Green Line Extension to Medford & Union Sq

Going from 4 regular service patterns (in addition to the rando exceptions and peak-hour short-turns) to up to 16 regular service patterns doesn't seem to be particularly intuitive for new or infrequent users.

B1 B2 B3 B4
C1 C2 C3 C4
D1 D2 D3 D4
E1 E2 E3 E4

Also, woe be the service planner, operations manager, and operators/motorpersons that have to create/manage/get some likely wildly-interlined schedules.
The point of variable pairings would be precisely to not "create/manage/get some likely wildly-interlined schedules." You don't schedule out 16 regular service patterns, you schedule out four (the letters) to roughly the the same extent they do now, and then you on-the-fly assign number termini to the the trains when they turn around at their letter termini. Or to the extent you do schedule out all 16, you don't publish that schedule. So when a train turns around at, say, Riverside, the system dispatcher looks at what's ahead of it on the line and determines whether it should go all the way to Medford/Tufts or Union, or whether needs to short turn at North Station or GC. A rather simple little scheduling algorithm could run behind the scenes guiding those dispatching decisions, but also crucially this would allow for a lot of flexibility when unplanned service interruptions or disruptions pop up.

Infrequent users don't need to know what's going on behind the scenes at all. All they need to know is, e.g., "I need an E train to go to the MFA" or "I need a 1 or 2 to go to the Science Museum" or "any number will take me to Park Street from the Back Bay." This is no different from, say, "I need a B, C, or D to get to Fenway Park."

With the set pairings across the GL branches on both sides of the Charles, you'll end up with a situation where a car accident or a track problem at, e.g., Brigham Circle will shut down service between Medford/Tufts and Lechmere. That doesn't seem to make a lot of sense. With more flexible pairings, you'd be able to optimize more consistent and shorter headways across the entire system. And you could very well end up with fewer short turns at GC or NS, which would be a win for connectivity.

With set pairings you'll also end up with a situation where a commuter from Ball Square will never have a one-seat ride to their job in Kenmore Square, while a commuter from Allston/Brighton will never have a one-seat ride to their job at North Station. With more on the fly mix-and-match flexibility, those commuters would usually-but-not-always gets a one seat ride, and on the days that they don't they could wait for the next train in any station along the central subway just like they do now.

Given how much tech workers love Somerville, I feel like switching the Medford/Tufts branch from D to E could actually be a big deal for the growing tech cluster in the Fenway / Kenmore area. With more flexible pairings, that wouldn't be a thing.
 
Last edited:
The point of variable pairings would be precisely to not "create/manage/get some likely wildly-interlined schedules." You don't schedule out 16 regular service patterns, you schedule out four (the letters) to roughly the the same extent they do now, and then you on-the-fly assign number termini to the the trains when they turn around at their letter termini. Or to the extent you do schedule out all 16, you don't publish that schedule. So when a train turns around at, say, Riverside, the system dispatcher looks at what's ahead of it on the line and determines whether it should go all the way to Medford/Tufts or Union, or whether needs to short turn at North Station or GC. A rather simple little scheduling algorithm could run behind the scenes guiding those dispatching decisions, but also crucially this would allow for a lot of flexibility when unplanned service interruptions or disruptions pop up.

Infrequent users don't need to know what's going on behind the scenes at all. All they need to know is, e.g., "I need an E train to go to the MFA" or "I need a 1 or 2 to go to the Science Museum" or "any number will take me to Park Street from the Back Bay." This is no different from, say, "I need a B, C, or D to get to Fenway Park."

With the set pairings across the GL branches on both sides of the Charles, you'll end up with a situation where a car accident or a track problem at, e.g., Brigham Circle will shut down service between Medford/Tufts and Lechmere. That doesn't seem to make a lot of sense. With more flexible pairings, you'd be able to optimize more consistent and shorter headways across the entire system. And you could very well end up with fewer short turns at GC or NS, which would be a win for connectivity.

With set pairings you'll also end up with a situation where a commuter from Ball Square will never have a one-seat ride to their job in Kenmore Square, while a commuter from Allston/Brighton will never have a one-seat ride to their job at North Station. With more on the fly mix-and-match flexibility, those commuters would usually-but-not-always gets a one seat ride, and on the days that they don't they could wait for the next train in any station along the central subway just like they do now.

Given how much tech workers love Somerville, I feel like switching the Medford/Tufts branch from D to E could actually be a big deal for the growing tech cluster in the Fenway / Kenmore area. With more flexible pairings, that wouldn't be a thing.
interesting idea. I'm having a hard time with the idea that each train would have a specific color, letter, and number tho. Seems needlessly confusing.
 
interesting idea. I'm having a hard time with the idea that each train would have a specific color, letter, and number tho. Seems needlessly confusing.
Under the letter / number framework I'm thinking of, each train would only have the letter OR the number of which way it is travelling. It doesn't have both. So a train travelling from Heath St towards Medford/Tufts would be a 1, while a train travelling from Medford/Tufts towards Heath St would be an E. There's no need to attach both 1 and E to any given train. No rider cares where a train is coming from, only where it's going.

