Strategic Uncertainty – Keeping Strategy Games Fresh – Extra Credits
- Articles, Blog

Strategic Uncertainty – Keeping Strategy Games Fresh – Extra Credits

Hey we’re back! I hope you all had a good holiday. Now, who is ready to talk game design? We’ve all had those moments in civilization or total war where we’ve won. We know we’ve won There is absolutely nothing the computer can do to come back But it’s going to take another four hours of repetitive slogging before the computer realizes that we have won. Before we get to the win screen and the game acknowledges our victory. Even in great strategy games like those this is often the weakest part of the experience. So today let’s dive in to why it happens and And why it’s so hard to design around The problem of having to sit around and go through the motions of finishing a strategy game that any reasonabe human being would clearly say you already won stems from one very important, but very difficult to manage of strategic game design Uncertainty Uncertainty in a strategy game is that element that makes readjust your plan Makes you reasses the situation and consider all the tools you’ve built up to that point in a new light Without uncertainty a strategy game simply becomes a puzzle And there is nothing wrong with a puzzle, in fact, they work some of the same mental muscles But the key difference is that once you know how to solve a puzzle all that’s left is the execution and that’s not what we’re looking for in a strategy game. In a strategy game we are interested in the planning, and coming up with strategies and testing them in all sort of scenerios And the only reason to replan, to continue to strategize once you’ve got a strategy that’s working Its because the enviroment that your strategy working in might change on you Uncertainty forces you to plan for contingencies and to rethink your strategy because that’s what we play strategy games for Now it’s important to not confuse ‘uncertainty’ with ‘randomness’ here Uncertainty comes in lots of forms and randomness can be one of them but more often uncertainty comes in the form of hidden information, Unexpected events or even simply playing against a player who might make some unpredictable move that will force you to reconsider your strategy. Taking civilisation as an example, you’ll find that the early phase of the game is filled with uncertainty. The map is almost blank you don’t know who your opponents are, where they are or even how big of a landmass you’re on. Then as you start to explore your plan shifts as you discover What resources are within reach as you run into other civilisations and have to work around their borders or even as you discover goodies in those friendly villages or the less friendly barbarian camps but you’ll notice that this uncertainty slowly declines over the course of the game bottoming out once most of the early wonder building is done and you’ve figured out which victory conditions most of the computer players seem to be going for. Now this lack of uncertainty is compounded by the fact that in civilisation, like in most single player strategy games advantages are cumulative. In most single player strategy games if you;re doing well you’ll have access to more units more resources more abilities or powers to help you deal with whatever gets thrown at you. This creates two problems. First at the point where many strategy video games devolve into being puzzles because you no longer have to adjust your strategy to win. You’ve actually already solved the puzzle there’s nothing engaging or challenging left to strategize around while finishing the game because you’re already so far ahead but second and perhaps more interesting is the fact that this makes uncertainty much harder for the designer to build in as for the player who is already way ahead a minor uncertainty is practically irrelevant it essentially doesn’t function as an uncertainty anymore because the player is winning so Hard that they don’t have to change their plan at all to deal with it, if a game suddenly hit you with an unexpected Penalty that results in you producing two less food per turn well if you’re already producing a surplus of 50 food every turn You don’t really have to change what you’re doing, do you? “Oh no my two food what am I gonna- ah oh no…” And like if some opponent decides to attack you with their flintlock rifle squads, but you’ve got laser tanks You’re just going to sweep them aside No problem their attack may have technically been unpredictable, but it sure didn’t give you any reason to rethink your strategy and this is a problem because if minor uncertainties are no longer effective that brings up the question of major uncertainties and major uncertainties are difficult to design because as much as we’ve talked about uncertainty being a good thing in some ways uncertainty is antithetical to strategic play creating a strategy for anything involves making predictions about The future about what results your actions will have the more information you have the more effective a strategy you can create or to put it another way the less uncertainty there is the more Strategic you can be so how do we balance this how do we balance creating strategic play while giving the player enough uncertainty that they need to keep making new strategies to deal with or anticipate the unknown Well for one thing we can build our AI with this in mind If you’ve ever sat around playing strategy tabletop games with your friends you’ll know that often at some point Somebody ends up doing something you completely didn’t expect and it may not have even been a great play on their part But it forces you to craft a new strategy So when it comes to AI especially in games with multiple AI controlled variables allowing at least one of the AI actors to play suboptimally, but in interesting ways can help. Better still is hidden information Not only because it forces the players to build adaptable strategies that they can rework as more information becomes revealed But also because it lets us build AI models that can craftily fake you out on their goal giving you those moments Where you think you’ve got the game in the bag but then they slam down three roads clinch the longest road objectives and leave you scrambling to get back ahead. and then there’s one-off events. These can be a ton of fun but they’re the most dangerous option too as it’s not super fun to have a working strategy suddenly run into a random event that basically reads Oops you lose too bad. And unfortunatly it’s really easy for unpredictable one-off events to lead to that result many strategy games encourage players to essentially run lean to consume resources As fast as they can acquire them and to acquire the bare minimum in all the categories that aren’t pivotal to whatever strategy They’re trying to execute so they can spend all of their energy getting the pieces they want most But that’s sort of delicate balancing act can be completely destroyed By even minor Random events that affect them, now if the player has the ability to react to rethink their strategy To come up with new strategies to compensate for the event and barge ahead with a minor setback This can actually be a good thing but too often it just means that all of their work is undone And they have to reset or more likely just quit the game here one of the best techniques I’ve found it can make events positive to have unexpected random things give bonuses instead of penalties But if you do want to put in negative events, just be sure to allow for a lot of flexibility. For example you can make it so that the event effects players differently based on their ranking like say a Volcano blows up the player in first place has to provide rescue crews massive amounts of aid has to divert ships for refugees and such while the player in last place just has to provide a small amount of food to the effort and Optimally they’ll both be rewarded in some way if they successfully deal with the crisis You can also provide players several different approaches to dealing with your events. Let’s say the earth is dying event occurs You can approach this by racing to colonize other planets, or you could use science to find a way to essentially terraform your dying Earth You could use your industrial might to burrow below the surface Or rework your economy to live more in harmony with nature and let the Earth recover this way all the civilizations That didn’t focus exclusively on space flight aren’t simply doomed when that random event happens. Now there are a lot of other ways to inject uncertainty into your strategy game But hopefully this has spurred some ideas Mainly though whenever you’re looking at a strategy game if you find it getting staid and stale, if you ever find yourself just going through the motions long after you already know you’ve won Dig in, and see if the game makes use of any uncertainties in the late game, just don’t judge the designers too harshly it’s a really tough task to balance a well honed system which you can plan and strategize around while having uncertainties that make that system always feel fresh. No one said design was easy. See you next week. ♪ Outro Music: “Wisdom (Zelda’s Lullaby Link to the Future)”♪

