 Thanks very much for having me here today. I want to talk about revelation in data visualization today. Visual storytelling is my trade, and I've found just in the past couple of years that what I've been really interested in is trying to use the storytelling to reveal information to users, because I'm very focused on the people who are using the journalism that I'm producing. And I try to really start with building blocks. And one of our favorite building blocks looks like this. Leigh and I did this for you. You know, I try to think sometimes about information as pieces of Legos, right? Just how you take one and you build it. And one thing that's interesting about Legos is that it's actually, to me, not very much fun to go to a Lego museum and see what somebody else built. I want to build it myself, or I want to see the process of it being built. And so this is sort of one of my Lego building blocks. This came from a project that we published this fall on oil exploration in North Dakota, where I was actually fortunate to live for just a couple of years when I was a teenager. And so we're looking at the landscape here, and that's a very plain satellite image. And we've coded, we've color coded those yellows, which are all of the oil wells that have been built since 2006. And I wanted to be able to reveal those oil wells. So we're looking at the surface of the landscape. But I also want to be able to reveal the infrastructure that is underneath the surface of the landscape. And those are the fracking drilling lines that are actually underneath the surface. So we're kind of making the invisible, you know, what you can't see when you're standing right on the surface of the earth. I want to be able to make that also visible. One of my colleagues, Greg Arash, who I really enjoy working with said, you know, this is so fascinating, but we're really only being able to see it in two dimensions. You know, these lines are both, you know, you know, you know, they look like this. And so he said, I wonder if there would be a possibility of doing something a little bit different. So we're constantly thinking about building where we want to go, where we want to take our readers. And Greg Arash came up with this. So this is a three-dimensional of those fracking lines that I showed you flat, and it's to scale, so you can get a really good sense of how big these things are and how much they take over the whole landscape. And I feel like the tools that we have, a lot of the tools that we've seen yesterday and just this morning, you know, I work with my colleagues to use those tools as many as we possibly can to be able to reveal this kind of information to readers. You'll always have crazy ideas. It's part of the joy of being a journalist is being able to come up with crazy ideas, sometimes hypotheses, and it's really fun to be able to test them. So another example I have here looks at the Harvard Business School, a friend of mine, Jody Cantor, did a big piece on the Harvard Business School and how they were having some gender issues. A lot of the faculty members were male, and a lot of the female students felt that the classroom tensions had some real gender bias with it. And so with Jody, I decided to see what kind of data that we could get out of the Harvard Business School to be able to shed some light to reveal something about this. Harvard was not super excited about sharing a lot of the data that they have, so I thought, all right, well, what's available, what's publicly available that I could use to begin to sort of test this hypothesis that the deans were telling me that in fact there is a real gender disparity in the faculty at Harvard, but the reason can be blamed on a single reason, and that single reason is the tenure process. It's very difficult to get tenure. It's a tenure process, and so there's this immense lag in being able to recruit good people, and then 10 years later, that sort of once the tenure pipeline has been sort of filled, then you begin to sort of solve this gender disparity problem. And I said, great, that's a testable hypothesis. Let's test that baby. So I got all the CVs of all the professors who currently teach here and what you're looking at right now. It is how long those tenured male professors have been teaching. So you've got one guy who's been teaching there for 51 years. He's tenured, he's doing a great job. And then you have the one super lucky guy who earned tenure when he'd only taught for six years, which is really kind of unusual, it usually takes about nine, 10, 11 years. I thought, all right, so then now we'll look at women. Okay, like everybody knows they don't have that many women tenured professors. We've got that here. But if we test the hypothesis that in fact the tenure pipeline will be eradicating this gap, then let's see what happens. So that is the tenure pipeline. So that is male professors without tenure right now. And I thought, okay, so here's the reveal moment, right? If the deans are telling me that the pipeline, the pipeline is set up so that in about 10 years we'll have at least a lot closer to the gender equality in professorships there, then we should be all set. But in fact, the