ITV Interview: Alex Libkind, Co-Founder and CEO of Zodiac Gaming

Alex Libkind is co-founder and CEO of interactive TV games company, Zodiac Gaming. The company offers a range of arcade, logic, puzzle, sports, and console-level games for advanced set-top boxes, and has also recently begun developing games for lower-resource boxes. It is perhaps best known for the Logic Pak and Arcade Pak games packages it offers to digital customers of cable MSO, Cablevision.

Libkind recently spoke to [itvt]'s Tracy Swedlow about the technologies Zodiac has developed to enable its games, including its "DOB framework"; about its strategy for targeting the interactive TV market; about his views on how operators could improve the way they present interactive TV games; about the company's product roadmap for the coming months; and more.

[itvt]: Could you tell us a little about how Zodiac Gaming came into being?

Libkind: Well, the core of Zodiac Gaming is basically myself and [the company's CTO] Michael Rivkin. I had come from a financial services background, and Mike had come from Internet applications. We got together in 2002 and we started looking at the cable space--Mike had actually been looking to do something in that space for some time. We saw that the space had a lot of potential, because of the new digital set- top boxes that were coming out. They were, for all intents and purposes, mini-PC's. And, because both Mike and I had worked in software development in Russia, we had incredible connections into the Russian developer community. So we brought these two elements together: we saw that, all of a sudden, there were now these boxes that could run pretty complex and interesting games; and we knew all these developers from our past lives. We realized we could buy games from those developers and port them to the digital set-top--or else bring in those developers to build games from scratch.

[itvt]: The developers had built games for the Internet, right?

Libkind: Yes. Most of them were for the Internet; what we also looked for was experience in older console platforms, such as Sega, Atari, etc. Those developers were able to build very efficient applications and games. We thought it would be pretty simple to adapt them for the set-top box--all that would be needed would be to simplify them a bit.

So we assembled a very interesting and diverse team of developers, and we said to them: "Here's the platform. Here are the platform limitations. Take your existing content and bring it to the set-top box. Push the envelope as far as you can. Make sure that your sound is high quality. Make sure you're getting at least 24 frames per second, multiple active elements and animation. We want to make a big splash in the market."

We had a core engineering team build a high-performance library, tool-set and SDK, which we called the DOB framework--short for "Developed-On-Box." This basically allowed the games to figure out what was available on the set-top box, and to offer such things as 32-channel sound, 3D scrolling backgrounds, and many advanced graphics effects, including alpha blending and emulated 3D. We were also the first company to enable the USB port on the Scientific-Atlanta box to work with multiple peripherals, such as joypads and memory sticks. The DOB framework allowed us to very quickly create new games utilizing the full capacity of the box, without the game programmer worrying about things like hardware and system interaction.

So, through this porting process, we were able to get, for example, a space-fighting game, and a run-and-jump game, with multiple enemies. These were special effects that nobody had ever seen before on the set-top box. We were able to get great sound--not just action sound, but explosions and background music. And all of these sound elements would play at the same time, without any latency. This framework provides a unique technology advantage, and even today we keep trying to push the envelope in this area.

[itvt]: Which Scientific-Atlanta set-top box were you developing for specifically?

Libkind: We were working on the Scientific-Atlanta Explorer 4200 platform at that time. It was a very, very good box. It had 32 megs of RAM, and it had a decent-enough processor. We were able to push the envelope on PowerTV. We got a lot of early support from Scientific-Atlanta. When they saw what we could do, they said, "Okay. For the Western Show (this was the last Western Show, back in 2003), if you can get those same games running on the hard drive of our digital video recorder, we'll put you inside our booth."

So we were able to establish a good relationship with Scientific-Atlanta. We got a lot of exposure to the senior management over there, and they really pushed us. I think we helped them see the power of their boxes for the first time. They were actually very pleasantly surprised by the power of their boxes. The first thing we showed them was an emulator for Sega games. However, we subsequently shelved that emulator, because we realized that the games we could do natively on the box were actually superior to the games that were made for Sega. Sega had had good games for its time, but by then it had had its day. At that stage, the gaming video needed to have much crisper and more active graphics, as well as much better sound. So we decided that it would be much better to just create the games from scratch. Now, with our branded games group, we are looking to bring back the classic Sega, Atari and Nintendo games to allow people to experience the classic gameplay, which in some cases was quite brilliant.

[itvt]: Why did you initially focus on Scientific-Atlanta boxes?

Libkind: We focused on Scientific-Atlanta because they had large deployments of higher-end boxes. Now we have successfully ported games to other platforms, such as TVWorks, that will run on Scientific-Atlanta boxes, and also Pace and even Motorola. Additionally, we have the capability to run on the hard drives of any manufacturer's DVR's.

