Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
This was mentioned many times already. SW Ballpark - Navy Yard isn't even the most egregious example of it especially since SW Ballpark - Navy Yard is nowhere near being the neighborhood unit, as walk score has them, that's added the most people since they started updating. I also doubt the area as they've outlined it, and the way they outline neighborhoods is another criticism, has more than 20,000 people living in it. If the entirety of Ward 6 added about 30K people from the 2010 to 2020 census, then does it really make sense that this tiny sliver they outlined takes up the majority of that addition with the over 16K people you're stating? It seems unlikely especially as it's missing all that prime development along M street near the water.
This is one of multiple issues with walk score I outlined, though I think the unclear and slow update among US cities is probably the most reasonable and now longstanding criticism.
Well, I can tell you there are probably about 7,000 people living here right now:
You're right based on their boundary that there are less than 20,000 people in that footprint. I wonder why they said SW? SW DC doesn't start till the other side of South Capitol St. That's why I was confused.
As you're right based on their boundary. I wonder why they said SW?
Yea, that's one of the things about some of the names. It gets much worse with even smaller cities like with Tacoma mentioned earlier.
The census tract area might have have changed since they last sourced the data. The part north of M street and on the other side of the freeway as they had part of the neighborhood outlined still roughly follow the bounds of existing census tracts today, but not the eastern part of it though whose closest census tract now encompasses a lot of area that wasn't in the walk score neighborhood outline. Even including that extra, it wouldn't come very close to 20K people as they've outlined it, so that's why I was saying that the walk score neighborhood definitions can introduce some unexpected oddities. Unfortunately, I think development on the web site has been more or less abandoned for the last few years and any updates are probably just from anything that takes from third party services that have an API that still exists and works as intended before development stopped.
Last edited by OyCrumbler; 11-17-2023 at 01:33 PM..
Because cities vary dramatically in how much physical area they cover and what portions of the metropolitan area they cover, walk score for city vs city comparisons end up being nigh useless. This is similar to when people say Jacksonville is bigger or more populous a city than DC and Boston. Sure, that's true, but it's also kind of not.
Using a city walk score strictly on municipal legal boundaries is maybe ironic for something that's supposedly ranking attributes that are supposed to be quite fine-grained and on a human scale.
I think other things people have noticed about walk score is that:
- the methodology in terms of sources used seem to effectively short change Canadian cities perhaps due to different sources pulled for US and Canada; it seems to be even worse for Quebec
- some of the data pulled is unfortunately outdated such as neighborhood population counts being at least least several years old and not updated from the 2020 census (2021 for Canada), so it's unclear what the update cycle for the various sources are but it is clear that at least some of them can go quite a while before updates
- it used to go by sheer proximity as the crow flies though I think in somewhat later times it did some route-finding
- it doesn't take into account some elements of urban design though arguably did better when it seemed to have implemented route-finding which would have innately taken into account some of that
Overall, I like walkscore and the methodology and idea, but I think the lack of updating especially after being acquired by Redfin is disappointing. I'd like to do an open source project version of it and thus with parameters that one can tweak.
I also tried to do what I think is the first step in a better way of doing a ranking of cities in terms of walkscore to move away from the obvious issues of doing it by municipal borders. This was to try to go by contiguous clumps of high walkscore areas. The most straightforward attempt was to use a cut off of 90 for neighborhoods that walk score used and sum the population of those neighborhoods as cited in this topic and with the last one done in this post though I stopped updating it given that walk score did not seem to be moving towards updating their numbers anymore:
Agree that the averaging the walkscore value across entire municipal boundaries is completely useless. I think measuring the population / geographic area above a (relatively high) walkscore threshold is a pretty good proxy for how walkable a city is perceived to be. I think only 90+ is truly walkable. 60-90 is useful walkability but not enough to rely on. Once below 60 or so I don't think it matters much. At that point you are doing errands in a car.