About Ralph Robinson

Read All Posts By Ralph Robinson

100 thoughts on “Strategic Uncertainty – Keeping Strategy Games Fresh – Extra Credits

  1. Too many strategy games are over hours before you actually win the game. How do we fix that?

    PLUS! Vote for Extra Credits in the Shorty Awards:

  2. What if you add a negative initial event but depending on your reaction it can go form bad to great or the polar if your a twat

  3. Supreme Commander handled this well: because the game was centered around your (relatively weak) commander unit, a desperate player could probe the enemy's base for blind spots and target the commander directly, allowing them to win by subterfuge rather than by brute force. You could easily see a player controlling 3/4 of the map lose because he had lax radar coverage, or failed to build a counter to air raids directly over his commander.

  4. My best Civ games, or even strategy games in general, have always been when I have a rival. Not just other AI players, but one that is in as good a position as I am at a pivotal moment. I think strategy games should try to "let" one other AI dominate those around them so that when you hit endgame, you still have an opponent.

  5. I'm dabbling with the concept of adding in high-risk high-cost options that an AI wouldn't normally go for… the STEROID option, if you will. This would buff their dudes out (I keep visualizing AOE/AOK for this) but would cost an arm and a leg, but it costs MORE the more units you have, so players already winning would be less likely to take it because it's too costly. It would allow for an unscripted, nigh random, difficult challenge that the AI can employ and if it doesn't work, the AI is left so destitute that what would take 4 hours for the AI to realize he's lost, takes 4 minutes instead.

  6. The CK2 version: When you own five kingdoms and no empire.
    Then everyone joing the "gavelkind for REALM_NAME" faction.
    You get an ultimatum.

  7. This was all fixed in Genghis Khan 2 on the SNES. It's been 25 years and they still don't use the random Governors that turn against you, like they should in games like Civilizations.

