Fantasy T maps

[guitarguynboston]

What he said ^^^^. I wanted to ask why not after all the facts F-Line said but I was too lazy to note it. So again after all that F-Line stated about studies and facts is there another reason than what you answered with before? Thank you for your response.

 

[Riverside]

One more map for today, NSRL featured this time because while I may disagree on the value proposition, I do agree that putting things on a fantasy map is fun. (Turns out working in Illustrator means that things like this aren't a multi-day process, this took maybe an hour or two? Turns out professionals do things for a reason.)

At this point we're detached enough from reality that a bridge is no biggie I suppose.
View attachment 50743

I'm sorry I haven't been able to reply earlier in this series, but these maps look great! Lots of fun and engaging ideas here. Random comments:

The Philosophical Challenge of "Completist" Fantasy Maps: I really love your western BLX on this map. Feasibility aside, this kind of cross-Brookline/cross-Brighton route would make for a very compelling heavy rail line. What was interesting to me, though, is that (in one of your previous iterations), this "CrossBrookBright" subway stood in contrast to the absence of the NSRL. Without getting into the comparative merits of the two projects, it feels hard to imagine a world that is so pro-transit that a CrossBrookBright subway gets built but an NSRL doesn't.

This points to a larger philosophical challenge when making a fantasy map: where do you (no pun intended) draw the line? At what point does one say, "This is a fantasy map but even x is too outlandish for a fantasy map"? One approach is to add arbitrary constraints. Another is to intentionally subvert the expectation of a "completist" map, and simply focus on highlighting some key novel ideas. Your earlier drafts were good examples of this in their illustrations of your Pink Line proposal, your support for an Urban Ring via Harvard, and your Aqua Line proposal. (Plus, the aforementioned CrossBrookBright subway.)

That's not at all to criticize these latest more elaborate versions! Far from it. Mostly I just think it becomes an interesting creative and philosophical challenge.

Green Line "fingers": I think this works surprisingly well! This is always such a pain when making a T map.

I believe my original intention was to try and isolate the surface running from the Riverside/Needham branches. I'm a little confused what you mean by keeping Arborway on the main GL though, could you elaborate or show that?

I enjoy the visual of the Arborway <> Seaport Line, though its operational considerations seem less than ideal to me. Can you explain a bit more of your thought process?

• Added the Hudson Line because why not

For funsies, the Central Mass Branch could also be a candidate for the "River Line"-style service you suggested for the Bedford Line.

• Maverick renamed to Jeffries Point because yikes. (Credit at least partially goes to @Riverside for that one although I chose a different name)

Love it! Yes, for the most part my feeling about renaming stations is less about "de-naming" stations that are named after terrible people (I prefer to think of it in the more exciting terms of "Who can we increase our recognition of?")... but Maverick was just awful, and even his contemporaries thought so.

Station Names: continue to love what you're coming up with!

 

[TheRatmeister]

Alon Levy already ballparked costs for double-tracking enough of the Reading Line

With no trackbed expansion, just reusing some of the OL triple-track and laying a bit of extra track where the alignment is already wide enough, for example between Wyoming Hill and Oak Grove, the longest single tracked segment is about 2 miles, so even with the current glacial 15mph timetabled speed between Oak Grove and North Station on the CR that's still fine for 16 minute headways. Track is not a major factor for RER.

The crossings are going to be very expensive with RER because you're capped at a 2% maximum FRA grade vs. 4.0-5.6% for Orange Line specs, effectively doubling the construction areas for each of the elevation-change eliminations at untold cost bloat.

The steeper slope can't always help the stations though, there's still plenty of cutting required for the OL. With the steeper grade the 3 Melrose stops could probably be shifted without much consequence, Wakefield could be shifted but that would put it a quarter mile walk away from all the TOD so not great, and Reading could be fine where it is but the area should really be an open space, not divided in two by a station, so a cutting is almost certainly preferable here, and Greenwood needs to be in a cutting no matter what, even with the steep grade you can't go under the first level crossings, up to ground level and then back under the next crossing.

