eSim Games - Steel Beasts Marketing Q & A.
26 October 2004 - Bard

Nils Hinrichsen from eSimGames has kindly agreed to answer a Q&A about the trials and tribulations that a grass roots sim developer encounters breaking into the simulation market. I would like to thank Nils for taking the time. If you haven't tried steel beasts, you're missing one of the finest multiplayer tank combat simulations of all time.

1) First of all, Who are you and what is your developmental role?

I am Nils Hinrichsen. A short bio can be seen on eSim Games' homepage: http://www.esimgames.com/company.htm

As far as development is concerned: I participate in the following areas - conceptual work, providing both the 3D designers and sound engineer with "raw material" (photos, video, and audio recordings), and give some technical expertise to them if there are unvertainties and we must guess some elements (for example, thermal textures should be realistic, but footage is not easy to come by, so we have to make some educated guesses here and there). I also try to coach the map and mission designers, and quite naturally I spend some time testing the code. In the past months I have worked on a new model to calculate armor penetrations and component damages.

Al Delaney provided me with a 3D painting program where I can load the lowest resolution models (the "LOD3"), and literally paint the triangles. For every of the 62 colors, I can then define a significant table of statistical data for different ammunition types and associated component damage likelihoods. This is a radical change from previous versions where we had a rather global damage model. Granted, it still was among the best ever conceived for a tank simulation game, but we felt that we could do better, so we did. It's not a trivial thing to do. While I get pretty good estimates of armor strengths from Paul Lakowski (who kindly assisted us for SB1 as well), these figures don't tell anything about damage likelihoods for components. Here I can dwell a bit on my engineering background as well as my knowledge about the internal layout of the various vehicles to estimate the consequences of a penetration. This may sound as if I were doing most work of the development, but that's not true. Al Delaney still is the only programmer that we have, so HE is THE developer.

Ed Williams is still doing one hell of a job with the sounds - more on that later.

Raino Sommer, our lead artist - well, he's young, but an exceptional talent, and even more: He's also managing the other computer artists, and doing a good job as well. Everybody can see that, looking at the screenshots that we released at www.steelbeasts.com.

Curtis Ratica, he did work for SB1 as well, provided the nice looking trees, ground textures, and most of the buildings so far. Bruk Marat left the team, some freelancers added vehicle models, and then there's the many, many volunteers who are working as voice actors for the various localizations, who are translating, and doing other thankless, low profile jobs.

Leonard Norman and Sean Dooner are working on maps, and we have recruited many scenario designers from among the Steel Beasts community who are very busy.

Carsten Lauterbach accompanied me on some tours to capture video and audio recordings, André Henkel and Chris Schatton provided additional video material - well, I don't think I could mention everybody, but each and every single contribution was valuable and is appreciated.

2) The simulation genre is often called a niche market, where the PF primary product are flight simulations. Armoured sims are a niche within a niche. What were your initial thoughts on the commercial viability of Steel Beasts?

In the beginning we didn't think too much about it. It simply was a topic of interest for us, and I also think that we were blissfully ignorant about the obstacles ahead. What we did know was that we would address a minority, that we were newcomers in the game industry, and that we had absolutely no development budget at all. We knew from the start that we didn't have the resources to compete with then established big name developers like Microprose. Our competitive edge was being small and efficient, and that we had unique know-how and were reckless enough to focus exclusively on technical and tactical accuracy. Reckless insofar as it was going to cut us off from a broad success right from the start.

3) How did you maximize your chances of commercial success?

Shortly before the release, we were lucky. Panzer Elite already had attracted modders, and one of them contacted us to become our sound engineer. Ed Williams - Volcano - really did a tremendous job to improve the overall immersion and feel of realism with his sounds. After him, Steel Beasts was never the same, and only through him we learned to deeply appreciate the importance of good sound. Also, we couldn't spend much money on looks. Again, we were lucky to be contacted by Geoff Coovert who added more playable vehicles which were incorporated in version 1.10, about six months after we made initial contact. Commercial success - well, for a long time we were looking for a publisher, and all our experiences were a bit underwhelming. We never had a serious budget to spend on marketing, so we relied very much on releasing a strong demo version, and going into many discussion forums. We immediately found a number of people who were very enthusiastic about even the beta demo, and I guess they went out and told other people about it. That generated quite some interest. Marketing need not always be expensive to be effective, but certainly good marketing is absolutely essential for any success. It also seems that we were somehow lucky - being in the right place at the right time with the right product.

4) How did you fund development?

Well, in terms of actual bills to pay, there wasn't much. Over a time of five years we had to finance approximately $50,000.- to $80,000.- With about $10k per year, one might still treat it as an extreme form of hobby, but that's only half the story. Let's keep in mind that as a Silicon valley resident engineer and programmer during those crazy years of Dotcom hype, Al Delaney could have made several hundred thousand dollars at least, and probably way more. So, if one is analyzing opportunity costs, Steel Beasts certainly did not represent good business judgement. It was and still is a child of passion. I had a day-time job, we also had relatives to keep us fed, and as we were willing to abandon a private life, we could achieve as much as a conventional developer team of five or six guys with a "nine to five" attitude in the same time. Our collaboration - at least that's how I always felt it to be - was always cooperative, focused, oriented on the outcome. We didn't have to make many compromises, and we never settled for a bad one. It went remarkably smooth. Since then, we spent half a million on the new engine, I'd say. That's a tiny budget by industry standards, but almost ten times as much as SB1 cost us, and I think that it's quite evident in the recent versions where we spent the money, if you compare it to SB1.