Yea, that's one of the things about some of the names. It gets much worse with even smaller cities like with Tacoma mentioned earlier.
The census tract area might have have changed since they last sourced the data. The part north of M street and on the other side of the freeway as they had part of the neighborhood outlined still roughly follow the bounds of existing census tracts today, but not the eastern part of it though whose closest census tract now encompasses a lot of area that wasn't in the walk score neighborhood outline. Even including that extra, it wouldn't come very close to 20K people as they've outlined it, so that's why I was saying that the walk score neighborhood definitions can introduce some unexpected oddities. Unfortunately, I think development on the web site has been more or less abandoned for the last few years and any updates are probably just from anything that takes from third party services that have an API that still exists and works as intended before development stopped.
Yeah, I see that. I will say that in a few years, this area will probably have over 20,000 people based on the current pipeline, but who knows how Walk Score will divide the neighborhood. It definitely has between 10,000-15,000 in 2023.
Aside from sidewalks, most aesthetic components should be excluded imo. Those tend to measure not whether a walk can be done but how pleasant it might be perceived. And that not only starts becoming subjective, it becomes difficult to measure and quantify. Which is more detrimental for scoring, curb cuts or street widths? Sidewalks or block lengths? We all know a pleasant walk when we see it but quantifying pleasantness is hard and subjective. I get the issues with the scoring, but I tend to take such information fwiw.
Street width and frequency of crossings/connections aren't aesthetic. They affect real and perceived safety, waiting time, etc., and therefore people's willingness to walk.
Agree that the averaging the walkscore value across entire municipal boundaries is completely useless. I think measuring the population / geographic area above a (relatively high) walkscore threshold is a pretty good proxy for how walkable a city is perceived to be. I think only 90+ is truly walkable. 60-90 is useful walkability but not enough to rely on. Once below 60 or so I don't think it matters much. At that point you are doing errands in a car.
I agree with this alongside knowing something like the upper 80s especially with how large some of the neighborhoods are can be pretty fuzzy boundaries on where the cut off should be. I also think that for contiguous blobs that if it's a really small patch of something that's a bit below 90 in between larger patches of places that are 90+ and that below 90 strip between them is quite permeable, then there should be some way to factor that that in. It's an interesting concept to try to play around with, but too much work for me at the time to do too much with.
Quote:
Originally Posted by MDAllstar
Yeah, I see that. I will say that in a few years, this area will probably have over 20,000 people based on the current pipeline, but who knows how Walk Score will divide the neighborhood. It definitely has between 10,000-15,000 in 2023.
Yea, there are still quite a few large surface lots to develop so that doesn't seem out of the question. I don't know if it actually hits 10,000-15,000, but if it misses, it just barely does so. Note that tract 72.01 you're noting isn't in their borders. Instead they have tract 71 though tract 71 encompasses a lot more that's not within walk score's borders.
Quote:
Originally Posted by mhays25
Street width and frequency of crossings/connections aren't aesthetic. They affect real and perceived safety, waiting time, etc., and therefore people's willingness to walk.
Yea, and it doesn't factor for that though at least the shift into walking distance routing instead of as the crow flies distance takes care of at least some of it and is a vast improvement. I also think there's a rough correlation in having a large mix and density of goods and services and walkability, so even the previous as the crow flies proximity method wasn't egregiously off most of the time.
I agree with this alongside knowing something like the upper 80s especially with how large some of the neighborhoods are can be pretty fuzzy boundaries on where the cut off should be. I also think that for contiguous blobs that if it's a really small patch of something that's a bit below 90 in between larger patches of places that are 90+ and that below 90 strip between them is quite permeable, then there should be some way to factor that that in. It's an interesting concept to try to play around with, but too much work for me at the time to do too much with.
Yea, there are still quite a few large surface lots to develop so that doesn't seem out of the question. I don't know if it actually hits 10,000-15,000, but if it misses, it just barely does so. Note that tract 72.01 you're noting isn't in their borders. Instead they have tract 71 thought tract 71 encompasses a lot more that's not within their borders.