It's all ultimately a decision between "Good service now, longer and more expensive timeline for amazing service down the road" or "Long but ultimately less expensive timeline for amazing service now."

 

[TheRatmeister]

The Philosophical Challenge of "Completist" Fantasy Maps: I really love your western BLX on this map. Feasibility aside, this kind of cross-Brookline/cross-Brighton route would make for a very compelling heavy rail line. What was interesting to me, though, is that (in one of your previous iterations), this "CrossBrookBright" subway stood in contrast to the absence of the NSRL. Without getting into the comparative merits of the two projects, it feels hard to imagine a world that is so pro-transit that a CrossBrookBright subway gets built but an NSRL doesn't.

This points to a larger philosophical challenge when making a fantasy map: where do you (no pun intended) draw the line? At what point does one say, "This is a fantasy map but even x is too outlandish for a fantasy map"? One approach is to add arbitrary constraints. Another is to intentionally subvert the expectation of a "completist" map, and simply focus on highlighting some key novel ideas. Your earlier drafts were good examples of this in their illustrations of your Pink Line proposal, your support for an Urban Ring via Harvard, and your Aqua Line proposal. (Plus, the aforementioned CrossBrookBright subway.)

My philosophy for the NSRL-less map is that I'm not really considering the local political will to build rapid transit, I'm going from basically a pure Cost-Benefit Analysis perspective. (Is this an arbitrary line? Absolutely.) Will a line generate more benefit that it adds in cost? Great, on to the map it goes. At some point I'll probably revisit the "Severe political/financial restriction" map because I do think that's in some ways a more interesting scenario.

Green Line "fingers": I think this works surprisingly well! This is always such a pain when making a T map.

The B is the only part I'm not super happy with, but space is getting quite tight so I didn't have a ton of options. One of the main arbitrary restrictions I've put on myself is that I haven't rescaled anything, all the text uses the same font size and all the lines are the same thickness as the official map. There's a couple places where this is really frustrating but for the most part it's actually fine, surprisingly enough. (The other arbitrary restriction that lasted up until I did Teal Line to Ruggles was no 90º angles on rapid transit lines.)

I enjoy the visual of the Arborway <> Seaport Line, though its operational considerations seem less than ideal to me. Can you explain a bit more of your thought process?

Both parts will require some surface middle of the street running, best to pair them together and avoid hurting either Riverside or Needham reliability more than sharing the Huntington Ave subway already will. It also avoids taking frequency away from the Boylston-North Station segment for the Needham/Riverside branches, which will probably be a lot more popular than the Seaport segment. It's also a pretty easy transfer so it's not exactly a severe drawback to this service pattern.

For funsies, the Central Mass Branch could also be a candidate for the "River Line"-style service you suggested for the Bedford Line.

I think it's better if it's full CR because that lets you use both branches to provide better frequency between Weston, Waltham, Belmont, and Cambridge.

Station Names: continue to love what you're coming up with!

Thanks!

 

[Riverside]

My philosophy for the NSRL-less map is that I'm not really considering the local political will to build rapid transit, I'm going from basically a pure Cost-Benefit Analysis perspective. (Is this an arbitrary line? Absolutely.) Will a line generate more benefit that it adds in cost? Great, on to the map it goes. At some point I'll probably revisit the "Severe political/financial restriction" map because I do think that's in some ways a more interesting scenario.

Yeah, I like this approach.

Both parts [Arborway + Seaport] will require some surface middle of the street running

Hmm, interesting. So for the Seaport, are you referring to the Eastport <> Design Center segment? To me, that doesn't really seem comparable to the 2+ miles between Huntington and Forest Hills. Saddling the Seaport exclusively with something similar to the B Line seems less ideal to me.

(Also, I don't think it's ever been discussed much, but if we ever do see SL2 converted to LRT, I don't think that the entire Drydock loop will get converted. My guess is that there would be a new off-street terminus built near the intersections of Tide St, Northern Ave, and Drydock Ave -- maybe through a conversion of 24 Drydock Ave, or by grabbing some of the parking lots along Drydock Ave. Or, hell, you could massively renovate the Design Center building itself and plop the station in there.)

