6 ms·

NYC Bike Stats

52 points by thatcherclay 10y ago

stinos 10y ago
Is it just me or are line charts with fills harder to interpret while not adding any value over normal line charts? Instead they add ambiguity because while the surface itself doesn't seem to mean anything (?), it does have two bounding lines and depending on the graph only the upper or lower bound is the line which contains actual data - the other one is the line with the data from one of the other data sets. At least that what I make out of it but maybe I'm completely wrong? E.g. take the first graph: the red surface starts at 0 and the upper bound is the actual 'overall ridership' for Second Avenue, right? (note to author: even when the units seem obvious to you, they might not actually be obvious for everyone). So the lower bound of the green surface (Lafayette Street) has the same shape as the data for Second Avenue. Why? What does that mean? It's just the upper bound of the green surface which is the actual data for Lafayette street, no? On topic: glad to see bicycle usage is rising, but would be interesting to see if e.g. car usage is declining and how the total number of people on the road is changing.
- Ended 10y ago
  The surface is the data. E.g. the ridership for Lafayette Street is the upper line of the green area, minus the lower line of the green area. This is called an area chart and is used to show cumulative data and how each part contributes to the whole. In this case it shows the total NYC ridership increasing from ~7000 to ~18000 (units are unclear!), while e.g. the Central Park West ridership has remained approximately constant.
  - Someone 10y ago
    My first thought when seeing that chart was "How to lie with statistics". Stacking those graphs in that way only is valid if there are no double-countings in the data. New York is large, but I think/guess (I know little of street layout in NY, but looking at coordinates, Layafette street at Astor plaza and Second Avenue at 7th Street seem fairly close to each other) that is unlikely.
thesehands 10y ago
Also interesting to see the breakdown of usage of the Citi bikes. Does increased usage of Citi bikes lead more casual users who don't wear helmets?
- stepanhruda 10y ago
  I'd expect so. If I had my own bike I'd wear s helmet, but can't be bothered on Citi bike.
untilHellbanned 10y ago
Nice analysis but Excel default color schemes are unbearable. I wonder how many people their hideousness turns away before even contemplating the data.
kingosticks 10y ago
The article says they do the survey on a particular day each year whereas the data set is less specific on that fact. Surely they do the survey over multiple days...? If available, there must be a ton of citibike usage data one could use to verify the citibike stats collected by this survey and generally validate this data a bit.
- aries1980 10y ago
  Why they not rather contact with Fitbit, Garmin, Under Armour, etc. to provide data?
  - mertd 10y ago
    Perhaps because it would be biased? Cyclist population who wears fitness trackers might look different than the population of all cyclists. So now you have sampling bias to deal with.
    - kingosticks 10y ago
      Yes, this. But even so, I'd probably take that over basing my annual cycling stats on sample taking on a single day.
  - danielsamuels 10y ago
    Strava, MapMyRide and Garmin Connect would probably be the best data sources.
    - thrownblown 10y ago
      google fit probably has more
ericmo 10y ago
> Let me just start by saying that there is no good reason not to wear a helmet when riding a bike in NYC. One common argument is that drivers have more respect for cyclists without helmets, i.e. they're more afraid of hitting a cyclist without a helmet. The discussion goes back a long time, this article for instance is 9.5 years old: https://www.scientificamerican.com/article/strange-but-true-helmets-attract-cars-to-cyclists/ https://www.scientificamerican.com/article/strange-but-true-... I guess, most cyclists think it's more likely that they'll be hit by a car, not a pedestrian. It could also be that some cyclists don't have the money to buy a helmet - or other types of safety gear.
- CalRobert 10y ago
  There is also no good reason not to wear a helmet when driving a car. Head trauma kills regardless.
  - Retric 10y ago
    > wear a helmet when driving a car. Seat headrests and side impact air bags provide this function much more effectively than loose helmets. Which are not safe for use in passenger cars due to neck trauma. Further, your skull hitting the pavement is more rapid acceleration than a survivable car crash. That said, NASCAR drivers do use helmets and a five point harness instead of air bags. But, very importantly they strap the helmet to the head rest, but that only works because of the five point harness and several other safety systems missing from consumer cars. https://en.wikipedia.org/wiki/HANS_device https://en.wikipedia.org/wiki/HANS_device TLDR: The issue is neck trauma. Which NASCAR avoids, by strapping both the driver and their helmet down.
    - CalRobert 10y ago
      Well, for what it's worth I do wear a helmet when I cycle. I have a suspicion that one saved my life when I was 16 and flipped my bike, landing directly on the crown of my head. The helmet smashed around it (an odd sensation), but I managed to stay alert enough to roll out of the way of cars approaching at ~50mph. Even so, it does seem like we could save a decent number of lives by using helmets and HANS devices while motoring.
Tiktaalik 10y ago
> Don’t bike on the sidewalk. Seriously. A prevalence of biking on the sidewalk in an area is a signal to designers that cyclists feel it is too dangerous to bike on the road amongst traffic. It means that a protected bike lane needs to be built.
- pavel_lishin 10y ago
  It might also signify a tourist area - near Central Park, I usually see people who look like locals riding in the street, and people who look like tourists riding on the sidewalk. (Tourists tend to have bikes with bike-rental logos, or Citibikes; locals tend to have their own bikes, or if on citibikes, are dressed more like locals.)
  - pc86 10y ago
    Out of curiosity how does one dress like a NYC local?
    - pavel_lishin 10y ago
      Tourists tend to overdress; they see that it'll be 45 in the morning, and dress for the arctic circle, whereas locals will typically just toss on a jacket, with maybe another layer underneath. Locals also tend to wear darker colors; we joke that New York's official color is black. A tourist will typically not have a messenger bag or a briefcase with them, either - those are typically commuter items. Tourists are more likely to have a bag from a "big" brand-name store like Macy's, although not as much on a bike. And obviously, an I<3NY t-shirt with a brand new Yankees hat is a pretty dead tourist giveaway.
    - cmurf 10y ago
      October, the black parka comes out of storage and is worn, no matter the outdoor temperature. Temperature dictates the degree to which the person buttons up.
revelation 10y ago
It's hilarious that you would interrupt a blog post on analyzing a data dump for a sermon on helmets that is not remotely supported by the empirical data. Cyclists die in traffic because they are hit by multi-ton motorized vehicles, not any of the things enumerated here.
bluenoteslur 10y ago
Interesting analysis. After ten years, I wonder whether the authors' choice of sampling intersections should be updated as new bike lanes make certain routes more preferable. For example: There's a heavily-used bike lane on 9th Street which is on the official NYC bike map; 7th street has no marked lane and doesn't get as many riders in my experience.

