which means they can stay late which makes their boss think that they are more committed to getting the job done which makes them more likely to get promoted or get a bigger raise, which creates a gender wage gap.

But that way of telling stories is systemically discriminatory. I could tell the story in reverse: It is pointed out that there is a gap between the wages of men and women at a company which causes people who are in charge of determining wages to come up with a rationalization for that gap that doesn't make them look sexist, which causes them to latch onto working longer hours as a reason to promote people and give them higher wages, which causes them to actually promote people based on that criterion despite the fact that everyone knows it's a poor criterion.

I could tell the story starting in the middle of the chain and cascading in both directions. I could tell the story as feedback loops with no clear beginning or end.
We can write stories in lots of ways.


The same goes for the story that women don't negotiate as hard for salary when they start a new job, or that women take maternity leave. We know that paying people more because of their salary negotiations and that paying people less if they took a year off to look after dependents are two ways of doing things that result in women being paid less than men. But we refuse to acknowledge that those two things are themselves sexist policies because they are coated in sex-blind language.

I think one way we can emphasize how sexist these policies are by pointing out that they serve no real world purpose. Rewarding people for working longer hours is likely counterproductive. The Vox article briefly mentions this:
It also means not giving disproportionate rewards to those willing to work the longest, either. Numerous studies find that long hours aren’t always productive. One study published last year found that managers couldn’t tell the difference between those who worked an 80-hour week and those who pretended to. 
"The research is clear," the Harvard Business Review declared last summer. "Long hours backfire for people and companies."
This isn't a revolutionary idea from last summer, though, it's been standard, accepted theory of business administration since Henry Ford if not longer. People who work longer hours are not more productive.

Similarly, is paying people more because they drove a harder bargain during their interview actually a way to get better employees? It might make sense if you are hiring people to do sales since for sales staff their ability to drive a bargain is a direct asset. For most jobs I think the answer is almost certainly no.

When we have widespread implementation of a policy that seems to sacrifice better outcomes in the name of producing more sexist ones, it's probably pretty easy to think of that policy as sexist. What about a policy we can make sense of, like paying people more if they have worked more years?

Maternity leave seems to connect to the idea of seniority. You pay people who've been there for five years more than you pay people who've been there for one. Therefore you pay someone who has been there for ten years more than someone who has been there for nine - that is, ten but took one of those years off.

But if you think that pay for seniority comes out of the idea that people with more experience are better at their jobs, I think you're applying a contemporary reasoning to a policy that has existed in many places for a long time. It's just as true that pay for seniority is about rewarding people for staying because turnover is hard on companies. For that latter explanation, a year of maternity leave doesn't change the reasoning for why you would pay an employee more. The rationalization we choose for our cultural tradition affects women's equality, and we chose the rationalization that goes against equality.

My arguments that these polices are sexist are, in a way, speculative. What I can say for sure is that discussing any of these polices without acknowledging their contribution to the gender wage gap is definitely sexist. Unequal pay based on gender is a bad thing. If it is the consequence of a policy, that policy had better be at least good enough to outweigh the harm.

Like if a company pays sales staff by commission and is convinced that paying by commission greatly increases the success of the company, they may acknowledge that it can also contribute to a gender wage gap by rewarding working for longer hours but say that it is nonetheless a policy they need to keep. Firefighters don't relax their rules about how much you have to be able to carry even though those result in discrimination based on sex. You can justify a policy as being important enough to overcome it's downside. But choosing not to even evaluate gender-based wage inequality as a bad thing is
explicitly
promoting sexism. It's shouting sexism from the rooftops.

Trying to frame policies in a way that is blind to discrimination is directly promoting discrimination. Policies may promote discrimation, they may reduce it, they may have no effect. If you don't care which one of those your policy does, you don't care about discrimination, and that means that you
support
whatever form of discrimination shows up in popular culture.

Tuesday 12 December 2017

Solving the American Healthcare Debate

Whenever someone raises the idea starting a single-payer healthcare system in the United States, someone says it would cost too much. They act like people are naive and say things like, "I want a free pony too!"

According to OECD numbers, Canadians spent $4071 of public money per person in 2015. Americans, by contrast, spent $4692 of public money per person in 2015.

The Canadian system doesn't just cost less, it costs fewer public dollars than the American system. The private dollars poured into it by individuals are on top of this public spending.

We aren't talking about free ponies here. We are talking about stopping paying the full cost of a pony plus 15% to make sure no one has a pony, and then telling people they have to buy their own pony if they want one. I can't think of a reason for such a public policy other than lawmakers who feel indebted to the pony industry.

But I started thinking about those numbers. Suppose American lawmakers could wave a magic wand and having the Canadian system. In addition to all the private money that could be used for other things, the government would save $621 per person for 323.1 million people. That's just barely over $200 billion.

So you might think they are throwing $200 billion in a hole, but they aren't just throwing it in a hole, they are spending it. They are spending it to kill Americans. A 2009 study concluded that being uninsured meant about a 40% higher risk of death. America has a death rate of 823.7 per 100,000. The uninsured rate in America is was 11.3% in the first quarter of 2017. That death rate, is therefore composed of the 11.3% of people who have a 40% higher risk of death and the 88.7% who have a "normal" risk of death, which means the "normal" risk of death is 788.1 per 100,000. So the excess death rate caused by lack of insurance is 35.6 per 100,000. With 3231 groups of 100,000 Americans, that would give us
115,092 Americans dying each year from lack of insurance
.

Divide that into $200 billion and you get a price of about $1.74 million per american killed.

Now, let me ask you, what do you think it costs to hire an assassin?

That's not something I can google easily, nor is it even something I want to type into google. But I feel like it's safe to say you could procure that service for less than a million dollars.

So I have a solution to the current American healthcare problems. First, implement a single payer system that works much like Canada's does. Second, in order to mollify the people who don't like assistance from the government, also hire assassins to kill about 100,000 Americans a year.

Better health outcomes, lower prices. That's a win-win.

A Note for Those Interested in Making a Counterpoint
The math in this post looks simple but it's not that simple. Estimating the number of people who die from a lack of insurance is hard to impossible, and some people dispute there being any causal relationship between those two things at all. Some people want to say that no one dies from lack of access to healthcare, some people would probably point out that my figure of 115,000 is much higher than any other estimate, almost three times as high as the estimate from the study cited in one of my linked articles that was conducted before the ACA when there were more uninsured Americans.

So why would I use such a high estimate? I was being generous towards the current system. My calculation was the cost per American killed. More Americans killed by the current system means a lower cost per American killed. If a million Americans died a year from the lack of single payer healthcare, the cost would only be about $200,000 per death. At that point, you might say, "Humbabella, can you really get assassins for $200,000 a target? Maybe you could under some circumstances, but through government procurement processes?"

Then I'd have to admit that my plan probably wouldn't save money. But if only 10,000 Americans die for lack of healthcare then the cost is $20 million per American killed. There's no way that assassins aren't cheaper than that.

And if you think that no one dies for lack of healthcare, like some American politicians seem to, then I have two things to say.

First, you are transparently disingenuous and think I am stupid. Otherwise, you wouldn't want healthcare for yourself. If something that saves lives costs a penny more, other things being equal, some fraction of a statistical person dies. There's no way around that math.

Second, in that case you are paying $200 billion and not killing even a single one of your citizens for that? What the hell are you paying for?