Horseshoes, Hand Grenades, and Government Work

Last year, I described my proposal for improving the WMATA metro system: basically, it’s captured in the picture below – basically, it involved one new Potomac bridge, one new under-Potomac tunnel, and a cross-town tunnel. I didn’t add in the link between Silver Spring and Grosvenor because that’s already in development by Montgomery County as light rail (due to the fetish of light rail in many places – even though it has the drawbacks of rail without the throughput of heavy rail).

I’m glad to see that society is starting to catch up: yesterday’s Washington Post included discussions about a major capital overhaul to the system designed to improve the carrying capacity, which will do some things but not every thing. (image below)

Washington Post’s rendering of proposed metro improvements

Half a loaf is better than no loaves, but I’m concerned about one of the fundamental omissions: the plan they’re describing does not add any Potomac crossings, and thus does not actually increase the aggregate throughput of getting people from Virginia to DC or Maryland. Additionally, it preserves the DC-centric plan, where there is no ability to bypass the city for those folks who live in Maryland and work in Virginia (or vice versa) – of whom there are a non-trivial number.

WMATA historically has made extremely short-sighted decisions about capital investment – the decision to build a two-track system rather than four-track means both that express trains are completely out and also that any issue on any one track or train completely screws up the whole line. The decision to build very high stations with vaulted ceilings means that overlapping lines are harder to accomplish. The decision to build the Silver line through Tysons as an overhead line means that it will be vastly more vulnerable to weather disruption, and maintenance will be more expensive (because fixing problems 30′ above ground is trickier than underground or at grade, likewise snow and water removal gets trickier). The decision to eschew multiple points of interconnection between the red line and other lines means that those interconnections are a complete balagon, and the decision to build the system to spec rather than to 400% of spec means that they routinely under project the actual demand for service. This is not to mention whoever had the brilliant idea about exposing escalators to rain.

I’m a fiscal conservative – I only want the government to spend money in limited and appropriate ways. However, it’s better to spend the money to do it right the first time, rather than spending a lot of money to not solve the problem. The new plan will solve problems inside the city, but I don’t see why, in absence of a second under-Potomac tunnel, it would solve problems between Virginia and the city, which are a lot of them. The improvements to Red-line Marylanders will be modest – the ability to ditch at Farragut is a long-overdue improvement (although you see that my recommendation would allow Dupont, Farragut, and Union Station to function that way). Hopefully the new plan is open for suggestions – I know what I want is expensive, but I think it will be cheaper in the long run than not doing it.

This seems like an appropriate thought for Tu b’shevat, the Jewish “new year of the trees,” a day on which environmental concerns are traditionally raised. Personally, I think we should be worried about environmental concerns every day, one day of the year prevent over noodginess.

Forward!

Any damn fool can predict the past.

Niven’s Laws, #8, Analog magazine Nov 2002

Today, Power line had a retrospective of failed predictions, and slashdot pointed out the new Berkeley retrospective study on CO2 and climate.

Interesting. Now, I have an open mind, but when I hear that a claim that a single variable completely controls a complex system, my skeptic hackles get raised. That picture sure looks like an extremely perfect fit – but in the times I’ve conducted experiments, I’ve never seen data which perfectly fit predictions (it should be within an error range, there should also be noise).

So retrospective analysis which precisely fits a specific variable makes me wonder about methodology more than it inspires confidence. What would impress me more is prospective predictions, made in a clear manner, saying that “we expect X increase in CO2 to correlate precisely with Y increase in temperature, unless there are Z volcanoes, and evaluate over W period”. That would be impressive. It’s also what hasn’t happened, and is why there’s so much sturm und drang about climate science – give me falsifiable predictions, à la Karl Popper (successor to the ideas of Francis Bacon et al.

Sadly, we’re not at the point yet where the medium term testable predictions have been formulated (which is extremely surprising to me), and the politics of the moment has injected itself into every part of the discussion, so any data point becomes the subject of partisan debate.

But the answer, as both Dawkins and Sacks would agree, to bad science, is good science, and this is what we need. But retrospective studies are great for validating whether a model has promise, but are not great for validating whether a model has predictive power, which is of course what matters.

Predicting the future is hard, but that doesn’t mean that it’s okay to predict the past and call it the same as predicting the future.

Note: I reserve the right to reevaluate any opinion based on any new evidence learned. I also reserve the right to be wrong, along with everyone else throughout human history.

On not Getting Killed on the Way Home

On my way riding home from the grocery store this morning, I was on the North side of the P street sidewalk headed West in Georgetown, between 31st and Wisconsin. Why the sidewalk, when that part of P street is pretty flat, if not downhill, so I wouldn’t hold up traffic? Because for idiotic historical reasons, on Sundays, parking is allowed on both sides of P street, making it nigh-well impossible terrifying for two cars to pass one another, so adding a bicycle to the mix sounded like a ticket to the ER.

32nd street intersects P street on the right, at a three-way stop, and is a one way street headed into P (ie do not enter).

So I’m just about to cross that intersection when some jerk decides to back up onto 32nd street nearly running me over so that I have to swerve to stop (glad I didn’tendo!). I raised my hands and hollered “what the hell do you think you’re doing?!” to which the response I got from the man driving was a rude gesture (no response from the female passenger), as he then forwarded on and drove away- you see this was a reverse U-turn, because apparently this idiot hadn’t ever learned that in the city one just circles the block to do that, or turns around at an actual intersection, or at least if you’re going to do something illegal at least make sure you’re not going to kill someone.

So. As if I needed more than coffee to get me going this morning.

(dons flame-retardant undies)

There has been a tremendous amount of discussion about gun control in the aftermath of the terrible events in Sandy Hook. I don’t write policy, and I don’t precisely know what, if anything, should be done. However, I think that some statistics may be useful to note (I’m using 2009 because of completeness of data available from awesomely reliable sources):

In 2009, there were 11,493 people killed by firearms (excluding suicides, but including accidents), which is a number big enough to worry about, but does not crack the top 20 causes of death (assault by all means is #18, but that’s 18k people). The source for this is the CDC factbook.

By comparison, in 2009, according to the US Census, 33,808 people were killed in automobile fatalities (moving, not suffocation suicides). Of those, 17,640, or about half, we’re the drivers, which means that drivers killed 16,168 people who were not driving at the time – roughly analogous to those being killed by any other deadly weapon.

Given these two numbers, shouldn’t we be spending three times more time and effort being worried about the multi-ton deadly weapons which are controlled by people with extremely poor judgement than we are about firearms, which honestly are a tough nut, given the specific constitutional protection and long American tradition of ownership? As a statistical measure of public health concerns, it’s clear that cars are a far bigger problem than guns – so I’d give a lot of credit to any politician who acknowledges this fact before making whatever policy proposals they make. It would certainly make me more likely to think they were dealing in an intellectually honest manner.