[wpietri]

Problem definitely not solved. There's too much uncertainty, and comparisons are too hard.

This happens all the time in software. Non-techies put a project out for bid. They will get estimates from a variety of consulting companies. The estimates could easily vary by an order of magnitude. Do they go with the lowest, because they are the most efficient operation? Or is the low price a sign that they're the dumbest, and are underestimating? Do they go with the highest, because in a market economy price is correlated with value? Or are those guys just the gougers?

And medicine is worse, because things are so contingent. Complications get complicated indeed. And the person trying to make the decision is pretty likely to be sick. I tell you true: deciding what to do about cancer is hard enough without trying to organize and judge a multi-round competitive bidding process.

[wfunction]

Claiming something is "too hard" never got anyone anywhere.

[wpietri]

Neither did waving away problems as if they didn't exist. If you can solve this, solve it. But arguments of the form "in theory there's a solution, so problem solved" aren't a step forward; they're a way to prevent others from trying.

[wfunction]

> they're a way to prevent others from trying.

What exactly was I trying to prevent others from trying?

[wpietri]

Well, I was assuming that you were participating in the conversation, rather than just seizing upon a sentence and nitpicking.

The conversation was about the role of the free market in health care.

User curt said that the only way to solve health care pricing was through the free market via individual choices. Mikeash said that would only work if we were happy to let people die of treatable diseases. Anonymoushn said that was a distraction, that the market would be fine if people could know prices up front. Grey-area said that predicting costs was basically impossible, so the market-based solution was unworkable.

You jumped in to say that prices were predictable through this new thing we hadn't heard of called "estimation", saying that this solved the problem. The problem, presumably, being how price opacity and unpredictability make a hash of individual-choice market mechanisms.

So I have no idea what you were trying to do, but as far as I can tell, it was either a) pushing people back toward an unworkable market-based solution for health care, thus preventing them from moving toward a solution that doesn't fit with the fundamentalist liberarian ethos so popular on HN, or b) you're doing this: http://xkcd.com/793/.

Given that you ignored the meat of my point twice in a row with nitpicky comments, as well as doing the same to the previous poster, I'm figuring b.

[wfunction]

I was actually responding just to the issue of predicting costs, not to the upthread issue of whether or not a free market is a good idea.

Predicting costs is a benefit regardless of whether or not we have a free market -- I was saying that nothing is completely unpredictable; estimating based on the average value of a treatment gets you much further than knowing nothing at all.

Claiming costs are inherently unpredictable is equivalent to calling B.S. on this report -- that the values in the report are meaningless. Clearly, they're quite meaningfull.

I never claimed estimation = "problem solved", I just said that it's a step forward that needs to happen. But you came along and just dismissed the idea because it wasn't a solution.

And my reply, which you thought was ignoring your comment, was that a step forward doesn't have to be a solution -- you can't just dismiss it because it doesn't stamp "Problem solved" on the bottom of the page.

Unless you or others are claiming that being given an estimate of costs HURTS anyone, I don't see how you can argue it's a bad idea. "Problem not solved" never solved any problem.

[wpietri]

Responding to some random sentence in the middle of a discussion as if you actually have something relevant to say is annoying, and a big waste of other people's time. If you want to go off on a tangent, say so. Then nobody has to try to figure out what you're on about.

Predicting costs is not an inherent benefit. That only is useful if somebody will change a decision based on the prediction. Otherwise, it's wasted effort. My mom was on Medicare throughout her cancer treatment, and I promise you: not knowing the costs was perfectly fine.

In fact, not knowing was better. The cognitive load of a major illness is huge. Not knowing is also better for most medical personnel. Their goal should be to maximize patient outcomes. Asking them to juggle some sort of cost-benefit tradeoff in the middle of treatment adds an insane and impossible burden.

Also, you quite literally did say "problem solved" after proposing SWAG estimation as a solution to a problem. Maybe you wanted me to have some other understanding, but I have no idea what it would be. And, since you admit you are off on some sort of tangent, I don't really care.