Swivel, the new Minor Ventures startup that we wrote about yesterday, is now live. Go compare the increase in U.S. wine consumption to the the drop in crime, or see how extreme temperature fluctuations (in Fresno) equal big profits for electricity utilities.
I already see a lot of fairly useless data included on the site, which makes the good stuff harder to find. As the active community grows the better data will hopefully rise to the top via the rating and feedback system. This is one of the freshest ideas we’ve seen in months – we’ll be tracking them closely.
http://digg.com...search?s=swivel
,
It is certainly NOT hard to see how important a TechCruch Write-up can be for a start Up, and that is not even counting the impressive 100 replies to the original topic yesterday……..
Compare the effect on Digg –
the TechCruch Review (with the new title) was submitted just minutes after a direct Digg to the Swivel Homepage was first submitted.
Both titles and descriptions were equally interesting – but look at the differences in the results Digg results.
You also realizr that people forget things easily.
wait for a week.
Just use search to weed through the datasets you don’t want
Swivel will enable us to do some great research.
Wow. This could be pretty sweet. Not that this will appeal to all personality types, but for me I love exploring how different variables are interconnected. We’ll see how submissions improve in quality — and what mysteries this site can reveal. I love the concept.
Swivel has great potential although their success will hinge on separating the quality data from the bogus.
I am a fan!
I just loved the idea that Swivel has …. want them to succeed… but some doubts though of how ranking and feedback will be able to weed out the useless stuff from the useful data.
I’m about as underwhelmed as I expected. this is *not* a research tool, it’s a clever gimmick. I really can’t identify and way to make money off this. plotting data set X vs data set Y is not “research,” it’s microsoft excel 101. really, this hype is starting to get worse than dot-com days…
I think this service can be very useful to measuere real experience in comperrison to assumption and subjective oppinion. I hope Swivel is going to reach a critical mass.
Thank you for sharing this story with me !
It’s cool, wish it all the best !
http://www.ezecho.com
I think this app will be a killer!
Just imagine how much power will this give to all the community groups that concern themselve with how the public money is spent. Assuming the data is out there, they can quickly do a regression of relevant variables and see whether a city government measures in the past 2 years helped to reduce crime in XYZ postcode…
Congrats Swivel!!!
And swivel is down….
Can I advise everyone using swivel to read this page first about the unreliability of statistics.
http://www.bbc....a/h2g2/A1091350
and maybe also this one specifically about how to decide if research is reliable.
http://www.bbc....na/h2g2/A852761
Still it’s very interesting and I will keep an eye on them!
The TechCrunchies must be hitting the Swivel server hard… I can’t even get through!
I have to admit, they did a good job with the site. They built the site using Ruby on Rails and I like how they managed to get the graphs generation so smooth and fast.
I always needed a site like swivel.com
Now the site is up and running.
I need the data and I need to be able to verify the data from the original sources. Swivel.com allows me to do that.
For me Swivel.com is not “a YouTube for data”, but a search engine for data which displays diagrams and graphics. That´s what I need.
Everything else is neat and fun and I like all them as well.
Copywright issues with Swivel.
I wonder how they will handle this as they have a much greater problem than the likes of YouTube.
At least there are millions of people creating their own home videos everyday for uploading to YouTube. Even with that there was / is still a huge amount of copyrighted data on YouTube.
However, there are not millions of people out producing their own datasets from their own proprietary knowledge. I would hazard to make a statement that there are far less people interested in creating datasets and posting them online than the numbers who would do the same with videos. Datasets generally tend to be harder to put together than poor quality home video.
People posting to Swivel are usually taking datasets from already published sources. It is a fair comment that not many people are likely to be using this published data because it is in odd formats etc., but the bottom line is that the poster to Swivel may not be adhering to intellectual property law.
I will be interested to see how this develops.
Swivel should (have) develop(ed) relationships with these sources of data and figured a way to ensure that this issue does not arise.
Yeah, Swivel really missed the big picture on this. Instead of really solving a problem, they got caught up in all the Web 2.0 hype. But that’s okay, I’ll just build the more useful version of this site.
Why use swivel when u can use stats canada or something… http://www.statscan.ca
Hmm.. time to invest in some infrastructure..
500 errors on homepage and other graphs.. ouch.
Seems interesting, although I don’t think LonelySQLServer15 will generate all that much attention (plus it’s staged anyway)
TERMS OF USE – IN CASE ANYONE MISSED IT.
User Submissions.
The Service provides you with the ability to upload, submit, disclose, distribute or otherwise post (hereafter, “posting”) content (including, without limitation, information, data, and images) to the Service (”User Submissions”). By posting User Submissions on or at the Site or otherwise through the Service, you:
* grant to Swivel, its affiliates and their assignees a perpetual, irrevocable, non-exclusive, sublicensable, royalty-free right to use, reproduce, display, perform, adapt, modify, distribute, make derivative works of and otherwise exploit such User Submissions in any form and for any purpose, including without limitation, any concepts, ideas or know-how embodied therein;
* represent and warrant that you own or otherwise control all rights to such User Submissions and that disclosure and use of such User Submissions by Swivel (including without limitation, publishing content on or at the Site) will not infringe or violate the rights of any third party; and
* acknowledge that such User Submissions may not be treated confidentially.
You shall not provide Swivel with any confidential or proprietary information that you desire or are required to keep secret. You are solely responsible for the User Submissions you post on or through the Service. Swivel does not endorse and has no control over any User Submission.
Swivel has no obligation to monitor the Site, Service, Content, or User Submissions. Swivel may remove any User Submission at any time for any reason (including, but not limited to, upon receipt of claims or allegations from third parties or authorities relating to such User Submission), or for no reason at all.