[itvt]: How is your business funded and how is it organized? I understand you are originally from Russia--do you have operations in Russia?

Libkind: The business is self-funded and operates in the black. Our US headquarters are in New York, and we have satellite offices in Russia: we have offices in St. Petersburg, Moscow, and a number in the Ukraine.

[itvt]: Are the teams located in those offices employees or contractors?

Libkind: Almost all are Zodiac Gaming employees.

[itvt]: When you set out to break into the interactive TV games market, what was your strategy?

Libkind: Well, as you know, there were a lot of companies back then--and there still are--that were playing in that space of 200- to 300-kilobyte games. So playing in the little games space, if you will. But we said to ourselves, "If we're going to create this company, how can we ensure that it stands head-and-shoulders above everybody else?" So the way we decided to differentiate ourselves was to focus only on the high-end. We decided that we would attempt to create the most compelling, sexiest games possible on the set-top platform--even if some of those games weren't immediately deployable. We wanted people to see what Zodiac was capable of.

Then, after a lot of hard work under the radar, all of a sudden, we came out at the Western Show, and everybody took notice. Everybody wanted our games. Cablevision was actually the first one. They gave us a contract for two game packages.

[itvt]: Those are the games they offer as the Arcade Pak and the Logic Pak, correct?

Libkind: Yes. So, once we had made our big splash at the Western Show, we started focusing on actual, deployable content. At that point, we already had a big library, so we just started rolling out games into these two packs.

[itvt]: What have been the main factors in your ability to produce high- end games for set-top boxes?

Libkind: Engineering. Well, engineering and Michael Rivkin. Michael was brilliant in that, at first, he didn't put any constraints on our people. He simply told them, "I want to see stuff that will make us shine." So the engineers didn't worry about real-world constraints. They created games that were huge. We then said to them, "Now apply your past console expertise and optimize these games for the set-top box." So, instead of having them start out with a limited perspective, governed by the constraints of the set-top, we had them start out with a free rein to use their creativity--and then had them adapt what they came up with for real-world deployment.

[itvt]: I understand that Zodiac Gaming's technology strategy is to deliver the game to the set-top box, and then let the box run the application. And, as you mentioned earlier, you are also developing games for DVR's--i.e. games that make use of a hard drive in the set-top box. Could you talk about the advantages and disadvantages of your set-top-oriented approach?

Libkind: The advantage of running games on the box is that you download the game just once. That's a very quick burst on the network, and the network is not taxed until you load up the next game. You don't need to dip into the precious video bandwidth. That means that you can have just one server at the headend, or even run the games on a carousel. It means that our games are not very capital-intensive. You don't have to put 20, 30, 40 or more servers into the headend. Everything is running right on that box. All you need to do on the network is just be able to deliver the game and store top scores and personal profiles. Of course, the great thing about DVR's is that you can build upon this set-top-oriented approach by streaming games to the hard drive, which makes for more powerful games. We've tested that and it works very well.

[itvt]: Isn't there a danger that the games can overtax the box, so that it can't do other tasks as well?

Libkind: We found that we could record two simultaneous TV streams on a Scientific-Atlanta Explorer 8300, while playing games at the same time. We haven't found any conflict there: it's all been tested and it works.

[itvt]: What is the average footprint of your games?

Libkind: The games range anywhere from 300 kilobytes to six megabytes for the console-level RPG.

[itvt]: That's larger than a lot of applications.

Libkind: That's correct. Most of our advanced games won't work on 2000-series set-top boxes. There is a certain profile of box--a minimum box, if you will--that we have to have, in order to run advanced games. However, we are now starting to diversify: we're starting to do a library of games that have a smaller footprint that will target legacy boxes, even some 2000-series boxes. But, fundamentally, the core games that we've built our reputation on are designed to run on any Scientific-Atlanta Explorer 3000-series box or higher, and on any DVR.

[itvt]: A lot of companies in the interactive TV space have resigned themselves to the fact that 2000-series boxes will continue to be around for a lot longer, and have adapted their software accordingly. To what extent does your focus on higher-end boxes put you at a disadvantage in the market?

Libkind: This is one of the reasons we have launched new initiatives to develop smaller games for PowerTV in the sub-500-kilobyte range, and also a whole suite of small Java games for TVWorks, OCAP and MHP. We are now also looking to develop similar games for OpenTV and for Microsoft TV's IPTV platform. Our new DOB2 framework is much more efficient and works exceptionally well across all of the above platforms. We believe this technology is years ahead of other interactive TV platforms.

[itvt]: Could you tell us a little about the games you're working on right now?

