The Rail Gap Weekend: Two Days of Buses and the Walks Between Them

The last passenger train left this valley long before I started keeping a field notebook. Freight still rolls through Boise at night, but the platforms that once served riders are gone. Planners call a region like this a rail gap, a place where the tracks exist and the passenger service does not.

I wanted to see what a weekend trip looks like when you stay on the ground anyway. The rules were plain: two days, regional buses only, no car, no rides from friends, and not a single flight. Wherever the schedules refused to meet, I would close the distance on foot and write down the cost in minutes.

The loop ran west through two river towns, with an overnight stop and a different pairing home on Sunday. I carried one bag, a printed set of timetables, and a stopwatch log that I filled at every curb.

The method

The hypothesis was narrow enough to fit on an index card. Published transfer gaps in a region without passenger rail survive contact with real arrivals only above a certain size, and I wanted that number. A transfer gap, for this trial, means the buffer between a scheduled arrival and the next scheduled departure.

I sorted every gap on the plan into three classes before leaving the house. Gaps of forty minutes or more counted as generous, gaps between fifteen and forty counted as workable, and anything under fifteen counted as fragile. The plan held seven scheduled connections across two days, and I refused to pretend any of them was certain.

Two of the links on the map were never bus connections at all. In those places the systems do not share a stop, so the timetable simply ends at one curb and resumes at another. I measured those walking gaps in advance, planned them at 3.1 miles per hour, and treated them as scheduled segments with their own departure times. That framing mattered more than any other single decision on the sheet. A walk with a departure time is a leg of the trip, while a walk without one is just an apology.

I also set a collapse rule before the first departure. If a missed connection left more than ninety minutes until the next bus, walking became the primary plan rather than the backup. The pack stayed under thirteen pounds so that rule would stay honest at mile two.

Field notes from the loop

Day one

The first bus left the downtown hub at 7:40 on Saturday, four minutes behind its card. The opening transfer, in a college town half an hour west, carried a published gap of twenty four minutes. We arrived six minutes late, the connecting coach loaded slowly, and the gap held with room to spare.

The second link was a walking gap of 1.3 miles between two systems that have never met. The route followed a signed sidewalk for the first mile, then a gravel shoulder past a seed warehouse. I covered it in twenty seven minutes, slower than the plan, and reached the far stop with nine minutes of margin.

The third connection was the fragile one, twelve minutes on paper, and it failed exactly as labeled. My bus came in sixteen minutes down, I watched the connection pull away from two blocks out, and the next departure sat two hours off. The collapse rule took over, so I walked 2.4 miles into the river town along a canal path and an irrigation road. The walk took fifty two minutes, and it beat standing at a signpost by more than an hour.

I stopped in that river town for the night, wrote up the log over dinner, and reset the sheet. Day one closed with three buses ridden, two gaps walked, and one connection lost by four minutes.

Day two

Sunday service runs thinner across the whole valley, and the first usable departure did not leave until 9:55. The morning bus ran on time to the junction town, where a fifty five minute published gap felt generous and proved merely workable. Roadwork closed the sidewalk on the direct block, the detour added 0.4 miles, and the posted board at the stop disagreed with my printed timetable by ten minutes. The board turned out to be right, which cost nothing that day and would have cost everything on a fragile gap.

The middle of the day held the longest planned walking gap of the return, 1.1 miles between a grain elevator stop and a park and ride lot. One departure left the curb two minutes ahead of its card while I was already seated, and I logged it as a warning rather than a problem. The last two connections held at eighteen and thirty one minutes, and the final bus reached downtown Boise at 6:12.

What broke

Every trial sheds parts, and this one shed them in predictable places. The honest list from the notebook reads like this.

• The fragile gap. Twelve minutes of buffer was decoration rather than a connection, and it failed on the first afternoon.
• The printed timetable. One posted board disagreed with the spring printing by ten minutes, and the board was the accurate source.
• My pace estimate. I planned 3.1 miles per hour and averaged 2.8 with the pack across sidewalk, shoulder, and gravel.
• Stop identification. One transfer point was a bare signpost with no shelter, no posted times, and no name I could match to the schedule without asking the driver.
• Sunday assumptions. The thinner schedule pushed the whole return about ninety minutes later than the same loop would have run on a Saturday.

The numbers

The log transcribed cleanly on Monday morning, and the totals carry the story better than adjectives would.

• Bus segments ridden: eight, split across two regional systems
• Scheduled connections: seven, with six held and one missed by four minutes
• Median published buffer: twenty four minutes, against a median usable buffer of eighteen
• Average arrival against the card: six minutes behind, with a worst case of sixteen
• Miles by bus: seventy six, and miles on foot: 5.2 across four separate gaps
• Longest walk: 2.4 miles in fifty two minutes, unplanned
• Pack weight at the door: twelve pounds nine ounces

One number deserves a flag before anyone copies this plan. The six minute average lateness sounds mild, but it consumed a quarter of the median buffer on its own. On a sheet built from fragile gaps, that quiet average would have broken the weekend twice.

The verdict

The rail gap weekend worked, and I want to be precise about why it worked. It held together because walking was a scheduled mode with its own departure times, never a rescue plan improvised at a curb. It held because I treated every buffer under fifteen minutes as fiction and planned nothing important behind one. And it held in April, in mild weather, on a valley floor where every gap offered a walkable shoulder or a canal path.

The caveats deserve equal billing with the result. Sunday service was the real constraint, winter would rewrite the walking math entirely, and a different region could put a highway where my canal path sat. Small systems publish thin data, so check the physical board at the stop and trust it over any printout. If I run this loop again, I will set a twenty minute floor under every connection and keep the collapse rule exactly as written.