LonelySQLServer15, that is hilarious.
Brian Mulloy
CEO & Cofounder
http://www.swivel.com
Swivel had a bit of indigestion this morning, hence the 500 errors, but we’re feeling better now. The hamsters are running hard making all the graphs. Thanks for bearing with us.
Dmitry Dimov
Product Chief & Cofounder
http://www.swivel.com
Howdy Alfred, thanks for noticing. We do use Ruby on Rails. We love it. It lets us be productive and have fun at the same time. We idolize DHH.
We explored a few different methods for making graph generation fast. Glad you got that impression — it means we’re at least on the right track.
Huned (member of engineering staff at swivel)
Again grats guys…..site looks good, very user friendly.
LMAO from the amount of game related data already on the site I can only assume you have a few gamers in your office
Swiveldudes/dudettes?–this is simply awesome. you’ve just introduced a new era of social science for the masses.
Hopefully your efforts will be used for the benefit of social change and education…
Another sweet tool.2.0.com I don’t have any use for. Good luck.
Before we usher in a new era of social science research, let’s start here with some questions for this graph, as an example:
http://www.swiv...hs/show/1001967
1)In 1980, there were 1,344,520 violent crimes in the U.S. according to the data as given below. What does a value of about -5% (instead of 1,344,520) on the y axis mean? What is being plotted on the Y axis? It’s not obvious to me.
http://bjsdata....tatebyState.cfm
2)What is the logic behind that oddly named Y axis on almost all graphs. Relative values? Percentages? Huh? Why not have an easily interpreted y axis at the beginning and end of the graph (one y axis per variable)? There is no explanation of the Y axis anywhere for any graph.
3)Using the NUMBER of violent crimes over time as your variable isn’t a good idea. As population grows over time, of course there will be more violent crimes — just like the NUMBER of violent crimes in a tiny state like Rhode Island will be a lot less than the NUMBER of crimes in a big state like New York. This is the issue of using levels versus rates in data. You should use a crime RATE to “normalize” things.
4)Regarding wine, in 1980 the data says U.S. per capita wine consumption was “2.1″ But 2.1 of what? Glasses, bottle, liters, what? I can’t find the exact data source (data sources are poorly cited in general on swivel) from the link given. And what does a y value of a little over 0 PERCENT for 1980 wine consumption mean? How did we get from 2.1 whatever to whatever is being plotted on the Y axis in the picture.
http://www.swiv...hs/show/1001967
…so I don’t know. Maybe swivel will be a good thing. Maybe it’s just introducing the the numerically challenged to excel graphs and providing links to data sources all over the internet for the unfamiliar data gatherers. It remains to be seen.
I think that the immortal Stevie Wonder said it best as far as armchair statistical analysis goes:
“If you believe in things that you don’t understand…then you suffer.”
(From the song Superstition)
Correction:
in my #1, that should be y value of about -12%, not -5%.
Swivel is very cool and I’ll definitely give it a try. 1010data.com has been around since 2000 and can handle many billions of rows and thousands of fields. In other words, if you need real industrial strength data analysis – all done on-line, give it a try. I know the NYSE uses 1010data to publish its daily quotes and trades.
david8,
Thank you for the criticism — every point you made is valid and we have a bunch of work to do there.
The mysterious Y scale is either relative to average, or relative to range of the data series. This is shown in edit view but it is not apparent anywhere else so we’ll have to fix that.
We also will be adding the ability to normalize values, for things like adjusting for inflation. One thing to keep in mind is that we don’t control what our users put into Swivel and can’t dictate that they normalize their data where makes sense, but we will certainly need to make it easy, so this is a good one.
Finally, we’re working on better support for units, so it will be apparent that it was 2.1 gallons of wine.
Sincere thank you again for the good feedback.
Dmitry Dimov
Product Chief & Cofounder
http://www.swivel.com
david8 (and everyone else):
Statistics are like a very small bikini — they reveal much but hide that which is most interesting/valuable, making you only want to see more!
This is fresh. Great job.
Best wishes for your new website, Dmitry and Company. If it gets people thinking about relationships between things and how numbers/stats can help folks move beyond ad hoc discussions of whatever into more substantive analysis, that’s a good thing.
I’ll check back as you fine tune things.
Take Care.
This is quite possibly the most retarded idea for a startup I’ve seen on TechCrunch yet (yes, even worse than the 3D virtual mall).
Just out of curiousity, are all the members of the swivel.com team just wannabe-web2.0-billionaire code monkeys? Or do you actually have a data analyst/scientist on the team?
Hint: Correlation does NOT imply causation. What a waste of bandwidth.
Hey Pablo (#34),
I dig the brutal honesty of your comment. And I refer to others as ‘wannabe-web2.0-billionaire code monkeys’ about once a week. You would fit right in at Swivel.
Dmitry and I both studied physics (although Dmitry is at least 10x smarter) in college. At University of Michigan I did research on complex systems and fell in love with the way patterns would emerge on a scatter plot or the way a simple delay-differential equation could generate crazy, yet repeatable time series.
If you look around, there is plenty of _software_ for data (Excel, Mathematica, etc.) but there isn’t a _Web site_ for data. We’re building Swivel so curious people have a Web site where they can explore data.
BTW, if you have a science or math background, we are hiring sharp-minded people at Swivel.
Brian Mulloy
CEO & Cofounder
http://www.swivel.com
Thanks, david8, for the comments. You can’t argue with a guy who quotes Stevie.
Congratulations guys; it will be exciting to see how this grows in the future.
Thanks, Dave Werner.
i was looking forward to try it