Explaining the Regional Connector

LA Metro is just finishing up a project that is very dear to my heart because I think it’s pretty important. Transit advocates sometimes fall into a trap sometimes–some transit lovers are uncritical and they will advocate for projects that really maybe never need to be built. Bad projects do exist, and the suck resources away from the agency. As a result transit skeptics find it easy to write off advocates.

Another problem advocates face, particularly in Los Angeles where transit riders are such a minority, is helping people understand why some projects really are damn good projects. I’ve found that to be the case with the Regional Connector: often, when somebody asks me why it’s such a nice new part of the network, I find myself describing how different trips will be possible with no or fewer transfers. Unfortunately, stated that way, the regional connector sounds like it’s solving a “me” problem–that is a problem that irritates me and the few people who are like me, but isn’t really objectively an improvement.

One nice way of characterizing the improvement is simply that kids in East LA will be able to take a one-seat ride to the ocean. That should matter to people–it does to me.

Nonetheless I’ve dipped in out of graph theory attempts to characterize why the regional connector is so nice from a *network possiblity* perspective. In transit, we have tons and tons of ways of measuring outputs (passengers, etc) and some inputs (usually monetary, but operator-hours, etc). But advocates have few ways of discussing the physicality of the network in short, summary measures. And I’m not suggesting that we should always just look at one metric to understanding a network–we should use a bunch, but I do think there are some key calculations that help me show just how nice the regional connector is for the network.

I first encountered some measures in Vulcan Vuchic’s 2005 textbook. He’s got a *boatland* of cool network characterization and performance measures in there. I set my class on working with some of them–I used the Madrid network for a class project because I had some students in there who would insist on just counting nodes and lines in the toy networks I gave them instead of using the equations, and Madrid has a very complex network that would cross your eyeballs if you tried brute force the count.

I got to thinking that one way to help us understand how the Regional Connector serves as a nice additional is to simply look at the new trips it makes possible. This would just be a matter of looking at the transit lines, the stations, and their connections–but if you know the proper number of stations, you know the possible origins and destinations. To get the proper number of stations, you count them all up and remove the duplicates:

Then, to sort the O-D possibilities, you just use that N and N-1:

Here’s what you get when you do this for three LA rail network states: Current potential trips (O-D pairs), when the Regional Connector comes online, and after Metro finishes all the new bits of the Purple Line and Crenshaw (All). (Note the base network includes the Orange Line, and that I set up the Regional Connector lines based on the promised new stations, Chinatown station being open, and Metro’s planned goals for interlining the Gold Line (I’m sorry, I never remember their letter names) and the Expo line for that East LA to Santa Monica Ride and Blue line to Gold line headed north).

This strikes me as a nice way to show just how much the Regional Connector adds to trip possibilities in the region. I mean, these are all nice projects, but…I think you have my point. This isn’t just a “me” problem.

Now these network measures are obviously partial. It might be cool, for example to weight all these O-D possibilities by existing and projected ridership by station (I don’t have those data because Metro thinks I’m a gangster or because I don’t know where to find them, one or the other.)

And, just by way of illustrating that, there’s another idea that Vuchic suggests–the notion of directness. you can figure out how many O-D pairs are possible without a transfer with:

Vuchic suggests a ratio of those direct trips to total trips:

In each network state, delta goes down–let’s interpret that. In theory, it would be nice for delta to be a relatively large number (between 0 and 1). The larger the number, the more one-seat trips you are offering. In our case, the amount of OD pairs just jumps so much that the delta goes down: in the order given above, from 0.18 now to 0.15 with the regional connector to 0.10 with all the new line openings in the next few years. Delta becomes much less an issue with greater frequencies. Again…this is just one measure it’s not meant to say the network on the whole is worse off–it’s that this one measure, directness, declines with the new lines. It makes sense.

Sorry my equations are wonky but I am no mood to futz with making everything perfect-looking.

So good-O to whoever thought up the Regional Connector. It’s a goodie. I raise my glass to y’all.

I can’t guarantee this is all 100 percent perfect because I’ve had kind of a bad week, but it is interesting enough to share. Hope you enjoyed.

All good things,
L