Saturday, March 28, 2015

Snowballing, Map Painting, Blobbing: Solutions

Note that this is a post about "meta" game, but not metagame, mechanics. Actual in world gameplay mechanics are not really discussed.

I have thought long and hard about how to handle an certain issue in almost all types of strategy games, but probably the most annoying in TBS, Grand Strategy, and 4X strategy games. In games that simulate you ruling over some kind of nation, and especially with a long time scale, there is a tipping point when you become untouchable. Below is the method I devised to resolve this issue. Be warned, I expect both that my method will work and be elegant, and that 80%+ of strategy gamers will despise it.

The reason that games attempting to simulate or abstract wars between various organizations and nation-states cannot do so is because of the player. the player ruins everything. Each different game attempts to account for the player but they all fail. The reason for this is that they are not identifying the problem correctly, whether because of intellectual shortcomings, or because doing so causes a sort of cognitive dissonance which they avoid by trying to solve a substitute problem, subconsciously rejecting the true explanation.

There are 3 major issues with the player as well as lesser issues related to the particular genre:
1. The player has limitless time to solve their problem.

2. The player has infinite redos if they mess up.

3. The player can crowdsource solutions via the internet.

In more poetic terms, specifically related to games that supposedly follow a nation-state over centuries, the problem is restated thusly:
In comparison to a true ruler in the medieval age, because even space games tend to simply change up the setting while retaining the underlying politics/theme, the player possesses many anomalous qualities. The player is immortal, relentless in their pursuit of a single overarching goal, almost perfectly pragmatic, immune to the constraints of time, possessing personal mental energies vastly outweighing that of their many silicon adversaries, and may gain the guidance of vastly more experienced entities of the same divine nature.

To use an example, this describes every person who ever copied a strategy first formulated by DDRJake in gaining one of the Steam Achievements of Europa Universalis 4. If you think about it for a second you see that this is the distilled essence of how a player approaches that situation.

As mentioned previously every game makes some attempts to level the playing field between both the player and the real life rulers and the player and the vastly inferior simulacrums they face off with in the digital game world. Some games give the AI powerful bonuses regarding resources and game data. Perhaps they provide map vision, better leaders if the game has those, superior starting resources, other AI who will suicide the player sacrificing themselves so another AI may win. Perhaps there are +1s to dice rolls. Yet it must be possible for the player to win. And here we see why crowdsourcing of solutions is an issue. If one person can defeat the AI on the highest difficulty through personal brilliance, invariably many can do so and will publish step by step guides online so as to gain the adoration and gratitude of other internet goers.

Indeed creating a challenging game based on knowledge and information has all the issues that plague people who possess great intellect. The law you discover can be applied by nearly any living human once you have done the hard work of figuring it out. Whereas talents that are physical cannot be spread in this way. Thus the greater popularity of physically challenging activities over mentally challenging ones. A human being cannot absorb the physical muscles of another through reading some text. Any attempt to create an intellectual challenge that remains static is doomed to fail as such challenges need only be solved once and introducing small random factors only requires the aspirant to hit reload or restart a couple times to gain a favorable result.

The first step to creating a robust intellectual challenge is thus obvious. A new challenge must be generated for each aspirant. This is not a perfect solution. You can still crowdsource the solution, and if arbitrary saving is enabled you can roll back a failure. But you cannot find a solution already written and the denizens of the internet are unlikely to solve a complex pussle for you with little benefit to themselves. You still have the benefit of limitless time as well. This simulates the real life experience of a historical general or ruler, where they could apply theory gained from solving related problems but could not study the exact answer to their current problem in advance.

The next step of the process has one major answer and many minor refinements. A time cap. Yes, you may grant the player 10 minutes a turn or some other arbitrary number, but you must have a cap that will eventually become less than what is necessary to plan out and flawlessly execute each minute, miniscule, microscopic decision. Thus contrary to the current design of most games, the tipping point is not the point where you become invincible, but the point at which diminishing returns on expansion begin to set in. This simulates the real life experience of a historical ruler or general, who simply could not personally manage every aspect of their army or state. Now you may formulate a system for them to offload that responsibility, while remaining able to take control of whatever particular aspect they deem most important, while delegating less critical tasks to an inferior mind, or simply a mind better attuned to those tasks. Specialization of labor of the art of sovereignty. An example of the classic axiom of economics, comparative advantage. Even if you don't create a system that would make equal or better choices than the player, as long as it isn't terrible, the time cap provides incentive to automate some aspects of their responsibilities because they can no longer accept the tedium of micromanagement to gain the benefit of optimal actions.

The goal of the time cap is that smarter players, or simply more experienced ones, would be able to make more and better decisions within the time limit. Compare to games such as RTSes or MOBAs or shooters where players with twitch skills have the advantage. This is intended to simulate the way that superior rulers in real life quickly forged great empires which quickly collapsed after their deaths. Europa Universalis was famously unable to create a system in which players, OR the AI, could perform a conquest which mimicked these real life rulers, because taking a lot of land fast merely meant snowballing harder, there was no downswing. It is extremely difficult to make a mistake when you have infinite planning time and can wait a couple days to ask superior players what to do.

This goal combines in my personal with attempting to have players roll their own unique map, although you could save and trade scenarios, so that you can make a choice without easily being able to plot all future consequences. A smarter player will be able to handle this better, thus allowing the game to replicate the real life situation of giant empires collapsing instead of snowballing as they grow.

The final step is what many games actually figured out and applied. No saves. Ironman. But importantly, a no save game in which one or even ten bad decisions will not permanently cripple the player. Paradox games do this very well. Although truly, if your time cap and randomized scenarios are implemented well, this step may not be strictly necessary. If you have reached the limit of your on the fly cognitive capabilities, no amount of save scumming will allow your empire to expand because you will keep making sub-optimal decisions. Granted you could simply spend your time limit grabbing game data, save the game and quit, and then work from there, but even then that would provide limited assistance unless you had human readable text files describing current game state.

I expect that many players will rage at a time cap in a turn based game, but you can't please everyone.


Disclaimer:
The inspiration for this post was the game I am making for which this blog is named. It applies mainly to turn based games, although variations for RT/WP games are simple to construct, for instance a time limited pause function. It primarily focuses on games that are single player and which you cannot "win". Dominions, on a suitably large map, or a Paradox game are probably the best widely known examples. Civilization might work as well.

No comments:

Post a Comment