It takes about 10-15 minutes in a car.

Can we please dispense with the canard that it is somehow the general case that Muni lines stop at the same block twice in more than one handful of blocks? It is absurd and foolish to assert that the end point extrema represent the general case.

Lines do not stop every block in the Richmond where the blocks are shortest. Lines do not stop every block in the Haight where blocks are mid sized.

If there are 20 vehicles serving the #9 line, then a 5% cut of the length of a trip would free up one vehicle.

I argued for a 22L but the MTA balked claiming that Fillmore was too narrow for vehicles to pass. I guess it would be too much for a local to remain pulled into the stop long enough for a limited to pass, similar to how trams in Amsterdam share what is in effect single tracking to cross narrow bridges.

As far as connections go, the limited lines all make the connections. But how many stops on the local could you cut without moving more than 1/2 of the way to the stop frequency of the limiteds?

-marc

"Is there evidence that a 20% cut in the number of stops translates into a 20% increase in speed? Or do the added boardings at fewer stops consume some of those speed savings?"

Indeed the evidence is that time savings will always be less that the amount of stops cut. This makes sense because cutting 100% of stops does not result in a 100% time savings. But Muni's own studies back that up with hard numbers. According to the MTA, cutting 8% of stops on the 9 would save 5% along the entire length of the trip.

Here's the quick back of the cocktail napkin math:

Assume that 70% of stops are too close. I'd wager that is a rosy figure.

Some 50% of those 70% are either transfer points or community facilities where stops cannot be eliminated.

That leaves 35% of stops to be reconfigured.

What would be your consolidation rate for the remainder of those stops?

If you were wildly optimistic and were able to cut 50% of those 35% of stops--my bet is that it would be more like 33%--then you'd be eliminating 17% of stops.

To what extent would a reduction in stops capped at, say, 20% best case, translate into an saved run on any given line?

Is there evidence that a 20% cut in the number of stops translates into a 20% increase in speed? Or do the added boardings at fewer stops consume some of those speed savings?

You learn quickly that the first 80% of an engineering problem takes 20% of the resources, and the last 20% takes 80% of the resources. If there were elegant solutions to these problems, they would have already been identified.

Wait, the elegant solution has been identified--its called express and local service and NYC figured this out a century ago.

-marc

I agree that the limited service can be faster than the local service under certain conditions.

Are you suggesting that we only run limited service because you feel that extra distance to a stop is worth the speed up? Do the needs of others who might not be so mobile need to be considered? What about the NYC MTA which runs both local and express service which compliment one another?

If we are going to eliminate lines, then we are going to make folks walk further to a line than currently. If we are going to eliminate stops on a line, then folks will have to walk further to a stop once they have walked further to a line.

If the City removed sufficient existing facilities which made Muni accessible to people with disabilities, then do you think it is worth a class action lawsuit that the City is violating the ADA by removing facilities?

Stops in lines need to conform with priorities, such as making connections to other lines and serving community facilities. Please, if this is so simple and so obvious, why has the MTA, which has been talking this up for years, been able to put forth a proposal that outlines stop reduction in concrete terms?

Again, nobody is contesting the notion that lowering load on a system makes it function more efficiently. But this belies the fact that the purpose of systems is to carry loads.

Healthy, abled people are the ones trumpeting the value of getting people from 48th avenue to the Financial District in 33 min or less, they have been for a decade now. The same people who pushed for Prop E and who will not concede that it was a failure are pushing this. But despite the best efforts to cleanse the population of San Francisco using a program of economic eugenics, there are still seniors and disabled here who are transit dependent.

Just like cyclist advocates felt that it was acceptable to slow down certain Muni lines in order to complete the bike network, even though they curiously supported a plaza which introduced new dangerous conditions as it disconnected the bike network, it is not "all about us" when it comes time to fashioning public policy.

Policy makers must balance all interests, and when barriers are erected to the mobility of seniors and disabled who could otherwise get to a transit stop, it screams to me that the strong are walking all over the weak, and is typical of the culture of political entitlement which has become de rigeur in this gentrifying city.