Libkind: Right now, we're focusing on branded and non-branded casual games. We've announced deals for branded games with ESPN and Speed Racer, and there are more to come. The next thing that we're doing is moving our games to other platforms, and optimizing them for those platforms. We're taking our DOB framework, which is now in its second generation, and porting the games to it. It is almost like a middleware that will allow us to build games in OCAP, MHP, OpenTV and all other platforms. It's a very robust and powerful framework, which we have even used to build other none-game-related applications on set-tops with great success.

So the DOB2 framework, as the second generation of our framework is called, is a combination of a rapid application development tool-set and a platform which can ride on top of another operating system, such as PowerTV, OCAP, TVWorks, OpenTV and others. This will allow our developers to utilize our best-of-class technology inside of the game and other applications, without having to create these functions from scratch. This also allows us to optimize the games to work with technologies that are proven and scalable without having to re-invent the wheel every time.

[itvt]: I understand that Zodiac has been working to demonstrate compatibility of its games with OCAP, correct?

Libkind: Yes. In fact, we were recently at CableLabs' Interop event: we had a very interesting SpaceShooter game, which again has a lot of animation on the screen, and has a lot of explosions, a lot of other special effects. It runs across pretty much all the OCAP specs, on all the OCAP boxes. It works well on a TVWorks platform, on Vidiom, Alticast and Osmosys. OCAP is a big focus for us right now. The OCAP vendors, such as Vidiom and Osmosys, have actually exhibited our OCAP games at numerous trade shows and we have a very close working relationship with all of them.

[itvt]: In general, what impact do you think the deployment of OCAP will have on the interactive TV games industry?

Libkind: OCAP and the lighter version of it, TVWorks--formerly Liberate--will allow developers like us to create a single game and then modify it simply for the requirements of the cable network, similar to what is done for cell phones. This is instead of having to rewrite the games, which is what you have to do today when moving, say, from PowerTV to OpenTV. We hope that OCAP evolves and gets adopted, but today we are very focused on making our products work on TVWorks, which is a reality at Cox, and on MHP, which is already deployed in Europe.

[itvt]: I understand you have a relationship with ICTV--though you have yet to deploy your games on ntl's or Grande Communications' systems, correct?

Libkind: The reason for that is that, even though we've found that our games work fine on ICTV's platform, we're waiting for ICTV to get enough critical mass to begin spending marketing resources moving our games from on-box to streaming.

[itvt]: I also understand you have a relationship with Arroyo to enable games-on-demand. Could you talk a little about the concept of games-on-demand and why it is significant?

Libkind: Games-on-demand is a very important trend. It is a combination of carousel delivery, which is prevalent at most cable systems, and DOCSIS, which is really only available at Cablevision. The limitation of carousel delivery is the amount of room on the carousel: you are limited in the number of games you can put in a package at one time, if you want fast loading times. DOCSIS allows us to offer as many games as we want over fast, out-of-band delivery. The problem with DOCSIS is that Cablevision is the only major vendor with this infrastructure. Therefore, delivering games-on-demand to the set-top box over the existing VOD pipes offers the best of both worlds: VOD servers are ubiquitous in almost all cable systems, they offer fast in-band delivery, and we can put as many games as we want on them.

[itvt]: Now, the flagship deployment of your games is with Cablevision, correct? What's the business model for your games on Cablevision?

Libkind: The business model is twofold. We have games both on their subscription package--which is priced at $4.95 per month for a package of 12-18 games. And then some of our games are offered on the pay-per-day model, which is $1.95 per day per game. We have games available on Cablevision both those ways. We just released Mahjong for them. And we built Tetris, Centipede and Asteroids for PixelPlay on Cablevision.

[itvt]: What made you decide to partner with PixelPlay?

Libkind: Well, our relationship with PixelPlay was our first foray into branded games. We wanted to see what it takes to develop branded games, and they had the licenses for those games.

[itvt]: Which model is working out better for you: monthly subscriptions or pay-per-day?

Libkind: The one that works best for us right now is monthly subscription. I think the economics and the structure of pay-per-day are still being honed. I don't think we have enough learning yet on pay-per-day to say how best to operate that. We would definitely like to see more pay-per-session gaming--similar to what you see in the UK. We feel that that's a very important way to deploy games. Another thing we'd like to see is this: let's say you've got an operator that wants to get into interactive TV games. Our feeling is that they should launch a super pack of games on a subscription basis, and price it at $7 to $10 per month. Then maybe add pay-per-play or pay-per-session gaming at $0.25 to $0.50 per player per session. That way, it becomes almost an impulse buy. From what we're learning, from what we're seeing out in the market, we think that that would be the most effective revenue model.