 

[TheRatmeister]

So for the Seaport, are you referring to the Eastport <> Design Center segment? To me, that doesn't really seem comparable to the 2+ miles between Huntington and Forest Hills.

D St to Drydock is about a mile, it's not exactly short. And what it lacks in distance it certainly makes up for in being located in an area designed around car dependency.

(Also, I don't think it's ever been discussed much, but if we ever do see SL2 converted to LRT, I don't think that the entire Drydock loop will get converted. My guess is that there would be a new off-street terminus built near the intersections of Tide St, Northern Ave, and Drydock Ave -- maybe through a conversion of 24 Drydock Ave, or by grabbing some of the parking lots along Drydock Ave. Or, hell, you could massively renovate the Design Center building itself and plop the station in there.)

Black Falcon Ave is almost 3k feet end-to-end, I think that's enough to warrant multiple stops, especially if density increases as dramatically as it has around Fan Pier. My headcanon SL2 LRT has one stop at the Design Center plaza thingy and one stop at the eastern end of Black Falcon Ave.

 

[Riverside]

D St to Drydock is about a mile, it's not exactly short.

Oh I wasn’t measuring from D St — Seaport LRT would all but certainly have a dedicated ROW at least to the current Silver Line Way.

More thoughts on Design Center later — I’m still not sold, but I do see where you’re coming from, and it’s an interesting comparison!

 

[WarpedReality]

At this point we're detached enough from reality that a bridge is no biggie I suppose.

Might as well add Northborough to the Framingham/Worcester line too, I guess

 

[Equilibria]

One more map for today, NSRL featured this time because while I may disagree on the value proposition, I do agree that putting things on a fantasy map is fun. (Turns out working in Illustrator means that things like this aren't a multi-day process, this took maybe an hour or two? Turns out professionals do things for a reason.)

At this point we're detached enough from reality that a bridge is no biggie I suppose.
View attachment 50743

Not that I think this requires a change, but as long as I'm nitpicking your Blue Line routing, I wonder why a tunneled route (which this would be) would follow a meandering right-of-way that crosses the river twice and doesn't really hit commercial or population centers? If we're doing fantasy here, I'd probably just suppose a cut-and-cover of Route 20 between Waltham and Watertown Square - the Pleasant Street corridor has some multifamily development but it's still mostly light industrial and low-density commercial. In that version, you'd have stops at Linden Street where it crosses the Fitchburg Line ("Bleachery/Bentley"), Gore Place, and somewhere around the intersection of Main and Waverly/Myrtle ("Myrtle", since "Waverly" is taken). On your map, if you wanted to you could make this change by only changing the station names.

You could actually use the same logic in the other direction and follow 20, then 117, then the Mass Central ROW to Weston. It misses Brandeis but hits more currently unserved density around Banks Square, which would be the intermediate stop. Brandeis folks could still get to a Banks Square stop pretty quickly, and Brandeis/Roberts could remain on Urban Rail.

Another quibble (I could just make my own map, I know ): Does it make it too busy to add thin blue lines for the Massport shuttle to Back Bay, Copley, and North Station? Also, should Copley and Back Bay have a walking line?

 

[TheRatmeister]

Not that I think this requires a change, but as long as I'm nitpicking your Blue Line routing, I wonder why a tunneled route (which this would be) would follow a meandering right-of-way that crosses the river twice and doesn't really hit commercial or population centers?

I talk about it a bit more in this previous post, but if you go via Pleasant Street you can actually do most of this sergment either above ground or in a shallow cutting along the old railway alignment. I'd also say this it does hit the major population/commercial centers of Watertown Sq and Waltham Center regardless of which path you take. I wouldn't really consider Rt 20 to fall into that bucket, but the Arsenal and the Chemistry both could with some existing and new redevelopment.

