Parallel development of desktop and computer games: how things got upside down in different versions

There is such a guy Vlaada Khvatil. He lives in the Czech Republic, loves to play and occasionally releases some small thing. Several years ago, another “small thing” took and broke up the market, first the Czech Republic, then Europe, and then the USA. Well, our little.

He made the game "Space Truckers", the second circulation of which almost did not reach the stores - was disassembled on pre-orders. The third edition was waiting for the whole of Europe - I remember that in Prague the seller in the store told me the exact number of days before delivery. They counted there every hour and kept a notebook of pre-order accounting. Thick.
')
With this game, Vlaada added a new atom to the list of the boardtop mechanic (which happens every couple of years) - real-time mechanics for turn-based games. So, I talked with him about the development, in particular - about what rakes were when transferring his game to the application.

It turned out to be a general conversation about what makes the game interesting . For such a life.

For a start, a short explanation of what is happening. So, so:

• The two most famous projects of Vladaad are “Space truckers” and Space Alert (“Space alarm”). We are about the first game. In it, you need to collect space ships from sewage parts and fly to distant stars through pirates, asteroid fields, space debris and so on. The ship is implemented by assembling from the heap the components connected by connectors, and the flight is realized by a chain of event cards with which you have to do something using the current assembly. For example, maps ask how many engines you have for a race, generate asteroids at certain coordinates (if you have poor assembly there, pieces of a ship fall off), require a command of a sufficiently large size to rob abandoned stations. And so on.
• Here is my review of Truckers . In short - in this game, the design of rules and interface self-documentation and generally thoughtful learning curve are surprisingly cool.

Game application process:



And exactly the same, just the other - in the desktop:



The process of flight (central field in the desktop and a separate view in the application):



The first interesting question was exactly how Comrade Vladislav so specifically resisted that he took and fastened real time to the board game :

- I have always loved board games: from small funny party games to huge epic monsters. Sometimes the big "meat" games are very tight, well, you know, players need to think carefully.

And I asked myself - what will happen if we make a big difficult game, where the main part will be held in real time? At that time there were quite a lot of games of skill and reaction, but nothing complicated. And I decided to try.

From the draft of the prototype to the final has changed, in fact, not very much. Of course, there was a lot of balancing: for example, how many connections and how many component types in a game. How each event card works and so on. But I think that the first prototype that I brought to the local table club, it already worked almost as well as the final version.

Naturally, I always ask after the words about the idea how well I managed to implement it. We know that only a couple of percent of such bright thoughts reach the production. Vlaada did not lose his head and immediately explained that the most difficult was to think about the points of entry into the game.

- The most important thing in the game, that is, the part where the main decisions are made - it should take place in real time. Players need to look for a middle ground between how much they think and how well they think. And also by how much information (open maps) they see. Accordingly, I did not want them to constantly think about the rules at this moment.

It is very easy to make a mistake if you have to follow a lot of rules, and at the same time, time passes. Another problem is that if something went wrong, you cannot take a look at the rules, sort it out and continue from the same place.

I needed very simple rules. In general, it turned out like this: you can connect the connector with the connector (two types), the engines should look back, there should be no modules in front of the guns and behind the sliders. Everything. In "Space Truckers" is not very difficult to assemble the ship correctly.

But while deepening the game, I wanted to show the players the difference between the subtlest nuances of assembling their ships. And also to give variability when assembling so that real dilemmas are formed, such as: put more engines or guns? Choose dual engines with batteries or put economical? Do I want to equip the infrastructure for the aliens inside the ship - and so on.

This place is very important. Notice how Vlaada talks about learning. He was able to create such an interface for gaming components, that when you look at them, you immediately understand how and what can be done. The rules are unloaded to the state when the game is self-documented. This is an extremely important principle.

- The biggest surprise in testing happened when we started playing the prototype for the first time in a club near my house. From my point of view, he was good enough. We ran it in virtual form, and I played with friends on the screen many times. As a result, at the time of the show, I knew that it was better than most of our prototypes. But still I saw this reaction for the first time. My friends finished playing a series of games - and instead of saying “I like this game,” as is usually the case with friends, they decided not to give the prototype to other players and continue on their own. Perhaps, “I want to play more” in this form sounds much more honest than “I like”.

About something similar says Richard Garfield (the creator of MTG): players, they say, can write hundreds of edits and generally smash a prototype to pieces, but if they continue to play - this is it. The rest is all garbage, the main thing now is to take away their prototype. The strength of the resistance determines the commercial success of the game. Judging by the sales in Europe, the testers should have taken its prototype home.

But let's continue with Vlaada:

- In the Galaxy Trucker iOS application, the main thing was to make a very good learning process that would not be boring for players who already knew everything about the desktop version. But at the same time, step by step, the lead players who knew nothing about the mechanics and principles of the game.

For the first departures, we left only the basic things. Advanced features of mechanics (such as infrastructure for aliens, their own team, viewing flight navigation and the like) come later. Players open up these opportunities during the single-player campaign, meet new challenges and challenges, and receive new game features as rewards (for example, cooperation of other alien races, ships ’class II and III schemes).

The method itself is not new: we regularly do the same in our games when learning. The principle of writing the rules: first, everything is basic, so that the player can quickly get used to it, then features for the experienced ones are wound up in additional sections. What is worth noting separately - Vlaada went further. There in the game are three piles of cards for different flights, respectively, I, II and III. So, he noted in the stack I eight of the simplest maps covering almost all typical situations in their purest form. And he made of them the training flight described in the rules. And, in fact, he scripted the first game, telling what to do and how.


