 Well hello there. Mark Risen Hopkins coming to you once again from South by Southwest 2011. I am here with Mark Phillip. The founder, yeah founder, making sure I get the title right. I called someone a founder when they were a CEO earlier and that didn't work. The founder was over there, got mad. So founder of are you watching this? And this is a website that alerts you when excitement happens in a sporting event. Have I encapsulated that right? I think you got it all. And don't worry about the title. It's the one-man shop. You can call me the janitor. You can call me whatever you want. Chief bottle washer. I'm familiar with that title, yeah. Exactly. So basically I've built a site that I've always wished existed. I hate more than anything missing great games, whether it's a no-hater through eight and I was sleeping or a triple overtime game and I was doing laundry. So I've built an engine that's able to figure out in real time if games are exciting and alerts you to run to the couch. That that sounds like a very necessary thing, very simple, but there's a lot of as you get into this, there's a lot of stuff going on the back end. There's a lot of stuff going on. Finding out what's exciting for a human is pretty easy. How did you get a computer to do this? Well a lot of it has been pretty tough and there's almost this philosophical question of if you have an NBA game that's tied with 30 seconds to go and you have a WNBA game that's tied with 30 seconds to go should the engine treat them the same because the followers it's slightly different in scale. So our objective is to have an engine that is completely cold treats everyone the same. So we sort of layer in users as well so you have this cold objective engine and then you have these very passionate by definition subjective fans and we blend those two together. So the opinion of the engine plus the opinions of the fans when you roll it all together you get the best signal of the games exciting or not. That's interesting so and you're pulling in social signals as well as stats and live stats in some cases or all cases. In all cases if the alert doesn't get sent out soon enough then we've kind of lost the whole reason for setting up. Right. The worst thing I can do is send out an alert that is by the time the person gets to the couch the game is already over. Right. At least they feel even worse that way because they missed a great game. So speed is very important real time is important. We process on some days as many as 500 games one in four days as over 100 sporting events. But our goal is to get an alert out. We run all of our calculations in under 100 milliseconds. So within the 10th of a second we're able to figure out is this game good or not and get the alert to the person within 30 seconds. Generally that's enough from the time we get an update to the time they can get to the couch that if there was a commercial break they get to the couch the game is starting right back up again. So tell me about because you've got a community that it's a lot of crowdsourcing going on. What role does that play in how your operation works. There are super fans. You know they're kind of like the editors that you might see on Wikipedia. There are point zero zero zero one percent of fan that no matter what the sport is they're going to watch it. If it's extra innings in little league they're going to run to the couch to watch it. So we're very focused on getting that super fan to help us curate the data because really the the main goal is to take this engine take this community and bring the data to the living room. That's our holy grail. If you're watching two and a half men might want to do these days. Yeah may have some problems with that now but yeah. Incindication most likely. And you get alert this is hey there's a triple overtime game two channels over or like there was today Northwestern is giving Ohio State a heck of a game. Maybe you want to run to the couch right now. You sort of take it to the next level of you're at South by Southwest all week. You can tell your DVR. I like the Longhorns like the Cowboys and everything ACC basketball. Start recording anytime anything interesting happens. Sports is the most DVR resistant genre on TV to steal from Mark Cuban. And it's a complete troublemaker. All sports fans and all non sports fans have dealt with the issues of trying to record something that sports either gets in the way of or runs long and you miss the end of a game. With our technology we're able to marry TV listings and sports data to really solve that that troublemaking genre of sports. Whether it's extending a game when it runs long pausing recording when there's a rain delay or if you're recording something where it gets preempted by sports we can automatically and proactively shift back the recording so you don't miss the end. Okay so I want to talk about DVR integration because that's something that's really interesting to me but before we do that I want to talk about like your social signals because there's two aspects of that. It's very interesting to me and probably our audience is first of all you've got to be processing a massive amount of social stream data which is you know really kind of the theme of the year is big data and how we how you manage that what tools you're using to to flume that in and what tools you're using to analyze that because semantic is just as much a part of sentiment analysis is just much a part of dealing with big data as housing it can be. So let's talk a little bit about that first. What what what's what are you doing with on the social signal side? It's we've been selective because there is so much data out there. You know big picture wise our goal is to whittle down information to curate with so many games so many channels so many teams so many leagues popping up every day. We really want to whittle it down so we play with certain things we look at Google Trends lightly we look at Twitter lightly the problem with streams like that sometimes is that it's over weighted to the most popular teams. You're always going to see more tweets about the Yankees and the Cowboys and the San Francisco Giants because they're popular teams. So we dabble in those but understand that we want to be unbiased because our main goal is to make sure we have a very accurate engine. The whole NBA versus WNBA thing. Exactly. So if you if you waited strictly there are problems so we try to normalize as much as possible and we definitely pick out for certain cues but we try not to rely on it completely. I think for any site if you rely completely on users completely on social streams you're going to get just the popular teams. If you focus just on algorithms you're going to get games that might seem good on paper but you'll miss sort of the subtleties. I think it's important to mix those two to really the objective side is interesting because it's so structured and so metered. The social side is interesting