Essentially, digging up the streets of Brookline and Brighton is probably not in the cards, nor necessary to get to the 50% number.

A route would be something like:

• Starting from Charles/MGH, dig up Storrow, dig a trench, then cover it over
• At Kenmore dive under the GL for a deeper station before surfacing and reusing the GL-D alignment to Fenway
• Dive into the main bored tunnel, hitting Coolidge Corner, Warren St, and Brighton Center (About 3 miles)
• C&C/Capped Cut along Arsenal St to Watertown Sq and then the Watertown Branch Alignment where possible, Pleasant St where not (Another 2 miles or so)
• Run in a shallow uncoverd cutting (Or elevated if you wanted to) along the rail alignment over the Charles and through the Chemistry
• Dive into one more deep(ish) section under Waltham Station before resurfacing to follow the Fitchburg Line to Brandeis/Roberts

Or alternatively instead of going via the Fenway/Longwood, Coolidge Corner, and Brighton dig a subway under Comm Ave (Maybe bury this bit of the B while you're at it), then surface at West Station and continue elevated above the Pike to North Beacon St and then pick up the route from there. This could let you go all the way from Charles/MGH to Brandeis/Roberts with no major bored sections, but missing Brighton and Coolidge Corner makes it not as useful of a connection closer in.

Does it make it too busy to add thin blue lines for the Massport shuttle to Back Bay, Copley, and North Station?

I'd lean towards yes, but I might try it. I have also now realized that NSRL makes the North Station Logan Express mostly irrelevant, I'll probably remove that on the next iteration.

Also, should Copley and Back Bay have a walking line?

The current arrangement of things makes that... difficult. I'm also not sure if that's a good connection to incentivize for crowding concerns. Arlington-Bay Village might be better in that regard. (And also easier to map.)

 

[F-Line to Dudley]

With no trackbed expansion, just reusing some of the OL triple-track and laying a bit of extra track where the alignment is already wide enough, for example between Wyoming Hill and Oak Grove, the longest single tracked segment is about 2 miles, so even with the current glacial 15mph timetabled speed between Oak Grove and North Station on the CR that's still fine for 16 minute headways. Track is not a major factor for RER.

It is a factor if you plan to interline it with the Fairmount Line via NSRL and hope to achieve something resembling SUAG (i.e. better than :15) on the Fairmount end, though. As a distinct "Indigo" pattern the Reading appendage weighs down the Fairmount appendage vs. the Fairmount appendage's service ceiling in an NSRL universe.

The steeper slope can't always help the stations though, there's still plenty of cutting required for the OL.

You'd have stations with 450-800 ft. platforms on RER bookended by max 2% grades on each discrete crossing elimination with depth of 17-18 ft. below each cross street. Versus 300 ft. platforms on OL bookended by 4% grades on each crossing elimination at depth of 12 ft. That's a hell of a big difference in length of cutting and size of construction areas. A piecemeal elimination on RER then figuring out OLX later is going to cost you a whole lot more, enough that ripping the band-aid is a very legit and very fraught decision. Don't minimize how big a cost difference that is.

With the steeper grade the 3 Melrose stops could probably be shifted without much consequence, Wakefield could be shifted but that would put it a quarter mile walk away from all the TOD so not great, and Reading could be fine where it is but the area should really be an open space, not divided in two by a station, so a cutting is almost certainly preferable here, and Greenwood needs to be in a cutting no matter what, even with the steep grade you can't go under the first level crossings, up to ground level and then back under the next crossing.

Reading would not involve a crossing elimination. The station, per the 1970's OLX proposal, would be shifted east to stub-out at the Ash St./Main St. crossing, with the adjacent storage yard occupying the long 3-track section towards Route 128. Hell...moving the station to Main has been floated as a current solution for the problem of installing full-highs around the historic depot building. The Main St. location, in addition to being more TOD'able, is closer and more direct by walking distance to dead-center downtown at Town Common than Reading Depot is. So that saves 3 crossing eliminations right there. Subtract New Crossing St. in Reading altogether because driveways to Ash St. from the buildings south of the crossing totally eliminates the need for a thru street; you could whack that one today with some town-level action. And make Broadway St. in Wakefield a road-over-rail overpass because of the long runup space both sides on the road side. That leaves only 9 crossings where the rail would need to change elevation, which is a considerably better prospect than treating 13 or 14 crossings.