^{Each step of the game reduces cognitive dissonance. Well, almost everyone.}

The learning curve in games is damn, just terribly important, and from the Galaxy Trucker you can learn how to work with it. Continue to learn from the experience:

- Why did we add the plot and make the campaign? When I play single player games on my phone or tablet, I see how polished they are: a lot of places to explore, many new unlocking features, a constant sense of progress, and constant rewards for it.

So why not do so with the board game when transferring? That's why we needed a campaign. And while I was doing it and prescribing interaction scripts, I remembered my programming years - it was very nice.

Now is the moment. What does the computer allow to do, which the “desktop” engine cannot? He can do a lot of automated actions for the player. In the simplest case, follow the rules. In the most complicated Neuroshim , for example, the computer itself parses the stack of moves, which eliminates any ambiguities and significantly speeds up the game. And here you can monitor compliance with the rules, removing part of the cognitive load from the player.


<sup>Bar "Space Bar":
- How did you get into this barrr?
- Clicked here and here.
- Ha! Clicked. I rrrsskazhu you, as was earlier, son ...</sup>

- Naturally, knowing the board game, it is much easier to start playing computer games. And the computer version is also good because it doesn’t tell you what is right and wrong - given the ability to control the rules at the interface level, you simply cannot collect something wrong. In the desktop version, it is important to assemble the ships correctly and assemble them well, but here it is enough to assemble ships that are functional for flying through a certain section.

This, by the way, gave his computer game models completely unusual etudes. The fact is that on the table a set of maps of events occurring during the flight is fixed, and, in general, obeys the laws of statistics. And in the application it can be easily changed. And here he had a very special areas - only asteroid attacks, a breakthrough through the pirate fleet, a large zone of hostilities and so on.

In the Nastolka, the whole "render" is made by the players themselves. And it is their great load. In computer games, you do not have an “estimated power of human systems” capable of producing at the maximum a conscious flow of about 10–20 bits per second at the output. On the table you are fighting for every quantum of attention. In computer games, most of the implementation can be hung on the processor, leaving the player only solutions.

This makes it possible to easily twist features on the game, if they do not add complexity to the implementation, more precisely, do not require additional training of players and do not occupy a place in their "RAM".

So there were unusual types of goods (such as "pig iron" bar, nuclear waste, etc.). So there were unusual companions (other players on the flight with you) - one must be accompanied and guarded, the second one must be overtaken, the third one must be substituted for attacks. Similarly, there were unusual ship formats that are not in the desktop version. Surely a lot of things were ruthlessly cut off at the optimization stage of the game under a bottleneck - the computational capabilities and memory of human systems.

Much was carefully taken out of notebooks for software implementation - and generated excellent events in the single-user campaign.

In multiplayer, by the way, most of these features are not.

Now pay attention to how he spins the learning curve:

... We wanted the players to get really many great impressions from the game. I played a big pile of board game ports, and was somewhat confused. Oh, the new game, I know its desktop counterpart. I set the highest level of AI opponents, I start ... and win the first game. And now what? No more than achivka. Until I start playing multiplayer, everything else in the game just looks like a waste of time.

Again, he echoes the ideas of Richard Garfield. When the MTG (chain of fights) was entered into a computer game, one of the implementations suggested a global map, which you had to follow as in King's Bounty. The mechanics of the game was in battles, but the global story added a lot to replay. I wanted to go. I wanted to collect more cards.

Accordingly, any game on duels can be beaten in a similar way. Even chess: you can add a campaign there that will add pieces to you, expand their capabilities in accordance with historical realities (unlocking the elephant features on the entire diagonal, and not 3 cells, as it was before) and, of course, throw up more and more difficult etudes and opponents. Actually, I wrote quite a bit about it, talking about plot-oriented games .

And now about the coup. When porting there is another significant issue format - the enemy. The first is AI. Iron brain is clearly able to do all actions in real time almost instantly. Understandably, he can register delays, but still hardcore will not work - in the field of speed we can compete with him and compete. But you can compete in creative areas. For example - as an assembly. This leads to the idea of ​​some mode where you can fight along with AI.

On the other hand, the negative (sometimes) difference between the computer version and the desktop version is that you do not have a live opponent. More precisely, it still needs to be found in multiplayer. Vlaada used the good old desktop game system of correspondence and allowed to run several parallel games with players around the world at once. To solve these two problems, a paradoxical solution was needed - to remove the basic real-time mechanics from the game and put a turn-based mode in its place. And it turned out to be even more interesting for some - so everything turned upside down:

- The biggest change in the game - in addition to the real-time mechanics, we added another turn-based mode. This was necessary so that players could play “by correspondence,” that is, in asynchronous multiplayer.

It is worth adding that the guys clearly wrote AI, using several evaluation functions (we did something similar for Shakal ), so they got the opportunity to do different "characters" by simply changing the weights of each of the subsystems of the ensemble. This is how the “military robot”, “trading cargo robot”, “balanced” and “paranoid” turned out.

Summary

Vlaada made a strange board game using computer game mechanics and won. Then he took and made a computer game using classical desktop mechanics, but at the same time, boldly using the additional features of the format. And again it turned out interesting.

Vlaada before nastolk worked in video games. Now he is known as a cool game designer from Europe. Securing success, he again goes to the video game market, but already with his ports nastolk.

Sly dude.