tenure pipeline for women is significantly smaller. So even in 10 years, they're still gonna be looking at some gender disparity there. And that was something that I was able just to pull out of the CVs. I worked closely with Jody. I did not work super closely with the deans at the Harvard Business School. But we had some good conversations about this afterwards. Digging around in the data, something I really like to do. I think it's something that probably we all really like to do. And we saw some great tools yesterday that help us dig around in the data. Sometimes a clean hypothesis doesn't just hand itself to you. And sometimes you get an interesting data set and you know there's good stuff in there. Sometimes it's through reporting or maybe a client hands it to you and you say, I wanna find the best thing in here. And Jeff talked a lot yesterday in our first keynote about using some of their tools to be able to sort of really look at those trends, look at outlier, see what's rich and interesting inside of it. This is a piece that looks at auto fatalities. And I had a data set that I've actually used several times. It's a government data set called FARs. And it's on auto fatalities. Anytime someone dies in an auto accident in the United States the government takes about 100 pieces of data about that accident. Where it was, who was driving, what their blood alcohol level was, who was sitting where, all kinds of things like that. And it's a very big, very interesting data set. And I wanted to look at it when the National Transportation Safety Board came out and said, we'd like to change the legal blood alcohol level. So right now it's at 0.08, that's legal. So it's binary, if you're above that, it's not okay. If you're below that, it's fine. They wanted to change it to 0.05. And I thought, all right, this might be an interesting moment in time to do a little bit of data acceleration to say, if you change that law, now the states have to do it, the National Government can only make recommendations. But if states start to change that law, who is affected the most? And I didn't know, I didn't have a hypothesis. So I started sketching and playing around and looking at the data and diving into it. And Aleister, Dan and I worked on this. And once we sort of started to work with the data, we realized that there was a real correlation to age. And so what this shows is along the horizontal, you've got age, you've got 16, all the way up, all the way up to age 60. And then you've got on your vertical is your blood alcohol level. So the first group is the very low blood alcohol level. And then where you get into all of the dark reds, that's where the blood alcohol level is really high. And then that middle part, that middle part is the levels and proposals. So right now that middle part is paired with the top, but under the new proposed regulations, it would then move down to the bottom. And it's very clear to see right now that younger people, that younger people are the ones where driving an alcohol is a really incredibly dangerous mix. And so really in the under 26 crowd is where, if this legislation were to go into effect in states, that's the population that would be most affected. And it was an interesting project to me because I didn't come to it with a very clear thought. But I do feel like I was able to reveal something interesting in this big data set that was very newsy and on point and really were able to show people the affected population. And I got some good feedback from readers that said, you know, this was a really good conversation starter with my teenager. Choosing form, form for the win. I think getting to play with form is something that is one of the great joys of being a visual journalist and being in this field. Finding the right form is like singing, right? I mean, it's just finding the right form to match your data is kind of the most magnificent thing in the world. And it's also sometimes the hardest. This is a piece that now you've seen three times. So I'm not really gonna talk about it. But what I wanna talk about is actually the sort of father of this piece, the piece that actually came before. And incidentally, the data for this piece also came from that far as data set that I just talked about. So that really related, very interesting. But this actual connected scatter plot was the first one that I did for the times a couple of years before I did the one that you've already seen three times now. And this one was for a business column that I had called Metrics. And I really wanted to look at the relationship between driving the price of gas. And that's kind of interesting right now because we're in, you know, we've got some pretty low gas prices right now. But this was actually published in early 2010. And like the other connected scatter plot that you've seen already several times, it actually has the same horizontal axis that is miles driven per capita every year. Because wow, Americans love to drive. So in the 1970s, an average American was driving 5,000 miles every year. And now, it's twice that. Now we all drive, each one of us drives more than 10,000 miles every year. And so that makes, that means