Developing the Twitter Game…Beat My Tweet

Michael talks about developing the Twitter Game “Beat My Tweet”

You can download the podcast here…
http://www.indiegamepod.com/podcasts/beatmytweet-interview.mp3?v=4

Or listen to it here…

Show Notes:
Michael: Hi, my name is Michael Newman. I am a web developer and graphic artist.

Interviewer: Great. What game did you recently work on?

Michael: I recently developed the Twitter game, Beat My Tweet.

Interviewer: How did you get into game development?

Michael: The middle school that we went to in Massachusetts had computers, and they started us with Logo, which is the turtle as people know it; then moving the turtle forward and drawing the lines and so forth. I think my brother had bought an Applesoft Basic book, and I would take it from him and bring it to school. And we would sit inside during lunch time and try to program, you know, draw a diagonal line across the screen.

From there, it kind of led to making games in Basic that were sort of Zorker, Hitchhiker’s Guide to the Galaxy knockoffs.

Interviewer: Nice. Once the web started taking off, did you think about doing games online, or was it just more of a hobby or what? Where were you thinking about taking your game development talents?

Michael: You know, I stopped using a computer, I guess, for about six or seven years, kind of in between that time in high school and just immediately after. I was more involved in skateboarding and running around doing stupid stuff, I guess.

In college, I went to school for painting and had a friend who was in graphic design and started using their computer. I quickly realized you could bring video into the computer, and you could start to combine things very quickly and not necessarily easily.

From there I got on, I guess, it was Director. I worked at a computer magazine as an advertising assistant and was able to get evaluation software. So, as soon as Shockwave came out, it was like: Oh wow, this stuff is on the Internet and very quickly just started and got right back into it, I guess, from where I was as a kid.

Interviewer: Oh, cool. What games did you develop in either Shockwave or Flash?

Michael: Quite a few. I was fortunate enough to be in New York City when a lot of the Internet stuff was happening and went to work at Viacom Interactive. It was great because it was a whole group of people from a variety of backgrounds who started doing web stuff. I was, I think, one of the few people at the time there who had been doing any Lingo programming. I did all kinds of games, from little sort of advertorial games for Pillsbury and little MTV games.

Viacom ultimately had me do a game for the defense college or the National Defense University which was this large classroom simulator that we actually installed at the National Defense University at Fort McNair so that was one of the bigger ones.

Interviewer: Oh, cool. When you were developing these games, were you experimenting much then, or was it mainly doing projects for companies, or how did that work because I know right now you have the Twitter thing and that’s more experimental. So, I wasn’t sure if you did that because back then Shockwave and all that other stuff was pretty new relatively.

Michael: You know, I think it was all experimental. I would be at work and I would work on things that were very specific to client projects and website projects. And then, I would go home and I would continue to work on stuff just for my own enjoyment.

In New York City you’re always out running around on the street and stuff, and I was very excited about creating a three car Monty game and went home and did it. It wasn’t video, but it was still frames from videos and stuff and took photos on the street. Then, people at work would see that and they’d go like, oh wow, we’ve got this idea for how we want to take that and leverage it into a marketing campaign or what have you.

Interviewer: Sure.Was it mainly Shockwave, or did you also experiment with Flash? I know back then Flash was pretty primitive. I’m not sure if it was enough to make games.

Michael: You know, I think when we got it, it was future Flash player or something along those lines. But, some of the early stuff, I was working on some games for Nickelodeon and I was creating elements in Flash and then bringing them into Director and manipulating them from Lingo. It was pretty cool back then, and it was painful to go from programming in a nice, rich environment to Flash’s – what was it – 300 pixel wide screen or little text box they give you.

Interviewer: So fast forward and then you start officeevil.com. What was the inspiration for that?

Michael: It’s working in a number of corporate cubicle jobs, I guess. It was trying to put together a site which would keep me focused and working on a variety of things, so it’s just a parody, hopefully, a fun parody of corporate life and cubicle hell, I guess.

Interviewer: OK. You also mentioned that it’s also used to just keep up on emerging trends or new technologies. Is that what motivates some of the projects for the site?

Michael: Definitely. I find and I’ve always found just on anything that I’ve done on the computer that I have a hard time with tutorials and just sitting down and just learning something for the sake of learning it. I love learning things on the computer, but I find it is best for me, at least, to have an end goal and say, I want to do this and then do it.

I find that the Office Evil site has been a really good motivator to come up with an idea for a project and dive right into it and then say: Now I need to learn this. Now I need to learn that. It takes you down all these different paths. I find at the end of it you’ve learned a whole variety of things.