-marc

Scientifically, bus stop reduction needs to pile up savings to save the cost to the system of running another vehicle. The argument goes that if there are fewer stops, you can serve the same line more intensively with the same number of vehicles or save a vehicle for use elsewhere.

Erecting barriers to using Muni has yet to increase confidence. Why would eliminating stops, making it more difficult to reach a stop and more likely to be walking to a stop and missing a bus, make people more likely to ride the system?

Often the most obvious solutions prove unworkable in the real world. Make with the case study already if it is so trivial, simple and obvious.

-marc

Beyond that, doing something about stop spacing would indicate to the ridership that Muni does indeed care about its performance and efficacy. This would go a long way to reducing cynicism and increasing support for other necessary measures.

Come up with a proposal on eliminating stops that generates enough time to make up a bankable amount.

In riding with an eye out for this, I've noticed that the nature of eliminating stops on line timings plays markedly differently during rush hour and during the rest of the day.

Perhaps switching certain stops to "off" only during rush hour would be the best approach.

-marc

The excess stops were painfully obvious taking the 27-Bryant this afternoon. I boarded at the Washington & Van Ness stop and then the bus went about 400' to the other end of this pretty-flat block to pick up another passenger at the Washington & Polk stop. By stopping for that second passenger the bus had to then wait for the next light cycle, but if we'd both boarded at the same stop the bus would have been able to make the light without waiting.

How about if the MTA puts proposed new stop configurations into a 511.org style system where riders can compare existing and proposed?

I'd asked the MTA during the TEP to set up a similar system with the new route configuration so that folks could compare how the different systems would work.

The impacts on changes to transit networks as complicated as Muni are not very difficult quantify and test using computational methods. The MTA has not taken its proposed TEP configurations of lines and stops, to my knowledge, and run an exhaustive analysis to discern who are the winners and who are the losers:

foreach $srcline ( @lines) {
foreach $srcstop ( @{$line->{stops}} {
foreach $destline (@lines) {
foreach $deststop (@{$destline->{stops}}) {
next if ($deststop==$srcstop);
$new = find_fastest_new($srcstop,$deststop);
$old =find_fastest_old($srcstop,$deststop);
print compare($old,$new);
}
}
}
}

The 'fastest' subs are just Dijkstra's algorithm to find shortest weighted path.

More importantly, such an analysis would shine the light on unintended consequences of significant changes that are easily overlooked at the level at which the TEP was operating.

-marc

I think we're in the "fog of war" period now, operating under the "Shock Doctrine" [1], where instability and the incoherence of other democratic voices is used as an excuse by elites to quickly push through their policies that would otherwise encounter

The TEP provides some guidance for what the MTA says they're going to do in other aspects of service modification. However, the mechanics of stop removal/consolidation remain elusive and were not elaborated by the TEP in anything but glittering generalities.

-marc

[1] http://www.naomiklein.org/shock-doctrine

So what? These delays are a "symptom" of the "problem" that the BUS system is being used.

Again, you are hypothesizing, in the absence of hard data about stop behavior.

Besides, you have not yet demonstrated that the aggregate time saved by stop elimination minus the additional walk time required to get from origin and destination is a net gain, not to mention given consideration to the impact of this proposal on vulnerable riders.

Perhaps this is because you are citing a paper that deals with bus service in Portland, 75% of which is SE Portland which has a street topology that looks nothing like anywhere in San Francisco.

Please, theo, humor me and show me a concrete proposal that shows a new stop configuration and the impacts for riders at points along the line in increased walk distances and increased trunk line speeds so we can do the math.

-marc

And having too many stops makes the red light/traffic problem much worse. If you've ever taken the 47 or 49 along Van Ness, you'll understand they interact in a bad way to make the trip extremely unreliable.

I would support more bus cameras, more yellow loading zones and bus priority signals. In addition to stop consolidation.