that the, that metric means it's really kind of clean and good for a connected scatter plot. But my vertical is the price of gas. And so what I wanted to look at is this sort of historical relationship between those two. So in the, in the 50s and 60s, cheap gas, long commutes, people were fleeing urban areas, but they were still working in urban areas. And so commutes got really long, but it was all fine because gas was totally cheap, no problem. And then Kaboom, Big Spike, 70s and 80s, energy crisis, Sarah Boyle embargo, you've got a lot of stuff happening there. And visually you can see that. Your eye goes right to that peak and you say, ooh, what went on? And there was an energy crisis. And then it fell, it fell in just a couple of years and then we started to have some record lows again. People started to drive a ton more. You'll notice during that peak in the energy crisis, a lot of people stopped driving quite as much. So you just didn't get that incredible pace of increasing your driving every year. But then in the 80s and 90s, you did again and then we started to see another big climb. And for the first time, you actually saw Americans driving less in 2009. Then they had before. What happened in 2009? Economic recession, a big one. A lot of people lost their jobs. People who don't have jobs don't drive as much. There's less merchandise on the road. There's less trucks out there. And so that number actually swung backwards. And I thought that was a really interesting moment in time to be able to say, hey, look at this. For the first time, we're actually swinging backwards in this metric. So choosing that form, I thought, was a little bit playful, a little bit interesting. I've actually been wanting to update it, but I haven't had time to update it. But I thought it would kind of be an interesting thing here. Along the bottom, you can see that I've got some small multiples. So it's the exact same chart shape. I'm just highlighting a few different little pieces and kind of giving you some analysis on the bottom. What's in there is basically what I just told you, so you don't need to read it. But I remember when we published this, somebody that I respect a lot in the community sent me an email saying, wow, having the only color on a whole color page, this was a printed version, the only color on the whole printed page were in those tiny, little, reddish-brown arrows. And that email was from Nigel. He said, that show is an incredible constraint. Hannah, I applaud that. So layering information, especially now that we are looking at so much design on mobile, layering information is really important. And I really feel that it can set the pathway for discovery because I want my reader to be the one who is discovering this information as well. And the piece that I just showed you, I put the small multiples on the bottom so that the reader could discover those annotations and be able to dive into it to get the big impression of the big data visualization, but also be able to discover those pieces of information. This is a piece that we did in the fall. I worked with Albert Sun on it. Albert had gotten his hands on the CDC model that looks at the relationship between international intervention in the countries that were stricken with Ebola and the number of cases that were actually counted. And it was a model. So it's got, obviously, inputs and it's got outputs. So this shows what the model says about the number of cases if intervention began in August 2014. That was fairly close to reality. It was toward the end of August and early September that the international community was really gearing up. I'm really helping, but this is pretty close. And counting those cases was pretty tricky business because there were so many that were unreported. So what we've got here is an estimate. So the light beige there on the chart is your high estimate and then your darker orange-ish is your low estimate. The black line is reality. So the black line's not the model. The black line is the actual number of reported cases. So what I wanted the reader to be able to do is to look and see what would have happened if maybe international intervention had happened in June. So if it had been in June, the number of those cases would have been significantly lower. Even we're just talking about a couple of months of getting money and resources and people, experts into there. But if intervention had happened in October, just two months later, that's what the chart would have looked like. And when the CDC came out with our estimates showing what this would have been, we did a couple of daily charts showing what it would look like. But I felt like readers saw that, but they didn't really understand kind of the impact or the import of this. And when Albert got this data, I worked with him very closely to be able to really work with the reader to be able to let them explore and discover what could have happened. I'd like being able to relinquish some of the control, some of the editorial control and let the readers discover that for themselves. And it's a slider, so you can move back and forth and see a bunch of different months. Always beware, always beware, the