because it's overwhelming. It's like trying to take a drink from a fire hose. Yeah exactly. And if you're able to get those two together in the right way with the right blend that's generally when you do the best. Right. So you're basically normalizing as an interesting approach to that. That's a term I'm familiar with as someone that does audio engineering hadn't really thought about it as applied to social stream signal data. But so you normalize the streams and you apply I guess some sort of scoring to that and then compare that as an element of like the stats and other elements that you're using in terms of of excitingness factors in games. Trailing averages are huge for us. If you look at just the social streams understanding that people are always talking about the Yankees. So things have to be relative. It's really important on the objective side we use trailing averages also because engines actually able to learn and understand which games are going to be good before they actually even happen. It understands Yankees Red Sox. It understands Texas OU Duke UNC because able to look at the past games understand what people have done previously which games were exciting and understand OK these teams always play good games. This game coming up in three weeks. You need to make sure you're in front of the TV. Right. Very interesting. So let's let's go back to talk about the integration is this is where it's all this is where it's all going is is almost a divorced experience from the cable providers and the telco providers. Yeah. Sports is probably one of the last hooks that keeps you in there. You've got you've got a you know Xbox and ESPN doing that thing. Yeah. I guess it was a ESPN three is what they call it. It's in some places obviously not a hundred percent coverage of all all the sports. Everybody's going to watch. But you see stuff starting to go that way. And of course the the major providers are also enhancing the experiences through the DVR's. What are you guys looking at. Where do you see everything going. And where are you focusing your efforts in terms of integration of these experiences. My big word is curation. You know whittling down the noise. You have two point five million different combinations of zip and postal code cable and satellite provider across the US and Canada. No site out there really does a good job of telling you here's a great game. Here's the channel number to turn to. It's that simple frustrating thing that we all deal with this yelling of I can't find the game. Someone tell me where it is. So the ability to whittle down these hundreds and hundreds of channels we get and tell you here's what you should watch. You're the top three things that matter to you. Click on one and we'll take you right to it. Right. We have a proof of concept Google TV app that's out right now that does exactly that. And it's the only app I've seen that actually changes the channel for you that actually complements the content coming over the pipe. ESPN is a juggernaut. They get four dollars plus a sub a month for everyone that has it. And they are the glue that is holding the cable together for a lot of people. So and we just came from the Samsung lounge where they're showing off these TVs with all this app integration. One of the most troubling things for me because we do a lot of video obviously and we're we're trying to figure out the best distribution strategy on as an independent producer. You know we can't go straight to cable. You know that's a big investment. You know come up with a million dollar bucket of dollars and you know then you're then you're good but we got to go kind of more the DVR approach the Xbox approach the mobile approach. But when we look at the the opportunities with apps on TVs it's exciting at first until you see the fractured landscape of all these different app platforms. You're obviously in that world too. So what are your what are your thoughts on the fracturing of these app platforms. I think it's very similar to what we saw on mobile. I think there will be a similar land grab as well when the set top boxes finally become robust enough to support these apps. It is something that we're going to have to deal with. I don't think there is going to be this sort of HTML5 underlaying layer that will be able to develop to the way we did with our mobile apps. I think they're always going to be fractured and it's something that's very frustrating. Do you see it living on the TV or do you see like players like PlayStation Xbox and and like Foxy and Roku coming up underneath and supplanting any apps that may live on a TV. I think yes and yes and then there's even the Comcasts of the world. They are coming out with new set top boxes later this year that are going to blow people away. I can't remember his name. A guy came from Plaxo. Runs social. Mike Berkley maybe? Ryan Cain? No no I'm going to look like an idiot because he's going to watch this and I've got a mental block on his name but anyway he was talking about they brought in Plaxo for that very reason so they could run social signals as a part of the set top boxes so as Comcast likes to get public likes to beat up on them sometimes for good reason. Sometimes. But in terms of being social savvy I think they probably got a better handle on it than most of the cable co-telco providers. Absolutely their social technology group is very sharp and I think they're really going to lead the way for the next generation of set top boxes. The type of boxes that I've been waiting years for when you bring an app like mine into a platform like that. I think you can create some really compelling experiences. We love the Red Zone channel. I remember Flying Cross Country one Sunday on JetBlue and the Red Zone channel made the trip feel like it was an hour long and it was the best trip I've ever taken. Imagining it to the point where you have a Red Zone channel for every sporting event on TV. Being able to bounce from great game to great game from buzzard beater to overtime thriller. I think there's a lot of innovation that can be done in the living room. So how deep do you go into the sports? Do you go all the way like ESPN 8, the Ocho? We cover it all. Everything. So Little League if it's if there's some fans around there ready to curate the data you got it. We got it all. If you give me a feed I will give you an algorithm for it. Anything from NASCAR to minor league hockey to minor league arena football which until a couple years ago I didn't realize even existed. Yeah, that's pretty awesome. All right. Well, I really appreciate you taking the time. Thank you. You can go check it out at are you want are you watching this dot com? That's right. No, yep. There you go. So go check it out. Really cool stuff here at South by Southwest. We'll come at you more with, we'll come at you. See if I can close this out and buy a vowel. Come at you with some more interviews coming up next. Don't go away.