Hacker News Reader

Comment by woodruffw | Hacker News Reader

First: I agree with the overall conclusion of the article. We absolutely should invest to make public transport faster, and thereby attract more riders.

Second: At least for NYC, some of this data is slightly misleading. NYC has (at least) three public transport networks: the bus and subway within the city’s borders, and commuter rail between the city, Long Island, and the rest of NY/NJ/CT. It would probably be most interesting (and correct) to compare car commuters that could take each; my intuition is that the subway generally outperforms driving within the city, the bus slightly underperforms versus driving (since it gets stuck in traffic, and there’s no meaningful enforcement of bus lanes). I don’t have an intuition for the commuter rails.

View on HN

We should invest in making public transit more consistent and frequent rather than faster. If something comes every 5 minutes… you don’t need to worry about schedule and the expected wait time from end of task until end of transit is lower.

Plus if you have transfers, it really drives down the time.

If the segment is really fast but only happens hourly, now you have to plan :( If it only happens hourly AND the timing is inconsistent or segments get cancelled, now you just have to possibly wait for hours.

This is complicated by the physical reality that different trains can't occupy the same physical space concurrently. One line coming more frequently displaces another.

There is a fair deal of complexity to this, that the common populace seems to assume isn't being implemented already. Transit is a very complicated prediction/optimization problem

Most transit (in the US at least) is buses. If you can have buses arriving every 5 minutes, it’s night and day from every 15 or especially every 30. I don’t think issue if overlapping lines applies to buses.

In SF, reaction to pandemic ridership decline was to decrease frequency of most transit. This also unfortunately makes transit less appealing :(

> Transit is a very complicated prediction / optimization problem.

It also depends what you’re optimizing for - is the goal to increase total users of system? Ensure there aren’t empty trains / buses? Have ridership “pay for” operations?

Because having your light rail continue into late night even if sparsely used means people can rely on it as an option. Frequent bus routes even on lesser used routes still can make the overall system more likely to be used.

> We absolutely should invest to make public transport faster, and thereby attract more riders.

People don't use public transport because it is faster, but because it is consistent, it is cheaper than many other options and it reduces the stress of commuting. Being a passenger is much more relaxing than being the driver. Not having to find a parking spot is liberating for your mind. Not getting fines is good for your finances.

Public transport doesn't attract more people because almost everywhere is at peak.

NY subway, for example, moves 5.5 million people a day (the peak capacity is slightly above 6) but it is harder to make new lines or increase the number of rides, while with cars you just ride yours to the highway and stay stuck there in traffic for as long as needed. You can't stack trains like that or the delay of a single ride will cascade to the entire rail line possibly for many hours.

People who prefer cars IMO are either outside of the public transport coverage, bad at planning or anxious.

It still has to be fast enough. It also needs to evolve beyond just the hub and spoke model that expects everyone to commute to the big city from the suburbs.

There is need for lot more circular routes. Connecting the spokes to each other.

> It still has to be fast enough

Fast enough is relative.

I'm European and don't know the USA situation very well, but IMO consistency beats speed every time.

The route between the city of Monza and Milan in Lombardy takes 10 minutes by train and 40 by car (not including parking)

People take the car because that train line is not reliable and trains are often canceled last minute.

In theory the train is much faster, in practice it's a Russian roulette.

Reply on news.ycombinator.com