  8. try occupy less than half of the world map, and purposely help one AI take the rest of map and come to you, some great fight just might keep u interest little while

  9. Stupid population revolting due to economic scares and culture wars that will never happen in game becuase you lose some territory but it is strictly possible, means you can focus on that section of the game which is now unstable instead of what you have already built up. Vice versa

  10. In civ V i was at war with my neighbor whos army was around three times the size of mine, but i had a workforce that was rediculously powerful (i shit you not i could make a nuke in 1 turn was unmodded. I turned my capital into a fortress(however they had the great wall of china and it was filled with several generals and more troops then I could ever get through) I flooded the border with defensive buildings(cant remember the name) and built as many troops as I could. I didnt use my nukes because we both knew i couldnt reach his capital, but he could reach mine. This war ended around sixty turns later (or at least what felt like it) when we signed a treaty.(if the war went on for even five more turns i would have likely lost. My front lines lasted the first three quarters of the war and after it broke things steadily went downhill for me. This was just a fun story and it shows that what for them should have been an easy war could have turned the other way if i just didnt make a few mistakes

  11. I guess this is why the board game Risk is kinda genius. Almost every turn is uncertain, yet there is still a surprisingly good amount to think about.

  12. WAIT WAIT WAT: Most Civ players try to calculate what victory conditions the AI are going for? THE AI CHOOSE VICTORY CONDITIONS?!?!
    I'm such a noob. (I've played civ 4 for years and never knew that DX)

  13. I thought you were British for the longest times until I heard your real voice…….. specifically southern British accent

  14. This episode made me think of one of my favourite tactical uncertunties in video gaming. I like to call it the "multiple fronts factor". I play a lot of total war games and this is mostly where it has applied to me. As you get a better economy and military, you will make more enemies. This means you will end up fighting two enemies on opposite sides of your kingdoms many times. The better you do, the more people you piss off, the more you need to pay to support several armies to fight on more fronts and the fewer people want to trade with you. What's more, you can't always predict how each fight will go. If you are winning every war you are a part of, but lose your best army to an your arch nemesis who is now marching on your capital, suddenly you need to pull troops from another front to combat this enemy, jeopardizing your other wars as they too begin to want for manpower.

  15. Nothing the AI can do? Then why does that FUCKER GHANDI always manage to pull a nuclear sub out of his ass at the last minute?

  16. I say the solution to this problem is throwing something at the player that they can't anticipate or can but can do little about, something that doesn't necessarily need strategizing but to just adapt and out live; throw them a curve ball problem that will delay them so that the other powers of the world in Civilization per se have time to catch up and thus they become a prominent problem again and strategy becomes essential again.
    Bringing that flintlock rifle Division versus laser tank situation up again, now the laser tanks have to deal with say an alien abductions that take some units or population from cities, or an earthquake that wrecks their cities so some or many city works or districts need repared, or a plague that ravages their cities and or armies rendering their units doing half damage at half health and their cities at half production or you can put different levels of severity on the disease with varying levels of severity in regards to Effectiveness with their troops and cities; even a meteorite that wipes one of their cities off the board would be good.
    And I think the new DLC might a dress this problem with the environmental disasters we saw, or I could be wrong.

  17. the issues you were talking about in here come mostly from 4X games and not for RTS games, I know the genre is seemingly dead but the trueth is even more sad then the reality, the genre is being replaced by 4X in the excuse of "looking for more depth" and "players can't absorb so much information on the go", the dedicated fans of the genre are left with Indie games that don't have the same budget and influence of AAA productions, Starcraft 2, old games and going to 4X…
    I liked Total War in the first 40 hours of play, but the gameplay felt off to me. I went back to C&C games for a while there till I got my hands on Supreme Commander – a game that have Awesome dripping out even from it's scenario, though some part of it(most of it) is unfair, the game have 1 objective – kill the enemy commander, everything else is meaningless.
    that was probably the best strategy game I've ever played, you could be losing left and right but because you decided to build some gunships and won a single air-fight suddenly the tides were changed and you rushed the gunships on the enemy commander killing him before his overwhelming horde of tanks crushing your base like it's nothing.