Interviewer: Absolutely. How long does it usually take then to do a project when you decide to do a project for a site?

Michael: Pretty fast. You know, I was calculating the other day. I think I’ve spent over 30,000 hours in front of a computer over the last 10 years, and I actually get out quite a bit for God knows how. I guess I don’t sleep a lot, so I spend a lot of time just sort of cranking things out.

Interviewer: Sure.

Michael: The more you do it, the faster you are adapting to new things, and certainly all the open APIs, like Twitter has done a really great job of having their API and all the documentation for it, and it’s really intuitive to be able to dive into it and say: Oh, here I want to make an adjacent call and then you set up a basic class in Peachtree [?] and you just read and code at that point.

It’s fun because you can start to do these things really rapidly. So, once you come up with an idea you can sort of, at least, get to testing it quickly and then start combining it with other things and get it out there.

Interviewer: Before the Twitter game, have there been any other projects that you’ve done that have just by accident taken off or gotten interesting exposure?

Michael: I used to do a lot of experimental short films, and we have some stuff in film festivals and by the way it would be the group of people we were working on things with. Yeah, some of those start to take off in a sense that small niche groups of people would see it and say: Oh, I really like the credits in that piece. I’d like to talk to you about either coming to work and doing some of that type of work for us and stuff like that.

Interviewer: What then inspired making a Twitter game?

Michael: You know, my wife and I were driving to IKEA and I had been talking around Twitter. Twitter had been using it with the persona of Office Evil and using it to communicate with people and start talking and stuff and developing a dialogue. And just started playing games, throwing out words and playing hangman and stuff.

She and I were talking about different types of games that would work with Twitter, and it just kind of came from there, just sort of a drive across town, I guess.

Interviewer: And so you discussed ideas. You came up with some ideas, and what was the next step then?

Michael: You know, it was really just to start testing formats. I was on Twitter and testing, throwing out word scrambles manually and different types of questions and seeing what kind of responses would come in from the group of followers that I had on Office Evil.

Interviewer: Cool.

Michael: People very quickly got it, and I tried to send out the questions in the way I would if it were an application, like the way the computer would throw them out, and see if people sort of intuitively got it and they did. People, the Tweeple, I guess, it’s such a great community of people that are engaging and sort of inspire you to talk in Tweet and do all kinds of stuff. It’s fun. It was great.

Interviewer: Sure. What was the response when you were doing it manually? Was it just people were getting addicted to it, or was it creating more followers for you, or how did that work?

Michael: Yeah, it was pretty small. The Office Evil list of followers is in the high 300’s. It’s really not huge, but it’s a lot of regular people just sort of bouncing things back and forth and stuff. I found people just started diving in, wanting to know what the answer was and play. I was surprised how quickly people got it and were playing it and saying, oh, throw out another one. Throw out another one. This is great.

The immediacy of that is amazing. I talked a little bit about films and all that development time. You just don’t get the immediacy of what you have with social networking and all these new technologies. It’s amazing.

Interviewer: Absolutely. Once you did it manually, what inspired you to be like, OK let’s figure out how I can automate this or make this its own little game in and by itself?

Michael: I think it was a combination of making something new for the site for Office Evil and saying: Oh, I want to have something fun that people will interact with in a multitude of ways so they can do it from their phone. They can do it from the browser. They can come and see this leader board and search it. Then, it was the challenge of saying: Can I do this?

The first thought was yeah, it’s probably very easy to do because prior to that I had already been using the Twitter API. I had made a Google Map Smash Up. That tweets when somebody posts and it’s stored in the database, so it was really easy to send out Tweets. And I was like, OK it’s pretty easy to read replies and stuff; just the challenge of mashing it all together and getting it out there.

Interviewer: In terms of technology, you pretty much then just had to mess with the Twitter API?

Michael: Well, it started with the Twitter API and my first thought was just to run it all through My SQL database because that’s so easy to interact with, but then I was thinking, well, if all of a sudden this goes out or if I’m running four games an hour, 24 hours a day, and I’ve got 10 answers coming in per game. Very quickly, I would be inserting 960,000 rows of data into the database. You know, that’s not a huge amount, but my thought was: let me separate it. I’ll set up another My SQL database.

I signed up awhile back as an Amazon Web Services Developer or for my account, and I very much wanted to use that in some type, in some way and started thinking, maybe, the simple DB application or, I guess, the package that they offer would be a good way to do it.