false reveal. In my field, things move incredibly quickly. I mean, we do projects within hours. Sometimes we have the luxury of doing them for weeks or even in the case of snowfall months. But we need to be incredibly careful about the false reveal. So not to pick on the BBC at all, but there's something just slightly off perhaps about the slope of this line going from unemployment 7.8 down to 6.5. That just doesn't quite sit with me and you can see they've got a little bit of a broken axis. But fortunately, this community also has a good number of people who are always on guard for this kind of false reveal. We're pretty good at quickly peer reviewing each other. Next thing I wanna talk about is mobile because a lot of my interest in energy and focus in the past year or so has really been moving to mobile and how we can start to tell complex, wonderful, rich, revealing data visualizations on teeny, tiny little screens. And Jim got into this yesterday and I think his presentation was magnificent. But he didn't talk about how hard it is. It's so incredibly hard. This is the one that Jim showed very briefly. This is one we published pretty recently. It was on the end game in Ukraine. So there's, we've written dozens, hundreds perhaps of stories about the machinations within the Ukraine. And I really wanted to be able to distill it down to something that you could see and understand on your phone. And we used swipe rather than scroll for this and sort of give first one. I worked very closely with Tim Wallace and Derek Watkins on this. And there's actually, I think there ended up being about 14 or 15 of these different panels. I'm just gonna show you a couple of them. But we distilled the information down to just the most sort of minute nuggets. So one of Russia's goals is for the two oblasts to become fully autonomous. I can show you that and I can give you a map. During the war, when the separatists were failing a little bit, they got a lot of help from Russia. And so the border was incredibly important. One of Russia's really important pieces in this is that they wanna keep that border as porous as possible so that they can send whatever they need from Russia to Ukraine. And then lastly, one thing that they're really focused on is being able to have a land connection to Crimea, which is one of the things that started all of this because Crimea has some amazing natural resources, but Russia right now doesn't have any way, any way over land to be able to sort of get the equipment that it needs to be able to enjoy those natural resources. So being able to distill all of that down into something that somebody can have on their phone and swipe through took longer than I would actually like to admit. This is a piece that came, it was probably 18 months ago or so, and this is next to the motorcycle and helmet piece which some of you have seen and Jim also showed. I decided to show this one, which wasn't quite as popular, but this was me trying to figure out what we can do on mobile. Is it gonna be scroll, is it gonna be swipe, are we gonna use animations, how tight can I get the text? And when Obamacare was passed, I thought, all right, well this is exciting. This is certainly historical and momentous, and eventually, who knows, maybe 10 years down the line, we'll actually wanna see not just sort of economically, was it a good idea, but health wise, was it a really good idea? So what metrics are available for me to be able to measure it? I can't measure it now, I don't have any results, but I'm gonna keep my eye on it and I wanna be able to look at what metrics are available. So there's a metric called amenable mortality, who knew? And what it does is it takes the deaths from diseases that researchers have decided could have been prevented if the patient had access to good healthcare. And it's a metric that you can compare across countries. Fantastic. So I looked at the United States in 97 and then 2007. 2007, although it feels ages ago, it was one that we had a lot of country by country data on. So within the United States, we have more recent stuff, but country by country, the comparable year was 2007. So I said, okay, well, we're looking better than in 1997, but wow, the slope of that line, compared to Britain or Germany, the slope of that line is not so good. So all right, well how do we stack up to other countries, maybe non-European countries? If you look at that and you look at the slope of Poland or Mexico, I mean they, their amenable mortality rates in 1997 were really high, but wow, they have made some incredible improvements in that decade. In the US, that slope doesn't look too great to us. It looks like we were really sort of beginning to lag a little bit. So those are, that's the metric itself. And then the next question that I wanna ask, and probably all of you are asking, what's in that amenable mortality metric? What diseases are we talking about specifically? Circulatory diseases take up almost half of it. And you can see that circulatory diseases deaths by population in the US, not so good. We're pretty much on the high side. We're right up there with Korea and Mexico, while a lot of our other counterparts like