52 points by thatcherclay 10y ago

Is it just me or are line charts with fills harder to interpret while not adding any value over normal line charts? Instead they add ambiguity because while the surface itself doesn't seem to mean anything (?), it does have two bounding lines and depending on the graph only the upper or lower bound is the line which contains actual data - the other one is the line with the data from one of the other data sets. At least that what I make out of it but maybe I'm completely wrong? E.g. take the first graph: the red surface starts at 0 and the upper bound is the actual 'overall ridership' for Second Avenue, right? (note to author: even when the units seem obvious to you, they might not actually be obvious for everyone). So the lower bound of the green surface (Lafayette Street) has the same shape as the data for Second Avenue. Why? What does that mean? It's just the upper bound of the green surface which is the actual data for Lafayette street, no? On topic: glad to see bicycle usage is rising, but would be interesting to see if e.g. car usage is declining and how the total number of people on the road is changing.

Ended 10y ago
The surface is the data. E.g. the ridership for Lafayette Street is the upper line of the green area, minus the lower line of the green area. This is called an area chart and is used to show cumulative data and how each part contributes to the whole. In this case it shows the total NYC ridership increasing from ~7000 to ~18000 (units are unclear!), while e.g. the Central Park West ridership has remained approximately constant.
- Someone 10y ago
  My first thought when seeing that chart was "How to lie with statistics". Stacking those graphs in that way only is valid if there are no double-countings in the data. New York is large, but I think/guess (I know little of street layout in NY, but looking at coordinates, Layafette street at Astor plaza and Second Avenue at 7th Street seem fairly close to each other) that is unlikely.

thesehands 10y ago

Also interesting to see the breakdown of usage of the Citi bikes. Does increased usage of Citi bikes lead more casual users who don't wear helmets?

stepanhruda 10y ago
I'd expect so. If I had my own bike I'd wear s helmet, but can't be bothered on Citi bike.

untilHellbanned 10y ago

Nice analysis but Excel default color schemes are unbearable. I wonder how many people their hideousness turns away before even contemplating the data.