Separating out numbers and letters in this way would also help people understand which direction they are going. If you're at Park St and you want to go to Copley, take a letter train. If you're are Copley and you want to go to Park St, take a number train.
 
Under the letter / number framework I'm thinking of, each train would only have the letter OR the number of which way it is travelling. It doesn't have both. So a train travelling from Heath St towards Medford/Tufts would be a 1, while a train travelling from Medford/Tufts towards Heath St would be an E. There's no need to attach both 1 and E to any given train. No rider cares where a train is coming from, only where it's going.

Separating out numbers and letters in this way would also help people understand which direction they are going. If you're at Park St and you want to go to Copley, take a letter train. If you're are Copley and you want to go to Park St, take a number train.
so the green line which already has 4 branches in one direction and growing to two in the other and doesn't use "a" in its lettering system should have letters going one way and numbers going the other?
Seems like making already poor usability needlessly more complicated.
 
so the green line which already has 4 branches in one direction and growing to two in the other and doesn't use "a" in its lettering system should have letters going one way and numbers going the other?
Seems like making already poor usability needlessly more complicated.
In my view it is less complicated, not more complicated, but agree to disagree.

And even if it were slightly more complicated, having more flexible routes that result in better train ops at the expense of a slightly more complicated map is a trade-off I would take any day.
 
"The destination of this train is: _________"

Don't need alphanumeric codes to orient yourself that way...just an endpoint and a direction of travel. It is quite literally old habit that way for the inbound termini.

I think we're ginormously overestimating the wayfinding complexity of the Green Line.
 
"The destination of this train is: _________"

Don't need alphanumeric codes to orient yourself that way...just an endpoint and a direction of travel. It is quite literally old habit that way for the inbound termini.

I think we're ginormously overestimating the wayfinding complexity of the Green Line.
so you'd do away with the letters?
and why is there no A?
 
so you'd do away with the letters?
and why is there no A?
The A was the Watertown Branch, bustituted 1969.

The letters have never been used for inbound termini, so we're already a half-century accustomed to looking at the full name on the destination sign. If there were no letters and we just looked for BOSTON COLLEGE, CLEVELAND CIRCLE, RIVERSIDE, HEATH STREET or any short-turns therein it would make zero wayfinding difference. The letters *never* did the heavy-lifting we assumed they did because parsing full destinations was always non-optional in one direction.
 
The A was the Watertown Branch, bustituted 1969.

The letters have never been used for inbound termini, so we're already a half-century accustomed to looking at the full name on the destination sign. If there were no letters and we just looked for BOSTON COLLEGE, CLEVELAND CIRCLE, RIVERSIDE, HEATH STREET or any short-turns therein it would make zero wayfinding difference. The letters *never* did the heavy-lifting we assumed they did because parsing full destinations was always non-optional in one direction.

There's an option now, though. I guess the issue is: if you're trying to catch a Union Square train, how do you know which of the western termini to go to (and vice versa)?
 
The A was the Watertown Branch, bustituted 1969.

The letters have never been used for inbound termini, so we're already a half-century accustomed to looking at the full name on the destination sign. If there were no letters and we just looked for BOSTON COLLEGE, CLEVELAND CIRCLE, RIVERSIDE, HEATH STREET or any short-turns therein it would make zero wayfinding difference. The letters *never* did the heavy-lifting we assumed they did because parsing full destinations was always non-optional in one direction.
so you would get rid of the letters?
 
It seems simple enough. the D line goes from Riverside to Union. Display the D and the last stop of the train to show direction.
If anything the color is redundant, but it works in the context of the overall system.
 
There's an option now, though. I guess the issue is: if you're trying to catch a Union Square train, how do you know which of the western termini to go to (and vice versa)?

Definitely not by looking at the system maps in their current state. A plurality of them still show Arborway, for chrissakes. And at the rate the T cracks down on outdated signage, it might be 2050 before every stop even depicts GLX. Drop an out-of-towner mid-block on Chestnut Hill Ave. today and task them with finding which of the 3 options in two-blocks' walk gets them to North Station on a one-seat, and they're as good as fucked. You can't know without asking someone or going to MBTA.com tell which western termini maps to which inbound termini most of the time. We're already at that flaw in the wayfinding system.

So the issue is more fundamental. If you're going to placemark it on maps, you better damn well step up the enforcement on signage accuracy because that's anti-useful the way things currently stand. And if you're going to start tying the letter designations to inbound termini, you better BOTH (1) ensure accurate signage replacement, and (2) make sure the inbound termini aren't subject to changing every several years like all of Green Line history (and the magnitude of upcoming GLT streamlining) suggests they will continue to such that signage isn't perpetually staggered mid-changeout.