[itvt]: How did you come to that conclusion?

Libkind: Basically, from looking at Europe--particularly BSkyB--to see what's worked there and why it has worked. If cable operators want to put games out there, they have to look at what has worked in the past. Subscription packages work. That's been proven in the US by Cablevision. Those packages have been very effective: Patrick Donoghue and Jonathan Boltax at Cablevision have really been the leaders in making that happen. But our feeling is that the next evolution of this is impulse-buying. That's worked well on Sky.

[itvt]: How does your business relationship with Cablevision work?

Libkind: We do a revenue share. In fact, when we talk to any cable or satellite operator about deploying our games, we are interested primarily in a revenue-sharing relationship. That way, we're successful if they're successful, and both sides benefit.

[itvt]: Can you say what the revenue split typically is?

Libkind: I can't discuss that, as it is governed under our NDA.

[itvt]: Can you say how monthly buy-rates vary over time: do you see spurts during certain times of the year, certain games becoming more popular--that kind of thing?

Libkind: Basically you are seeing a lot more activity on days that people are home: weekends, holidays and even rainy days. That is where we get the most volume.

[itvt]: What kinds of marketing do you do for your games?

Libkind: At this point, our marketing has been primarily to the cable operators. Zodiac isn't a consumer brand, we're an industry brand. But as we get into new markets, we're very interested in proactively supporting the operators' marketing efforts. For example, in the case of Cablevision, we provide them with a lot of the consumer art that they've used to promote the games.

[itvt]: How has Cablevision been marketing your games?

Libkind: They've done some print advertising and they've done banner ads on the user interface, and so forth.

[itvt]: My understanding is that for interactive TV games services it's very important to keep refreshing your content. Could you talk about your policy for refreshing your content--for example, how frequently you refresh your games--and about anything else you do in order to sustain consumer interest in your games?

Libkind: You're right. That's a very important part of our business model. We try to release about two new games a month, to put out into the packages. We also try to do two to four reskins of our games on a monthly basis, and also to reskin them on a seasonal basis. For example, we have a single-button game called "Kick the Cooper," which has been very successful. It gets a lot of play. We recently released a Halloween edition, with a Frankenstein Kicking Cooper. Now we've released an "Elf Delivery" version of it, where Santa is rocketing an elf through the Arctic.

[itvt]: I take it you see a jump in usage when you refresh or reskin your games…

Libkind: Absolutely. We see a significant increase in interest. The game gets a new lease on life if you reskin it, and a game package gets a new lease on life every time we release a new game.

[itvt]: Are there any particular types or genres of games that are proving more successful than others?

Libkind: Yes. The kinds of games that we're finding to be extremely successful are single-button games and those that have a very short gameplay--three to five minutes--or that allow users to save their game and come back later.

[itvt]: Could you explain what you mean by "single-button games"?

Libkind: Yes. I mean games where the only button you need, in order to play, is the "select" button. So with "Kick the Cooper," for example, you select the angle, you select the power, and then you press "select" to kick him. Or, with "Treasure Hunter," you use the select button to manipulate the swinging arm and pick up jewels.

So these kinds of games tend to be very popular, because they're very easy to understand and very easy to learn. We're also finding that "swap-and-fall" or "swap-and-collect" games tend to do very well, as well.

[itvt]: What do you mean by "swap-and-fall"?

Libkind: Games where you collect three or four of an item, and they disappear, and then more appear on the screen. Right now, though, our most popular games are our "bubble-shooters."

[itvt]: What is a "bubble-shooter" game?

Libkind: You aim a gun and you shoot a bubble. When three bubbles are together, they fall down.

[itvt]: Which of your games are bubble-shooter games?

Libkind: We have a game called "Aqua Bubble," and from that we've also created new games called "Aqua of Sherwood" and "Sherwood Tales." Those are all bubble-shooters. With the Sherwood series, we combined the bubble-shooter format and the single-button format, and created a super-powerful game. It is almost a new genre.

[itvt]: Could you talk a little bit about your product roadmap for the next six to 12 months? What kinds of games will you be focusing on developing?

Libkind: One thing we'll be doing is developing games for multiple platforms--especially mobile phones and PC's--with a focus on turn- based and simple games. We'll also be working on games that work cross cable operating platforms, so that you can create bigger game communities. Multiplayer will definitely be an important trend next year.

We're also going to be developing various "run-and-jump" games that will probably be affiliated with existing brands. These will be for the teenage male audience.

[itvt]: Could you explain the concept of run-and-jump games?

Libkind: They are games a la Sonic or Mario--meaning that you've got a character, and you make your character run through a world and collect certain things. Those kinds of games are very engaging and work well with a remote.