Yea, and it doesn't factor for that though at least the shift into walking distance routing instead of as the crow flies distance takes care of at least some of it and is a vast improvement. I also think there's a rough correlation in having a large mix and density of goods and services and walkability, so even the previous as the crow flies proximity method wasn't egregiously off most of the time.
Well for starters, there are more than 10,000 units within the tracts I posted. I think there are a little over 12,000 units. Please note this is also a little behind. It hasn't been updated since 2020. There are about 1,000 units that have either delivered or are under construction that are listed as pipeline.
Street width and frequency of crossings/connections aren't aesthetic. They affect real and perceived safety, waiting time, etc., and therefore people's willingness to walk.
Debatable and even if true, it's hard to quantify how they impact walk decisions. I won't disagree that I find those same qualities attractive but we've also discussed many times how people complain that street widths in LA are too wide even though Google Maps shows them to be more or less similar to street widths in Manhattan.
Personally I find curb cuts and parking lots more of a hinderance to walking than street crossings and block lengths, as long as those two things aren't egregious. Weather and crime are probably bigger factors when influencing walk decisions but I don't think things like those are easily quantifiable either.
What can I walk to and how far is it? Really basic and may not tell you what you want to know but everything else is subjective and open to preference.
Well for starters, there are more than 10,000 units within the tracts I posted.
That wasn't the point being made, but okay. The three census tracts you posted with that do not add up to 10K population nor to 10K housing units as far as I can tell.
That math doesn't add up to 10K on either metric. That's with the links you posted. I certainly think it's possible they're over 10K now because the data those links are drawing from are from a 2021 ACS estimate. It's just that that's not what the links currently say.
Now, again, not the point. I think I've probably voiced in the past just how very little I care about DC development news minutiae. Not that interesting to me in the past, still not interesting to me now, and the most important part of this is that it is not the point. The point was that walk score has outdated data and that the neighborhood boundaries they used aren't necessarily going to be sensible given the name. As noted before, tract 72.01 that you're linking to and does make sense for the name of the neighborhood they used is *NOT* included in their neighborhood boundaries. Instead, their neighborhood boundary includes a small part of what's now tract 71 which was probably different the last time they pulled their data.
That wasn't the point being made, but okay. The three census tracts you posted with that do not add up to 10K population nor to 10K housing units as far as I can tell.
That math doesn't add up to 10K on either metric. That's with the links you posted. I certainly think it's possible they're over 10K now because the data those links are drawing from are from a 2021 ACS estimate. It's just that that's not what the links currently say.
Now, again, not the point. I think I've probably voiced in the past just how very little I care about DC development news minutiae. Not that interesting to me in the past, still not interesting to me now, and the most important part of this is that it is not the point. The point was that walk score has outdated data and that the neighborhood boundaries they used aren't necessarily going to be sensible given the name. As noted before, tract 72.01 that you're linking to and does make sense for the name of the neighborhood they used is *NOT* included in their neighborhood boundaries. Instead, their neighborhood boundary includes a small part of what's now tract 71 which was probably different the last time they pulled their data.
The Census Reporter website is way behind for DC. DC builds way too many buildings at such a high rate for the Census Reporter website to ever be up to date on construction and population for a city like DC. You can count the units for yourself here:
Every single building listed as under construction has been finished for years. Now, even a few buildings listed as pipeline are finishing up construction or have broken ground. The author of this site stopped posting during COVID.
Here is one that was listed as pipeline on that website: Under Construction
A second building listed as pipeline on that website is now under construction on this lot: Under Construction
Again, there are over 12,000 units in that footprint right now.
Looking into the future, Phase II of the Yards is just starting with an additional 3,400 units (but the residential unit count will probably double or triple because most of the 2,000,000 sq. feet of office is expected to go residential.).
Last edited by MDAllstar; 11-17-2023 at 02:59 PM..
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.
Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