It's all ultimately a decision between "Good service now, longer and more expensive timeline for amazing service down the road" or "Long but ultimately less expensive timeline for amazing service now."

Again...at Orange headways this outslugged the ridership of RLX-Hanscom (not even Arlington) handily the last time it was studied. So "amazing" SUAG service being > than "good" SUAW service is a pretty big get and pretty big threshold cleared. You have a very tough road to hoe trying to incrementalize this because the costs of each elimination on the RER mode. On a completists' map its absence sticks out like a sore thumb given some of the more speculative and extremely expensive builds that did make the cut.

 

[TheRatmeister]

Reading would not involve a crossing elimination. The station, per the 1970's OLX proposal, would be shifted east to stub-out at the Ash St./Main St. crossing, with the adjacent storage yard occupying the long 3-track section towards Route 128. Hell...moving the station to Main has been floated as a current solution for the problem of installing full-highs around the historic depot building. The Main St. location, in addition to being more TOD'able, is closer and more direct by walking distance to dead-center downtown at Town Common than Reading Depot is. So that saves 3 crossing eliminations right there. Subtract New Crossing St. in Reading altogether because driveways to Ash St. from the buildings south of the crossing totally eliminates the need for a thru street; you could whack that one today with some town-level action. And make Broadway St. in Wakefield a road-over-rail overpass because of the long runup space both sides on the road side. That leaves only 9 crossings where the rail would need to change elevation, which is a considerably better prospect than treating 13 or 14 crossings.

True, but this also works for an RER line. If we're fine with the ride being a bit of a roller coaster there is still a big advantage to the OL for cutting length and station siting, but if we want to keep things relatively flat and not make a line resembling a wave then the difference goes down to 7500ft from longer slopes. Station sites depend on just how NIMBY Melrose and Wakefield want to be. If the answer is very, then it will probably be cuttings for all the stations (except Reading), if it's almost not at all, and we don't mind a 300ft push for the Melrose stops, then it's probably just Greenwood and Wakefield. The former makes the difference between RER and OL smaller, the latter doesn't.

It is a factor if you plan to interline it with the Fairmount Line via NSRL and hope to achieve something resembling SUAG (i.e. better than :15) on the Fairmount end, though. As a distinct "Indigo" pattern the Reading appendage weighs down the Fairmount appendage vs. the Fairmount appendage's service ceiling in an NSRL universe.

Having remeasured the Oak Grove-Rivers Edge segement it's more like 1.6 miles, I'm not sure where I got 2 miles from. That brings the minimum headways with the existing trackbed layout to 13 minutes. If we sacrifice a small part of the DTA's parking lot it goes down to 1.1 miles for minimum 9 minute headways without rebuilding Malden Center/Oak Grove stations or double-tracking that segment. I still think this is a non-issue, frankly.

Again...at Orange headways this outslugged the ridership of RLX-Hanscom (not even Arlington) handily the last time it was studied. So "amazing" SUAG service being > than "good" SUAW service is a pretty big get and pretty big threshold cleared

The study wasn't (AFAIK, please correct me if I have misunderstood) comparing ridership between OL with 5-10 minute headways and RER with 15 minute headways, it was comparing OL with the existing (terrible) CR service. Obviously that's a favorable comparison, but how much of the projected OL ridership would be captured by 15 minute RER on the existing infrastructure? 50%? 80%? 90%?

 

[F-Line to Dudley]

The study wasn't (AFAIK, please correct me if I have misunderstood) comparing ridership between OL with 5-10 minute headways and RER with 15 minute headways, it was comparing OL with the existing (terrible) CR service. Obviously that's a favorable comparison, but how much of the projected OL ridership would be captured by 15 minute RER on the existing infrastructure? 50%? 80%? 90%?