  18. After conquering the entire world as persia in Empire total war, i decided i wasnt quite satisfied.

    So, i disbanded most of my militaries and war fleets. Then, i jacked my tax rates up to max and let the rebellions come. I expect the rebels to play. At least a little bit. I wanted to conquer the world all over again with my fractured empire to see how it would play differently.

    It didnt go great lol. I mean, America became a thing and the biggest problem. But after 15 turns of letting it all burn. i was left…. underwhelmed. A few elite janissary armies could sweep the map in another 20 turns and be done with it.

    I had hoped for a ton of new minor nations to rise up and become a problem

  19. I like the uncertainty in Mount & Blade: Warband. In that game, it's really hard to get into that state of having the whole game in the bag. In fact, it's never happened to me, even after successfully uniting the land under my banner once. It always felt like a massive endeavor, a struggle that took many many in-game years, and months of real world time. A whole mess of things can happen at any point. A good example is that some time after the early game large bandit parties spawn, seemingly usually in the territory of the richest faction, which can really hinder the progress of that faction. Others include how some of your lords may defect from your faction. The larger and more powerful you become, the more difficult it is to maintain everything. And your skills with diplomacy and understanding and adapting to circumstances really get tested.

  20. Stellaris does something interesting with this concept with the end-game crises. The end-game crises are essentially galactic-scale catastrophes that you and your allies will have to fight off. The advantage here is that, if all goes as planned, these crises are always serious efforts, and run a very real risk of ending your game. Though I wish that the End of the Cycle could occur in AI players (would be an interesting challenge).

  21. I actually play two RTS games: Homeword (my favourite) and They Are Billions.
    Other strategy game I like are War Thunder, World of Tanks and Battestations Midway.

  22. A colonial expedision if they find the land good ther Will come more when they leave but if they find the land hostiele theyll leave and never come back and the colonist and expedisions has far supiriur tecnolegy than anybody in the game

  23. The obvious solution is to include inherent elements of chance that are almost impossible to calculate but also fair. Look at the game Diplomacy for a good example.

  24. That's what is great about Stellaris, or maybe just Paradox games in general because anything could happen. I could have the most massive empire in space, and then some space fleets from somewhere else come along and I juuuust about beat them, after having the entire empire torn in half (and that is why I have never activated a single Gateway since)

    Even right now I know I'm going to win, but I'm worried that something may happen that ruins everything, and I'm trying to formulate a strategy against what I believe will happen, but something entirely different could happen. My strategy could be wasted. On multiple occasions, I thought I could go on forever with the resources I'm collecting, but several minor events happen that add up and all of a sudden I'm trying not to fall in the red. It's fantastic.

  25. Do you need to give a major umcertainty to the player,why he's doing too well?
    Nuke him(not one nuke,but LOTS of them,targeting his major cities,and a nuke attack sent not by the players,but as a random event,but with the sidenote that will only happen to players that are going way too well,and are much superior than his enemies.).that will be a major problem to him:he'll need to rebuild his military,economy,his cities,contain the unrest in the population,re-populate,clean the radioactive areas,and do all that stuff while he defend his country from all the enemies that will take this opportunity and attack him.

  26. one of the best ive seen so far, is in barbarian invasion total war, right of the bat your warned the hordes are coming, so your always mentally bracing yourself for the hordes.

  27. I don’t know about you but the last time I played cig I Conquered a continent and basically waited until the game ran out of turns

  28. Paradox does this wonderfully. Fog of war, rebellion,and all sort of random events keep you on your toes. Also there is no winning

  29. So space fakers are just gunna sit around lazily as they figure someone else is already on fixing their bail attempt? Banish em for being self serving elitists

  30. Loyalty is the fix that civilization used in the rise and fall expansion. It makes you lose cities depending on at least 6 separate factors that you have to balance, including in a rough sense, how many cities you have, meaning that if you are doing better, you are more likely to have cities desert you.

  31. How dare a video on this topic not mention a single Paradox game. Firaxis and CA are good, but EU4, CK2 and HoI4 run rings around both of them when it comes to strategic uncertainty and developing AI that tries to compete. They're not perfect, but there are very few developers who are able to deliver strategy games on the same level as the big three by Paradox.