kingosticks 10y ago

The article says they do the survey on a particular day each year whereas the data set is less specific on that fact. Surely they do the survey over multiple days...? If available, there must be a ton of citibike usage data one could use to verify the citibike stats collected by this survey and generally validate this data a bit.

aries1980 10y ago
Why they not rather contact with Fitbit, Garmin, Under Armour, etc. to provide data?
- mertd 10y ago
  Perhaps because it would be biased? Cyclist population who wears fitness trackers might look different than the population of all cyclists. So now you have sampling bias to deal with.
  - kingosticks 10y ago
    Yes, this. But even so, I'd probably take that over basing my annual cycling stats on sample taking on a single day.
- danielsamuels 10y ago
  Strava, MapMyRide and Garmin Connect would probably be the best data sources.
  - thrownblown 10y ago
    google fit probably has more

ericmo 10y ago

> Let me just start by saying that there is no good reason not to wear a helmet when riding a bike in NYC. One common argument is that drivers have more respect for cyclists without helmets, i.e. they're more afraid of hitting a cyclist without a helmet. The discussion goes back a long time, this article for instance is 9.5 years old: https://www.scientificamerican.com/article/strange-but-true-helmets-attract-cars-to-cyclists/ https://www.scientificamerican.com/article/strange-but-true-... I guess, most cyclists think it's more likely that they'll be hit by a car, not a pedestrian. It could also be that some cyclists don't have the money to buy a helmet - or other types of safety gear.

CalRobert 10y ago
There is also no good reason not to wear a helmet when driving a car. Head trauma kills regardless.
- Retric 10y ago
  > wear a helmet when driving a car. Seat headrests and side impact air bags provide this function much more effectively than loose helmets. Which are not safe for use in passenger cars due to neck trauma. Further, your skull hitting the pavement is more rapid acceleration than a survivable car crash. That said, NASCAR drivers do use helmets and a five point harness instead of air bags. But, very importantly they strap the helmet to the head rest, but that only works because of the five point harness and several other safety systems missing from consumer cars. https://en.wikipedia.org/wiki/HANS_device https://en.wikipedia.org/wiki/HANS_device TLDR: The issue is neck trauma. Which NASCAR avoids, by strapping both the driver and their helmet down.
  - CalRobert 10y ago
    Well, for what it's worth I do wear a helmet when I cycle. I have a suspicion that one saved my life when I was 16 and flipped my bike, landing directly on the crown of my head. The helmet smashed around it (an odd sensation), but I managed to stay alert enough to roll out of the way of cars approaching at ~50mph. Even so, it does seem like we could save a decent number of lives by using helmets and HANS devices while motoring.

Tiktaalik 10y ago

> Don’t bike on the sidewalk. Seriously. A prevalence of biking on the sidewalk in an area is a signal to designers that cyclists feel it is too dangerous to bike on the road amongst traffic. It means that a protected bike lane needs to be built.

pavel_lishin 10y ago
It might also signify a tourist area - near Central Park, I usually see people who look like locals riding in the street, and people who look like tourists riding on the sidewalk. (Tourists tend to have bikes with bike-rental logos, or Citibikes; locals tend to have their own bikes, or if on citibikes, are dressed more like locals.)
- pc86 10y ago
  Out of curiosity how does one dress like a NYC local?
  - pavel_lishin 10y ago
    Tourists tend to overdress; they see that it'll be 45 in the morning, and dress for the arctic circle, whereas locals will typically just toss on a jacket, with maybe another layer underneath. Locals also tend to wear darker colors; we joke that New York's official color is black. A tourist will typically not have a messenger bag or a briefcase with them, either - those are typically commuter items. Tourists are more likely to have a bag from a "big" brand-name store like Macy's, although not as much on a bike. And obviously, an I<3NY t-shirt with a brand new Yankees hat is a pretty dead tourist giveaway.
  - cmurf 10y ago
    October, the black parka comes out of storage and is worn, no matter the outdoor temperature. Temperature dictates the degree to which the person buttons up.

revelation 10y ago

It's hilarious that you would interrupt a blog post on analyzing a data dump for a sermon on helmets that is not remotely supported by the empirical data. Cyclists die in traffic because they are hit by multi-ton motorized vehicles, not any of the things enumerated here.

bluenoteslur 10y ago

Interesting analysis. After ten years, I wonder whether the authors' choice of sampling intersections should be updated as new bike lanes make certain routes more preferable. For example: There's a heavily-used bike lane on 9th Street which is on the official NYC bike map; 7th street has no marked lane and doesn't get as many riders in my experience.