5) How did you advertise?

We depended a lot on product reviews. We were lucky in a way as most critics were honest, appraised the strong points, and pointed our the weak elements that deserved criticism. I think I can easily live with being criticised for the right things, that's fair treatment. But if your work gets hammered because it is not what it doesn't even attempts to be - that hurts. Well, apparently the critics understood what Steel Beasts attempted to be, and consequently the reviews were usually very favorable. We won awards, and our publisher printed a lot of these on the retail box which probably contributed to a significant share of spontaneous buying decisions. So, yeah: Reviews are important! (editor's note: reviews here )

6) When it came to initial publishing, what kind of reception did steel beasts receive from publishers? Was it seen as a viable product? Were you forced to look at alternative methods to sell and distribute your product? Did this change as time went by?

We were looking quite some time for a publisher, and received many rejection slips. Hey, even Bohemia Interactive had a hard sell with Flashpoint, so one shouldn't overestimate this. Publishers are in a tricky situation themselves. 60% of all computer games do not break even. Of the rest, 25% are only moderately profitable, 10% could be considered serious successes, and of the remaining 5% real hits, only 1% are the real blockbusters - the Doom, Half-life, Warcraft league. Just look at the annual reports of the large publishers - EA operates with a turnover profit rate of just 5%. To cover a single title's loss of $100,000.- , they need to gross in $20 million with other games! Needless to say that if so much money is at stake, a publisher is hesitant to try a new formula. Just look at how many publishers stay in business for more than 10 years, and/ or are not being swallowed by others. It's a tiny minority.

Anyway, so we found a small wargame-oriented publisher who would be selling exclusively through his web shop. After a year we found that it made no sense to continue the relationship. We then wanted to self- publish SB, and did so for three months with version 1.15, then that other publisher contacted us, and we decided to give it another try. Well, the good news was that we could push Steel Beasts into retail channels that way, and that it sold way better than anticipated. Later the financial problems of the second publisher became apparent, so we decided not to extend the contract that was running out anyway, and the financial loss, again, could be limited. What really stinks is the loss in productivity that results from these struggles.

If people are asking me, why did it take such a bloody long time to come up with the sequel - well, we commenced work immediately after the release of Steel Beasts Gold in spring 2002. Publisher trouble cost us half a year. Another six months can be contributed to the massive changes between DirectX 7 and 8.1 that forced us to rewrite every single code line dealing with 2D graphics output - menus, map views, drawing simple lines - where MS didn't offer a soft migration path. Changes between DirectX 8.1 and 9 weren't affecting that much code, so the transition went by rather easily. So, if you strike six of those unproductive twelve months, we're down to two years of pure development which isn't bad. It's only bad if you're looking at the original release date of Steel Beasts (August 2000), but we spent one and a half year with free feature additions, and the business trouble of changing publishers etc. in the background.

7) Were there other obstacles you had to overcome during development?

The biggest problem always was, and still is undermanning. We could easily double or triple our staff, there certainly is enough work to do. It's just that we don't make enough money to hire. We're getting by. We're not starving. We made enough money with SB1 to advance from "no budget" to "shoestring budget" with Steel Beasts Professional and Steel Beasts 2 (which are sharing the same engine), so that in itself is a noticeable improvement. Since the release of the original Steel Beasts, we started to develop a second market - Steel Beasts as a training application for several armies. But despite the well-known stories of inefficiency and glaring waste of tax dollars, somehow they never get wasted on us. ;) These military customers are returning, so it looks like they're happy with our products. But sales are often very hard to come by, they usually need several years to transition from evaluation to buying decision. To overcome the huge inertia of the apparatus - boy... But it IS possible, that's the good news.

8) How would you characterise the sales trend for steel beasts?

I cannot give precise figures here. But Steel Beasts was more of a long-seller than an immediate smash hit. That also was a result of the complicated publishing story and their different strategies, so it's a bit hard to characterize it as a representative of a certain type of seller.

9) Has the sales performance of Steel Beasts opened more options as a developer for you?

Yes, like I mentioned, we at least have a shoestring budget now. We can afford to hire artists, and pay about ten times as much for their service than we could with SB1, and I think it shows. Likewise, the customization deals with military customers have provided a cash flow that was crucial now and then, and it pays a substantial part of the development. The game version - Steel Beasts 2 - will profit from those jobs as well.

10) How much (if any) pressure were you under from outside sources to release the product by a deadline?