The number of all-new transit riders attracted by OLX-Reading was third-highest amongst the 7 HRT extensions that the 2003 PMT evaluated, behind only the two BLX extensions at between 61-68% of the individual BLX's. It's almost double the next-highest (RLX-Weymouth), and way more-than-double RLX-Hanscom and OLX-Westwood. Lynn-Salem is a Commuter Rail corridor that already has de facto :30 frequency Regional Rail schedules in addition to some bus overlaps, and yet the difference between that and a Blue Line frequency was worth over half of the HRT extension's new ridership. The capture by doubling the RER frequency to Urban Rail doesn't even begin to scrape the demand there. Proportionately OLX-Reading traces the midpoint in the evaluated extensions. It cleaves to the higher end in total riders, and legitimately upper-crust in all-new riders. If scaling of CR frequencies was going to satiate all the demand, we'd either see the BLX's performing worse on all-new transit riders or the southern extensions performing much better on ridership generation. It of course needs to be re-studied vs. :15 Urban Rail, but there appears to be a lot of *oomph* still being left on the table.

And look at how much it outslugs RLX-Hanscom on both counts. Your map doesn't even bring Red past Arlington Heights. I don't think there's any disagreement that Arlington Heights is a good solid get, so if that makes the cut amidst all the other crayon mania...why not Reading?

Extension	On-mode new riders (rank)	Total new transit riders (rank)
BLX Wonderland-Lynn	21,000 (1)	7,900 (2)
BLX Lynn-Salem	15,500 (2)	8,900 (1)
OLX Forest Hills-Needham/128	11,300 (3)	600 (7*)
OLX Oak Grove-Reading	9,400 (4)	5,400 (3)
RLX Braintree-South Weymouth	6,700 (T5)	2,900 (4)
RLX Alewife-Hanscom/128	6,700 (T5)	2,000 (T5)
OLX Forest Hills-Westwood/128	4,700 (7)	2,000 (T5)

*large weighting to bus diversions/consolidations

 

[TheRatmeister]

Lynn-Salem is a Commuter Rail corridor that already has de facto :30 frequency Regional Rail schedules

30 minute headways is a lot different than 15 minute headways though. It also needs to be said that Reading is quite a bit different from Lynn, with a population density more than 3x lower and a per-capita income that is 2x higher. I'd be extremely wary of comparing the two cases.

I'm also more than a bit skeptical of the 2003 ridership projections. Take RLX to South Weymouth, which implies that 3800 people per weekday who already take the MBTA would use the station. This seems rather unlikely since 3800 is about the same as the total 2022 weekday ridership of the 226 and South Weymouth CR station combined with the total number of parking spaces at Quincy Adams and Braintree. This feels like a stretch. The OL to Reading numbers suggest 4000 existing transit riders, which is again about the total of all 137 weekday riders, Haverhill Line riders (Between Reading and Melrose), and parking spots at Oak Grove/Wellington.

 

[F-Line to Dudley]

30 minute headways is a lot different than 15 minute headways though.

And 4.5 is a lot different from 15. What's your point?

I'm also more than a bit skeptical of the 2003 ridership projections. Take RLX to South Weymouth, which implies that 3800 people per weekday who already take the MBTA would use the station. This seems rather unlikely since 3800 is about the same as the total 2022 weekday ridership of the 226 and South Weymouth CR station combined with the total number of parking spaces at Quincy Adams and Braintree. This feels like a stretch. The OL to Reading numbers suggest 4000 existing transit riders, which is again about the total of all 137 weekday riders, Haverhill Line riders (Between Reading and Melrose), and parking spots at Oak Grove/Wellington.

You're picking nits with the wrong ones. RLX-Arlington Heights is the weakling of the bunch, as duly noted. If that one is a consensus pick, and crayon fever dreams are strewn over the whole map...why in the hell are we pooh-poohing and over-squinting at the very substantial and many-times-over studied ridership for a rather straightforward conversion like Reading? The cognitive dissonance in expending all that energy is very confusing to an outside observer.