I mean, there are ways to do this. But half-assing won't do it, and thus a tweak of current half-assing practice isn't going to amount ot anything meaningfully better and will quite likely make things lots worse when signage versioning attrition rears its ugly head. It's not about coming up with a more-perfect Version 36 of the spider map like we're narrow-focusing on here...it's about changing institutional practice on the follow-through execution in a major way.

so you would get rid of the letters?

Get rid of OR make more thoroughly intuitive than it is right now. Something that's not a half-assing of contradictory signage that dates itself within years or require a bunch of situational asterisks when everyday occurences (shift changes, run-as-directeds dispatched anywhere, late trains getting looped early) contradict those assignments multiple times per hour. Absent a more robust scheme, I'd say less is more. We're accustomed to watching the name on the destination sign more than the letter, so that wouldn't be the worst thing in the world if they went minimalist and kept the signage errors from multiplying all over.

I don't have strong opinions on this. I just think people are waaaaaaaaaaaaaaay overestimating how much the letter designations matter in everyday wayfinding...because in one direction they haven't mattered at all in 56 years.
 
It seems simple enough. the D line goes from Riverside to Union. Display the D and the last stop of the train to show direction.
If anything the color is redundant, but it works in the context of the overall system.
But this hand-cuffs you into only allowing ~1/4 of trains serve Union. What if the system could allow for more than that? It would be better to let as many trains as possible serve the GLX while maintaining optimum headways across the system. And that means allowing trains from all four existing branches the possibility of also serving the two new branches at the opposite end.
 
The problem with depending on the terminus is that 95% of the riders dont care about it.

When youre in DC, and you want the red line, "Shady Grove" and "Glenmont" are easy to confuse because theyre places no one cares about. It just doesnt stick in the mind.

If youre directing someone to Agganis Arena, saying "take the B" is less confusing than "take the Boston College" = "wait I thought it was at BU"

Im having nightmares imagining if NYC tried that.

"Oh youre going to Times Square? Take either a Van Cortlandt Park, Wakefield, or Harlem train!"
 
The problem with depending on the terminus is that 95% of the riders dont care about it.

When youre in DC, and you want the red line, "Shady Grove" and "Glenmont" are easy to confuse because theyre places no one cares about. It just doesnt stick in the mind.

If youre directing someone to Agganis Arena, saying "take the B" is less confusing than "take the Boston College" = "wait I thought it was at BU"
It also gets confusing when you're told to "take the Riverside train" but then there are bus substitutions and all D trains become Beaconsfield trains. That Riverside train will never come. If you're told to "take the D" then this isn't an issue.
 
The problem with depending on the terminus is that 95% of the riders dont care about it.

When youre in DC, and you want the red line, "Shady Grove" and "Glenmont" are easy to confuse because theyre places no one cares about. It just doesnt stick in the mind.

If youre directing someone to Agganis Arena, saying "take the B" is less confusing than "take the Boston College" = "wait I thought it was at BU"

If it's an out-of-towner, I *never* go by the letter destination when giving directions. In the rollsign era it was a crapshoot that the train would even be showing the correct destination sign...that was useless. In the LCD era you can barely see the pixels on the destination signs without squinting yourself crosseyed. However, the ASA screens and countdown announcements clearly spell out the full name of the destination train. So..."Take the Green Line, get on a Boston College train...your stop is on that one" is the absolute clearest-most and confustion-free way to do it.
 
If it's an out-of-towner, I *never* go by the letter destination when giving directions. In the rollsign era it was a crapshoot that the train would even be showing the correct destination sign...that was useless. In the LCD era you can barely see the pixels on the destination signs without squinting yourself crosseyed. However, the ASA screens and countdown announcements clearly spell out the full name of the destination train. So..."Take the Green Line, get on a Boston College train...your stop is on that one" is the absolute clearest-most and confustion-free way to do it.

If youre at Government Center and directing a tourist to Kenmore for the baseball game, saying "take any B, C or D train" is much easier for them to remember than the termini

You could counter with "tell them to simply not take the Heath Street train" but the negative is a lot less reassuring.
 
I don't have strong opinions on this. I just think people are waaaaaaaaaaaaaaay overestimating how much the letter designations matter in everyday wayfinding...because in one direction they haven't mattered at all in 56 years.
with two termini at the other end, I think lettering is about to become important in both directions.
far easier to explain to someone getting on in copley looking to go to east somerville.
Take the train to medford
or Take the E line to medford.

Not much difference but seems important to me.
I also think there's an element of place making or ownership with each line having a distinct letter.
Anyway, probably getting lost in the weeds now.

Oh and digital maps might be a good idea re. updating, if they don't get smashed up.
 
If youre at Government Center and directing a tourist to Kenmore for the baseball game, saying "take any B, C or D train" is much easier for them to remember than the termini

You could counter with "tell them to simply not take the Heath Street train" but the negative is a lot less reassuring.
And if you had the letters as they are with numbers on the northside, then that tourist at Government Center wouldn't follow your advice and jump on a D train only to find themselves at Lechmere.
 

Back
Top