Just to give a sense what this extension would do at reshaping the bus loads at FH, the last PMT (2003) that included the W. Rox OLX in its universe of projects projected +11,300 daily Orange Line riders as a result of the new stops, but only +600 new transit riders using no current form of transit.
That proposal did spec going all the way to Route 128, so that +600 is majority from the 128/Hersey park-and-ride. And that's a very shitty bump for a line with a 128 park-and-ride, which is why I think the cost of building any further than W. Rox (especially when W. Rox-Hersey is 40% of the extension's total length) is a complete waste unless some later Phase II slam-dunk shows its there.
But this illustrates how much the buses are transformed. If 1400 daily Needham Line riders at those intra-city stops (per the '14 Blue Book) migrate to OLX, that's 10,000 diversions per day from that nine-route Washington St. concurrency between FH and Rozzie.
Here's the daily weekday boardings from the routes on the FH-Rozzie concurrency (IB+OB), plus their ridership rank out of 169 total MBTA bus routes.
#30 - 2496 (rank: 54)
#34 + 34E - 4030 (rank: 33)
#35 - 2422 (rank: 56)
#36 - 3323 (rank: 43)
#37 - 1593 (rank: 84)
#40 - 1344 (rank: 96)
#50 - 1310 (rank: 99)
#51 - 2145 (rank: 63)
TOTAL: 18,663
This route down Centre St. from FH to Bellevue is also relevant for the discussion:
#38 - 1021 (rank: 113)
Now, the diversions are not going to be 18,663 minus 10,000 because there will still be OL-to-bus transfers at the outer stops and not everybody is getting on or off at FH. You'd have to dig into the per-stop boardings and do some painfully boring math to figure that one out. The 35, 36, and 37 are the three routes that hit every OLX station and do not diverge until West Roxbury (39% of total ridership); those would have the most diversions to rail. Each only runs relatively short distance past W. Rox after they diverge, so the % of riders on those routes riding for the last few stops regardless of OLX isn't large. Again, maths and blah blah blah. Still, at pretty conservative estimates of how much diversion you're talking...it's a shitload of bus trips saved.
Maybe a quarter of the load off the Washington concurrency if you're playing it very conservative, which is still enough to do some consolidation.
So...with that in mind:
#37 - Least amount of diversion from OLX, lowest ridership of the 3 routes that duplicate the whole length of OLX. Loops around north of W. Rox station within 1/3 mile walking distance to the station. The 52 duplicates part of this route and swings close enough to your probable W. Rox OLX terminal behind Shaw's that the 52 would probably be drawn in to loop at the transfer. Whack the 37...you're covered. ELIMINATED.
#38 - Lowest ridership of all routes. Hits Bellevue then fizzes out on some side streets to the south after zigzagging around. All of the past-Bellevue stops are less than 1/2 mile walking to either Highland or Bellevue. I think you can whack the south portion entirely, then reassign one of the 35/36/37 to the Centre St. routing to cover the 38's former bread-and-butter. Probably at better frequencies than the 38 ever managed. ELIMINATED.
#35 - Longest route, middle-of-road ridership. It stays almost entirely on Centre St. from W. Rox. This is the ideal one to re-route on the 38's old Centre St. routing, to make it essentially "the Centre St. bus". Logical. PARTIAL RELOCATION, SAME TERMINI.
#51 - Very long route to Reservoir. Would probably work better and attract more ridership if shortened. TRUNCATE.
#30 - Diverges to Mattapan from Rozzie down Cummins Hwy. Short route, will hit Blue Hill Ave. Fairmount station in future in addition to FH, Rozzie/OLX, and Mattapan. Keep it as-is out of FH and upgrade this sucker because it's a hugely valuable transfer circulator. KEEP AS-IS.
#34/34E - Stays arrow straight down Washington. It's a Washington load-bearing bus and a faster trip to Dedham Mall than the 35. KEEP AS-IS.
#36 - Relatively short travel distance between W. Rox and the VA Hospital, highest ridership. Chances are this is going to have elevated numbers of handicapped passengers who need closer-spaced stops, and it needs to go through the gut of the neighborhood. Keep it as-is. Maybe increase the frequencies to offset some other losses. KEEP AS-IS.
#40 - Goes considerable distance down Washington before looping around at the south side of the cemetery. Lowish ridership, most of it probably on the Washington concurrency to Rozzie then Washington currency with the 34/34E. The 33 out of Mattapan, which has considerably higher ridership (rank: 34) meets it on the other side of the cemetery. Would have to study the numbers at the unique stops and determine whether the 33 covers the it well enough. I lean to yes because the 40 ridership is very low while the 33's very high, suggesting a demand skew to Hyde Park and Mattapan instead of Rozzie/FH. Some light reshaping of other routes might be the solution. TBD (merge routes?).
#50 - FH to Cleary Sq. Probably extended to loop at Fairmount sometime soon. The 32 is the most direct route from FH to Cleary, and is #8 rank which basically makes it a Key Bus Route. Doesn't suggest a lot of natural Rozzie-Cleary demand, and have to assume that at least some people at FH are grabbing a 50 in between 32's because it says Cleary/Hyde Park. Flanks the other side of the cemetery from the 40. That's three routes at the cemetery when the 33 is thrown in. Maybe the solution here is to combine the 40 and 50 into a loop-around and boost frequencies. If the jagged route takes too much time, terminate at Rozzie...but you'd probably support better frequencies keeping it running limited-stop on the FH-Rozzie concurrency. TBD (merge routes).
OK...so that's 3 routes on the FH-Rozzie concurrency, down from 9. One is a combined route with upped frequencies, presumably to get the formerly two separate routes out of the ridership gutter. One is a supersize mini-Ring route that should get very enhanced service. And one is the local-stop double-up of OLX. The routes that fizz out in league-average walking distance to rapid transit get eliminated. One re-route serves the highest-demand portion of one of the eliminateds and likely becomes a growth route sticking to one corridor with unique travel patterns instead of splitting the difference. One route gets a beneficial truncation that'll make its schedule more reasonable and encourage more ridership as a result. A non-FH route gets roped into a nearby OLX transfer and supplants all the unique needs of one of the eliminateds.
That's pretty well-balanced, no? You eliminated a lot of schedules on the concurrency but kept the strongest ones and strengthened some through reshaping that probably needed troubleshooting to begin with. So it's -6 routes. but say you retain two-thirds of the frequencies because the drops were generally weaklings that clogged more traffic than they brought in riders to FH.
I think that's entirely reasonable, splitting-hairs differences of opinion aside. It's a bit more dramatic change to the route lines on the detail system map, but not a drastic change in what goes where how often for the highest % of the route's riders when everything's weighted.