 

[TheRatmeister]

And 4.5 is a lot different from 15. What's your point?

Warning: Hypothetical and arbitrary numbers incoming
A TPH vs ridership graph will look something like this. If the right side of this graph is 1 minute headways at y=14, and the left side is 2 hour headways or something at y=0, where are 15 and 5 minute headways? If they're at y=7 and y=12.5 for example, then OL is a big improvement for ridership. (Almost 100%, in fact) But what I suspect is that it's more like y=11 and y=12.5, which is only around a 15% in ridership. Is this based on anything? No, not really. It's ultimately speculation. Until this is actually properly studied we won't really have an answer and your guess is as good as mine.

You're picking nits with the wrong ones. RLX-Arlington Heights is the weakling of the bunch, as duly noted

I can pick nits with this one too. Peak 77 ridership was around 4k riders per weekday, more than 2/3 of the projected total all the way to Hanscom. That one seems quite low, particularly on the Arlington segment. (OLX to Reading is also more than twice as long as RLX to Arlington Heights, so I would certainly hope it gets more riders.)

 

[Delvin4519]

A thought exercise:

What happens to the frequent grid if all the mainline RR ROWs carried rapid transit service in the inner core?

Some of the frequent surface routes would be re-allocated onto other routes within the system. For example, the 32, 35, and 36 currently all duplicate rail corridors due to the lack of rapid transit in the south west of Boston. If West Roxbury and Hyde Park had rapid transit, the 32 and 36 would all be scrapped, and instead the 30 and the 34 become the high frequency routes.

A thought test of what a fantasy high frequency grid of transit routes in Boston would look like. I went through and did a rudimentary run of cannabalizing a handful of duplicated frequent routes from the BNRD map, onto a new map that extends rapid transit; and re-allocating frequencies into other surface routes.

There are a few possible corridors not on this map that I wasn't so sure whether to add. Some corridors are Arlington - Medford Sq, the 119 in Revere, an Arlington - Mt. Auburn/Harvard crosstown, or a 92+95 combo route.

The map on the bottom right is a rough estimate of what Boston's frequent bus and subway network would look like if it had similar frequencies to some similarly sized city like Amsterdam. The rail corridors outside of the inner core are placeholders if some of them don't make any sense (those not covered by any MBTA routes are simply shaded as a very thin black line with no frequency marker, or I simply couldn't be bothered to include them). I simply tried testing what happens if every other train was through routed (essentially cut headway frequencies in half outside of each terminal). It was mostly an experiment/test, so outer core rail frequencies should not be taken at face value as actual "dream/fantasy headways".

For a comparison to today's pathetic service levels, see here. I actually made a similar map (but I had no idea what I was doing at the time), back in mid 2022 after the original BNRD map was posted, but the small size of the subway system really destroys how many high frequency surface routes can exist, since buses have to cover the RLX to Arlington, BLX to Lynn, OLX to West Rox/Wyoming Hill, BLX/GLX to Waltham, BLX to Newton Corner, GLX to Porter, GLX to W. Medford, the existing RR ROW to Hyde Park, etc., 5 RL infills between JFK UMass and North Quincy, etc., etc., etc.

With rapid transit on all of those corridors, the frequent network would essentially shift to cover other parts of the network. Take a look at Lynn and Roslindale on this fantasy frequency map compared the frequency map of today.

I've left the Quincy/Waltham/Lynn/Newton terminals as is without rapid transit extensions to Braintree/Swampscott/W. Newton, so the frequencies on this fantasy map kinda leave room for improvements. It was intentional, since the surface routes still have to cycle back to the main city center terminal, instead of hitting the first rapid transit station. Essentially, my thought is that rapid transit extensions would work best if they can eat away at excess bus running. I'm still not sure how to find info whether demand is greater a grid based system or hub & spoke for these terminals. I can make a 2nd dream fantasy map adding more frequent corridors. if it is so desired.