Right now, they’re giving 25 machine hours a month for free in your first gigabyte of traffic, and I was like, whoa it’s basically free. Why not? It just sort of went from there diving into their tool set and API to be able to get in. There’s a pre-existing PHP library that they have so I was just basically writing my own web services to interact with that.

Interviewer: Did it take awhile to set up an image on the Internet web services, or how did that all work?

Michael: It wasn’t really too difficult, I mean, in terms of settting things up. The concepts are a little different from My SQL or from a SQL database in that you don’t have the same flexibility. It was interesting because as I was developing it, they were coming out with the Select Statement. Like, oh now you can do select work. Oh, this is great. Midway through they came out with the Account Query. I was like, OK great, now I can count things.

Some of the challenges that would have been a lot easier, like things would be a lot easier to do in My SQL, like the leader boards and stuff, but I was just figuring out how to model the data. It was me laying on my back outside by the pool, staring up at the sun with my dog swimming in the pool thinking: How am I going to make this work? So, it just kind of came together, I guess.

Interviewer: Are there any scaling issues since you are using Simple DB? Are there any scaling issues that happen or do you even have to deal with scaling issues anymore now that it’s on Amazon?

Michael: You know, not yet. I think my machine hours are still well below the 25 hours a month, and my data is probably piling up, but what I’ll do is once it gets to a point, if it becomes a vast amount of data, I will go back in and export the data and store it somewhere else, store it in MS3 or something. Costs are just so economical that there’s not any reason not to.

Interviewer: So, you’re using Simple DB on EC2. Are you using pretty much all the different web services from Amazon?

Michael: No, essentially for this project, really just the Simple DB because they have data storage that’s associated with it, and my server itself is handling the crown jobs for the actual game.

Interviewer: Gotcha.

Michael: You know, the EC2 stuff is something I really want to get into and just actually using it for all of the hosting at some point, but just cost right now I don’t think it’s worthwhile for what I am doing at this point.

Interviewer: Were there any other technical challenges that you encountered while you were developing this game?

Michael: You know, one of the things with Simple DB at the time, I was trying to load questions in so I had 500 words or 600 words. And I just wanted to write a script that was going to literally look through the questions and stuff them into the database. That seemed simple enough to me, but when I ran the script I found that my data got messed up based on the latency issues with Simple DB.

So, what I ended up doing was actually writing just a real simple script that had a pause in it so it would insert a query. It would check to make sure the query went, but it wasn’t really as intuitive at saying: Oh, I want to do this batch. Upload data and pull it in.

I actually got an email from Amazon the other day that they added Batch Query so you can batch data in, so I think they are constantly tweaking and changing it. I am happy to see the rapid development on their end.

Interviewer: Great. On the Twitter side, did you have issues? Do they limit the number of tweets that you can send out automatically?

Michael: Twitter has rate limiting, and right now it’s a hundred queries per hour. I’ve been sitting there noodling with it, doing some research to figure out what actually counts as a negative one to your hundred queries in an hour. Certain things count. Certain things don’t count. The way I’m doing it, it’s not an issue. Apparently, it looks like you can pull in 3200 replies in one, I guess, increment of your rate limit.

Their documentation is a little unclear. One place it says 200. Another place it says 20. Another place it says 3200.

Interviewer: Exactly. Well, let’s move into the game design. Aside from the leader board, were there any other interesting game design mechanics you had to add or game mechanics that you had to add to make the game more compelling?

Michael: You know, it seemed that the leader board is pretty much the big component of it that got a tremendous amount of traffic, that wants to come back in and see how they’re doing, I guess, whether it’s on a daily basis, a monthly, weekly or yearly.

Interviewer: So you do have leader boards then for the different segments of time.

Michael: Yes.

Actin: Do you have leader boards based on, maybe, friends’ lists or stuff like that?

Michael: That’s an interesting concept, and we’re looking at different ways to sort of expand the social component of it. One of the things we quickly put in was… A good friend of mine who had worked with a number of things, my friend, Clay, had developed the – it’s all very tongue in cheek – the BeatOff. which you can see you can compare user names and stuff, and it’s just user to user. We’ve been bouncing around different ideas about how to scale it out and change it so I think that concept of groups and playing against groups is a great idea.

Interviewer: What has been the response then from the players? Do people just play for a little while and then take time off, or do they bring more people in to play, or how does that work?

Michael: You know, it sort of comes in waves, I guess. Initially, right when it went up, I started promoting it through Office Evil and quickly got 20, 30, 40, 50 followers, and they were all playing it sort of constantly and there was a write-up on Tech Crunch and that brought in 2000 page uses in a day and then subsequently went on Mashable. I think there is some carry through that some people probably come in and play it for awhile, and then it kind of moves on to their friends and just sort of trickles out.