Since we've always financed everything out of our own pockets, we enjoy an unusual degree of independence. There are however occasions where deadlines simply make sense, e.g. to profit from cyclic buying habits of the consumer market, and then we better adhere to such a deadline. But we have never rushed a job and released something where we had a bad feeling about code stability, or customer value.

11) It seems that besides Microsoft, there are no longer any 'big players' in the simulation market. Do you agree with this statement? and why?

Well, any of the bigger publishers could decide to publish a simulation game if a developer had a game ready, and if they were sensing a market opportunity here. But I cannot give profound statements about the general market situation. We concentrate on our field of expertise, there's enough to do for us. Why should we bother with a train or flight game, only because they use the methods of simulation tech- nology? A tank action game would be much closer to what we're doing, even if it were no simulation at all. I think that customers are not buying Steel Beasts because it is a simulation, but because it deals with modern tanks, and they have developed some interest in this subject matter. Whether the game promises an authentic experience - either explicitly by stating that it is an accurate simulation, or more implicitly with nice and convincingly looking screenshots - is a secondary consideration. Some may just be looking for action, and to blast their way through enemy lines with impunity. Motivations vary a lot here.

12) I understand that Steel Beasts is being used in armoured warfare training, how has this affected your future plans?

Yes and no. It was designed that was from the get-go. But at some point we realized that there sometimes are diverging development goals that cannot be met simultaneously in good compromise. We decided that if we wanted to cater to both the entertainment market and the market for serious training, we'd better split Steel Beasts into different variants that would focus on either target group while sharing a similar technological basis.

A common engine makes a lot of sense since the basic requirements are the same. So, we have now Steel Beasts Professional for "collective training", that is a group of students simultaneously in a classroom or computer lab, ideally in a local area network. It's "the full monty", so to speak, including a powerful instructor mode where the session host in essence has the same power as a mission designer in the editor, but at runtime during the action phase of a mission. This software is being sold exclusively to army customers. There are no big secrets here, it's just a different licensing model and a few additional features.

Soon we are going to release a public version for individual training, Steel Beasts Pro, Personal Edition. Due to the personalized licensing the price is a lot lower (although still about twice as much as a normal computer game). It still is not what we envision as the end state of SB Pro, so one could misunderstand it as a premature release. True, we are going to add features in the future, but this due to the very nature of an evolving product that will never be really "complete". Think of a web browser - they undergo a constant development as well, and still they would not be considered "selling beta ware". What we have right now is stable and mature code, and we offer a much higher value than the original Steel Beasts. So, it's justified to release the software in the next weeks.

We have added many new playable vehicles - the Bradley, the Leopard 1A5 and Leopard 2A5, the FIST-V artillery observer, and more. We added many engineering elements - five different types of minefields, three of them visible, three types of point obstacles, emplacements for vehicles and bunkers for infantry, as well as mine clearing assets. All in all a tremendous boost not only in quantity and looks, but also in tactical depth. The new vehicles will require a substantial shift in tactics to be applied to operate successfully with them - the Leopard 1 is a thin skinned death trap, but with an excellent fire control system and decent firepower. The Bradley is a similar case - even less armored, with a less powerful cannon, but missiles and infantry to dismount. The Leopard 2A5 is very similar to the M1A2 in many areas, and allows a similar tactical employment - so we are in essence covering armor technology of the most recent quarter-century, from 1980 to 2005.

13) What are your future plans?

While we are going to continue working for the serious stuff, it's time to turn our attention more towards the game again. Steel Beasts 2 is supposed to deliver a more accessible tank simulation game that is less complicated in handling. I'm not talking about "dumbing down" an armor sim into the realms of arcade action. It will always be a game of tactical challenges. But I think that we should take a step back from modelling even the finest and irrational elements of the different fire control systems. I think we can merge the capacities of certain vehicle classes into one, thereby reducing complexity without abandoning much tactical fidelity, and that we can simplify the user interface in selected areas. Corresponding with these mild simplifications, we think that we should add more story elements. I don't want to go into details here, it's too early to lift the curtain. Besides, we're really in the final stages of releasing SB Pro PE, so we don't devote too much time thinking about the steps after the next two ones. We have plans, we just don't discuss them until after a successful implementation

14) Was there anything you think you had done wrong in development of steel beasts? If so, how do you plan to avoid these problems in future?

We certainly underestimated the obstacles ahead. But I don't know if that was a mistake - full awareness might have deterred us from giving it a try.

We know that more programmers would be helpful - but I don't see how we could have better dealt with it, given the lack of funds.

Maybe the original Steel Beasts - as much as our most loyal fans still love it - was a bit too complicated to pass as a good _game_. Well, we are addressing this with the split into different versions.

Sure, we found dead ends in some areas during the development of SB2 and SB Pro. A lot of our plans depended on a risky bet about the acquisition of certain terrain databases, and we lost that bet. That was a serious setback, but if it had worked out, it would have been a real hoot, so it was justified to take these risks, and I would do it again in the same situation. We didn't lose too badly either since we can still acquire other terrain data, even if they are not quite the ideal type that we almost had in our hands a year ago.

Overall we could have done much worse, and not much better.

Developer Website: http://www.esimgames.com/