Another thing we're going to be focused on is sports games. We're developing those for ESPN. Now we're not trying to compete against the consoles. We never could: our platform's not powerful enough. But what we're trying to do is bring back some of that old, classic, wonderful gameplay you had with certain games from the '90's. We're taking those games and bringing them back to life with better sound, better graphics and more action. We're going to take them out to the market with brands on them. Once people see a brand like ESPN or a brand like Speed Racer, they'll recognize it. They'll want to play it and they'll want to come back.

[itvt]: Could you outline some of the things you consider to be important when developing branded games?

Libkind: Well, the key to making a branded game is not to dilute the brand. If you're talking about an existing game, you've got to stick pretty close to that existing game, without dumbing it down too much. That's why Zodiac has a very strong advantage in adapting existing branded games for the interactive TV environment: we can replicate the graphics and the sound that the original games had.

When you're working with media brands such as ESPN, you've got to be very careful with their image. You've got to make sure that the game you develop isn't just a matter of you basically taking their brand and slapping it on something that's not of high quality. Again, Zodiac's ability to do high-quality graphics and sound gives us a definite advantage. We're a proven company for creating the best game quality.

[itvt]: Would it be fair to say that most of Zodiac's efforts and resources will be going towards developing branded games in the future?

Libkind: Yes. We already have 100-plus games in the non-branded category, and, right now, MSO's can't eat everything that we have. So focusing on creating games out of brands is definitely going to be a priority for us.

[itvt]: How do you think operators could improve the way they present interactive TV games?

Libkind: This is an area that we feel very strongly about. We'd like to see more operators take a very personalized approach to gaming. What do I mean by that? There are usually multiple people in a household. So when a member of that household comes to a games portal, they should be able to log in and register themselves individually: the child versus the mother versus the father, and so on. They should each be able to see their personal top scores. They should be able to enter their name once, and, every time they hit a top score, have it recorded.

We also feel very strongly that it's important to have large top scores lists or leaderboards. Not just 20 top scores, but 50, 200, even 500 top scores. We don't want our games packages to just be for gamers. We want people to say, "Oh, my gosh! I scored 374! I'm now going to try to break the 350th position." That gives people an incentive to come back over and over and over again, and it also builds community. It's important that people feel that the games are for everybody, not just for the top scorers, and that they feel that they're making progress every time they play a game.

[itvt]: You mentioned earlier that multiplayer functionality is going to be a focus for you over the coming year. Could you talk a little more about that?

Libkind: We actually demonstrated our cross-platform, multiplayer games at the last NCTA National Show. We made games that worked the same on a PC, on an OCAP platform, and on a PowerTV platform simultaneously. This will enable multiplayer gaming for the whole cable community, not just for those with powerful boxes.

[itvt]: I understand Zodiac is also interested in something called "intelligent menuing."

Libkind: While we have always felt this is a great application of our technology, we've kind of backed off of that a little bit, because we don't see the MSO's requesting this at this time. What it allows you to do is to personalize the menu based on people's gameplay preferences: for example, show the games that they play the most at the top of the menu, rather than at the bottom. Our game portal offers this feature and we would love to see it in the field.

[itvt]: Could you tell us a little more about this portal?

Libkind: It's an end-to-end game portal solution, which we are now rolling out to operators as a private-label product. What I think is really important for us to do right now is to get that portal more broadly adopted--or at least to get the operators to adopt the principals we have built into the portal. The objective of our portal is not only to get people to play our games, but to make it possible for them to personalize their experience, and to feel that they're part of a community.

[itvt]: Do you have any deployments outside the US? Do you have plans to expand your presence in international markets?

Libkind: We have one international deployment. We are currently formulating a strategy for international business operations, which may include the establishment of third-party relationships.

[itvt]: Do you offer any applications in addition to games?

Libkind: We have created some unique applications that work with the USB port, and that allow us to upload photo albums and download music to memory sticks. As you know, people are interested in downloading shows from their TV to their video iPod and in moving pictures onto the TV. By utilizing some of Zodiac's USB technologies and our ability to work with the hard drive of the DVR, we can bring these solutions to market.

We are also looking to use the DOB2 framework as a tool that an operator can deploy to offer advanced interactive TV services. This framework allows the MSO to deploy small, interesting applications—or widgets--without a significant capital investment.

URL: http://www.zodiacgaming.com

Click http://www.itvt.com to subscribe to our free email newsletter, which contains all the news stories you see on this Web site, and additional breaking news and scoops, in-depth features, interviews, screenshots, videos, and other exclusive content you will not find anywhere else.