The way the game is actually set up, I had it set up that the engine that I built will work with basically any type of questions. Right now, it’s just doing word scrambles, but it works with Trivia. It works with Fill in the Blanks and so forth.

Interviewer: Are you thinking then of branching off into other categories? I know Trivia is very huge or popular thing among many people. I don’t know if that’s something you’ve considered.

Michael: Definitely, and one of my challenges is right now, what I am doing with the Office Evil site and all these projects and stuff, it’s just me working on all of it. So, it’s when do I sit down and write 10,000 or 1,000 trivia questions.

Interviewer: Oh, yeah.

Michael: Just the other week I put up a job post on Elance.

Interviewer: That’s what I was going to suggest.

Michael: Going through that and I think Trivia is the next step. One of the debates I was having was the Office Evil site being office parody related, and then Beat My Tweet was separating it to make it more sort of general trivia, or the idea of saying, should it be somewhat office related or work related. That might be boring.

Interviewer: And so then, what’s in store in terms of future Twitter games or future games for you?

Michael: Well, I’ve really been doing a little bit of iPhone development and just love being able to sort of dive into that. I’ve got my Apple business developer account for being able to release applications. So wanting to expand from Twitter and open things up, so you can play it on the iPhone. You can start to play similar games via Facebook so you can get in and play the game from anywhere, I think. That’s kind of the next step, and then expanding it more so it’s a lot more in terms of multiple types of questions and different formats.

Interviewer: Are you going to experiment with other types of game play, or have you thought of other types of game play that you can put into Twitter?

Michael: You know, I’ve been giving it a lot of thought. Right now, the trivia stuff seems to work very well, but I was thinking about the concept of – and it started this way initially, I guess – Hangman because Hangman is much different from a word scrambler trivia. Hangman is that you’re constantly getting answers wrong and so forth and ways that you can have multi-player games or group games.

Interviewer: Exactly.

Michael: I think that’s sort of the next step, and I don’t think Hangman is the answer for that by any means, but it’s some way that you can have these sort of group mass games where you’re developing swarms around different questions or tasks that come up. And getting people to participate in incremental ways that help move the game play on.

Interviewer: Will you have time to even experiment with that now that you are doing the trivia stuff and some of these other things that are more of an extension of your current game?

Michael: Yeah, I think so, actually. The Twitter game is nice in the sense of, it’s actually been running for a couple of weeks now non-stop so when I first got the post on Tech Crunch I woke up in the morning and had 800 emails in my inbox from Twitter followers. My first thought was panic. Oh my God, I hope this works and that it’s stable. So, it seems self-sufficient and part of it is just adding new features on. So, it was very modular. I just try to set things up that way that it’s easy to go back in and tweak.

Interviewer: Great. How is the post, the exposure on Mashable and Tech Crunch? Has it either helped your game or helped to develop or understand or get more ideas for more games for Twitter?

Michael: It certainly helps in the sense that a lot more people are playing it. I think the more full the leader boards are and the more the incentive is to have this time based component to it, it just creates a little bit more of a frenzy of people playing it. I have to say the biggest thing for me is the motivation. When you see something, it’s like: Oh, wow. People like this, and then it kind of prompts you to want to do more and better things from it.

Interviewer: Absolutely.

Michael: I think with positive motivation it can really inspire you to do just about anything.

Interviewer: Exactly. Can you talk about then where you see this social media explosion. You talked about going on multiple platforms with your game. Do you see it more as like a game that is just trying to use all these different social media channels to express itself, or do you consider each of these different social media spaces as something different, or how does that work?

Michael: You know, I think they are all different in a sense, but there’s obviously ways that you can start to link them all together, and people have been doing that. There’s a lot of great applications that are tying things together, and I think the Beat My Tweet game in and of itself might sort of stay where it is but taking what it learned from that and developing it into new games and new sort of interactive experiences that people can work with and play with.

Interviewer: Great. Any last words then for future Twitter developers or other experimental game developers?

Michael: I would say we’ve all got such a great emerging tool set and existing tool set, and there’s so much that you can do with open source technology. It’s just to keep developing things and just be inspired by what other people are doing. With everything on the web, how quickly you can get into stuff and start doing it. Just keep making great stuff.

Interviewer: Great. We’re talking with Michael from officeevil.com. Thank you very much.

Michael: Thank you.

Interviewer: Take care. Bye.

Michael: You too, bye.