France is doing enormously well. So that's very interesting. I'd like to be able to see next time I look at this. I kinda wonder if the US ball's gonna move a little bit. Infectious diseases is also another metric that goes into this. And this was particularly revealing at least to me this fall because of the measles outbreak. So our vaccination rates are significantly lower than a lot of other highly developed countries. And you can see that in this chart. And that histogram, we're really not doing so great. Because we're newspaper, we get to do a lot of this live. There is nothing quite like elections and Olympics to get your adrenaline pumping. And one lovely thing about elections and Olympics is that readers come to us. Readers are super excited to read and digest and discover and explore anything we wanna give them. And they're really fun times in newsrooms. So I wanted to give you an example from the Olympics. And this is one where an American, McKayla Schifrin, I wanna reveal to you how she won the slalom. Ski races are fast. They're like a minute and a half. I mean, it's like, here they come, there they go. I can't even see anything on that live video. You look at the time and they cheer and that's it. But how do I really begin to explain to somebody how she won? So this is the standing after the first run. She was drawing really well, but they've got a second run. And in the start of the second run, an Austrian actually had a better time than she did. So she really had to up her game. And there were a few of us working on this together, driving it, doing some reporting. And we had an expert, a state-side expert on the phone watching it and as we're watching it together again, like it's a minute and a half, we're doing the live video. I don't remember what time of day or night it was. You tend to sort of work at night if you need to for these things. We had him on the phone and he said, oh my goodness, she made this terrible mistake. And we said, I didn't see anything. She didn't fall down. And he said, that's it, it's over. And it was really interesting to be able to go back and take the stills and actually have him say, here at this very second, you can see she's skiing on one ski. There's no way she's not gonna fall over, but she doesn't. She regains her balance. And you can see where we've marked mistake and recovery. She actually is able to regain her balance and then she is actually able to pick up speed. So it was Will Brandenburg, the state-side that we had on the phone. And he says, it's amazing she was able to do that. And she gets right back on it. And then he tells us the points during the race where she's doing incredibly well. There was a problematic place that a lot of people were having to slow down. And she didn't. And he says, she's basically making it around the gate. She's just rolling her ankles. I mean, it's just like butter. And then so at the end, here she is, she makes the win. So really fun to be able to dissect something like that. In real time, she just raced and boom, we just put that up on the web where you can see right now more clearly than just the live video how exactly she was able to win the race. I like goofy. I think goofy's good. I think having fun is good. One thing that's awesome about this conference is that everybody seems like they're having a really good time. And I love that. And this is a piece that the Times did for the World Cup. Some of you may have seen it. It was a game called Spot the Ball. And you got a photograph like this and you picked where the ball was. I wasn't very good at spot the ball. That's me in the red, not any really near where the ball was. But really fun. And we had, there were a ton of iterations of this. It was sort of, I don't remember how many, but we do for any game. There'd be six or seven of these. People had so much fun with them. It was just like, it was endless good times. Goofy's good. But I think goofy and thinky is even better. So similar technique when New York City came out to say that New York City cab drivers no longer are being tested on their knowledge of the geography. We thought, oh, well, that might be fun. Like you could do a kind of similar spot the ball, only do it for landmarks in New York City. So I lived in Queens in New York for a long time. Now I live in Washington, D.C. But I thought, I'm pretty good at 12 years in New York City. I'm pretty good with New York geography. The Unisphere is in Queens. I almost nailed it. I was very proud of it. The next one starts to get a little bit harder. The next one, Central Park Zoo. I thought I had been there a lot. They have a polar bear, which is sort of sad, but also I grew up in Alaska and I think that anything Arctic is awesome. So I would go and visit the polar bear, although it was sad. And I thought, all right, I know exactly where the Central Park Zoo is. And I was very, very close. However, I was less good at finding New York City. So be it. One thing that is really much more powerful than any reveal is this community. And I'm really, really happy that I was invited to be here. And I really feel like it's an amazingly special place. So thank you all for inviting me, people.