 

[Riverside]

There are a few possible corridors not on this map that I wasn't so sure whether to add. Some corridors are Arlington - Medford Sq, the 119 in Revere, an Arlington - Mt. Auburn/Harvard crosstown, or a 92+95 combo route.

Now this is a fantasy T map. Fantastic stuff!

 

[Teban54]

I can pick nits with this one too. Peak 77 ridership was around 4k riders per weekday, more than 2/3 of the projected total all the way to Hanscom. That one seems quite low, particularly on the Arlington segment. (OLX to Reading is also more than twice as long as RLX to Arlington Heights, so I would certainly hope it gets more riders.)

Given than F-Line's summary table shows a difference of 4,700 between "on-mode new riders" and "total new transit riders", that corresponds quite well to the 77's ridership (at least on the same order of magnitude, if not the exact same numbers in 2003). The induced demand may be too low, but at least it's not inconsistent with the 77's numbers.

But I think F-Line's main point is that ridership projections for RLX and OLX should have similar methodologies. If you think the induced demand projections for RLX are too low (which I may actually agree with), then the induced demand projections for OLX should also be too low. This would make OLX even more attractive and its omission from the map even harder to explain.

--------------------------------------------------

On the broader discussion of OLX-Reading vs. other extensions, I have some thoughts -- or rather, questions -- for @TheRatmeister.

If I'm getting this right, your core argument for omitting OLX to Reading (and possibly others) is that, regardless of its own costs and benefits as a rapid transit extension with "great" service, there exists an alternative (regional rail) with "okay" service at lower costs, and its existence stops OLX from making the cut. Did I get it right?

If that's indeed the case and your only criterion for inclusion on the map, that does raise questions about your BLX idea -- because the same argument (existence of an okay alternative) can be applied even better to Coolidge Corner, Brighton, Watertown and Waltham.

• Coolidge Corner: The C already exists
• Brighton Center: Streetcar restoration (not without challenges, but much easier than an expensive TBM tunnel)
• Watertown: Watertown Branch restoration, or LRT conversion of the 71
• Waltham: Regional rail via Fitchburg Line

All of them are less ideal than BLX with 5-min frequencies, but they're as "okay" as regional rail to Reading. As for cost, my hypothesis is that all these combined will cost much less than fully grade-separated BLX. (If the argument for BLX is that it serves more people, F-Line has already explained in detail why OLX to Reading has high projected ridership.)

And while I really don't want to go there, the same also applies to other proposals on the map like Urban Ring. Going east from Harvard is expensive no matter what, and particularly so via Cambridge St. There are surely other less costly alternatives that approximate its job reasonably well, most of which present similarly compelling arguments as Reading Regional Rail. Same for tunneling north of Chelsea.

Perhaps another possible explanation is that your map focuses on otherwise unconventional ideas, but that's not the impression I got from earlier conversations, either.

 

[TheRatmeister]

If I'm getting this right, your core argument for omitting OLX to Reading (and possibly others) is that, regardless of its own costs and benefits as a rapid transit extension with "great" service, there exists an alternative (regional rail) with "okay" service at lower costs, and its existence stops OLX from making the cut. Did I get it right?

I'm assuming that, at the end of the day, fully grade separated RER through-running onto NSRL is an equivalent alternative to OL service. So the options are:

• Start a long construction process, end up with an OL extension with amazing after 5-10 years.
• Electrify the line with overhead catenary now, end up with a line with great service much quicker, and then spend the next 5-10 years slowly eliminating grade crossings to elevate the service to amazing. Yes, this works out more expensive in the long run, but it yields much more immediate returns compared to the OL alternative. It's not letting "best" get in the way of "better." (And as a nice bonus the RER approach gets you express service along the current OL corridor without sacrificing service to Sullivan, Assembly, Rivers Edge, etc.)

Although I realize that potentially the biggest problem here is that there is a very significant temporal element not really present elsewhere that isn't really easy to convey with a static map.