  32. 1:57
    Oh no the fishingbouts won because the destroyer couldn't see them
    Why did I say that
    I thought it was kinda funny though

  33. But there ist also the oposite.

    You played for hours and days and than finally you are winning abd finnaly want to destroy the enemy and the Ai Just steals the fun of Victory and surrenders…

  34. This spoke to me hard just within the first minute. Here I am, playing Medieval 2: Total War. Playing England. Only having lost a couple battles mostly early on when getting "sucker punched". I have the largest empire, the strongest military by a LOT, and the richest faction. Europe belongs to me completely except Rome because Pope (really irritating seeing that little bit of white tainting my beautiful red peninsula) and a couple of places that calling them "Europe" is questionable. Because of me controlling most of the western coast, aside Northern Africa, I was easily able to be the first to the Americas and I control all those trade goods there. Literally all I need to do is take Jerusalem, which I would have done sooner if the AI wasn't "stupidly" going to war with me constantly in wars and fights they have 0 chance of winning. It's only a matter of "when", not "if".

    I say "stupid" because the AI knows my win condition and can't just sit there doing nothing, and they ARE successful in delaying me if nothing else. It's just stupid from the PoV of real life or logic without using meta knowledge

  35. And this is why I love the gameplay of 40k. Never know when it’s going to change on you and need to mold your strategy turn by turn to beat the enemy player.

  36. Another thing that can help are predictable effects that occur as you progress and make the game progressively harder.
    Than can be in form of corruption. As your empire get's bigger it get's harder to manage and corruption spreads.
    Another possibility is for the other characters to team up against the player when playing aggressively.

    Besides simply making the game harder it is also possible to introduce new mechanics that might even be accessible to everyone but makes old strategies useless, e.g. once nuclear weapons are developed your strategy of expansion by conquering could get ineffective, because every attempt would result in mutual extinction which makes diplomacy, spying and sabotage relatively more effective.

  37. I don't quite like how Total War: Warhammer 2 handles the Vortex events, for my play the uncertainty of the amount, size and place of the spawned chaos armies is just a little too big. I might know, where those armies will be heading, but because of the way economy and "garrisons" work in the game, those armies would be able to just destroy any city without a commander easily. For me it's not fun to re-build an entire part of my empire, just because a randomly placed enemy took it from me. I can totally accept AI-Factions to betray me (it's total war after all) or having to deal with a random army of Rebels because their existence is my fault in the first place and i can at least forsee it coming and send out troops in the general direction. I miss being able to send single units around between cities and sometimes this one extra unit gives you the edge in an otherwise difficult fight, loved those moment in Medieval 2

  38. Maybe I'm just mean or something, but I actually like to eradicate the enemy at the end. Of course it wouldn't be fun if the game only consists of this easy puzzle solving, but once I've gone through the first part, I normally like to destroy all the buildings and units of the enemy…

  39. The real uncertainty: Ghandi has a anger and use of nukes of 13, and if using mods, he has a space empire.

  40. 4:10 I think in an rts like starcraft or warcraft 3 u should get a widenning area of play and a new enemy once you expand ,, like Overlord being a bboss and people going Ainz ual goan you are a threat to my nation

  41. One option is to build (nearly) deterministic risks into strategic moves. That is, make problems that get worse the better you are, but don't stop you from continuing your goals until they reach the breaking point. The history of real civilization comes to mind. Rising empires always fell and often fell hard due to the contradictions in their system. Rome fell because it relied on slaves from conquered lands and ran out of them. European empires fell because of the increasingly modest gap between master and subject and the contradiction between developing your infrastructure and extracting from your colonies and keeping them from revolting. Russia fell because its class structure, which enormously benefited the elites, triggered the Bolshevik revolution. The Soviets fell because by that point they had abandoned many goals of that revolution.

    I'm not saying you should build a system based on class analysis and dialectical and historical materialism into games like Civ and play as a representative state of some particular class interest, this would be far, far too hard. Nobody who joined at the start of the game would still be there by the end, and players would be stuck in the pre-classical era. But at least paying lip service to these kinds of ideas can help make a seemingly steamrolled game into a loss, in ways that the player can absolutely learn from and refine their strategy against.

  42. Maybe I'm weird but I like the part of turn based strategy games where I've already secured victory but the Win game state hasn't been achieved.

    When playing against AI opponents you're able to play around with powerful end game mechanics that would otherwise be to risky to go for if the game was close or play for build towards complex synergies that would be either too fragile if your opponent was in any reasonable position to break them up or to slow to achieve victory if your opponent was in any reasonable position to find another path towards their own victory.

    Against human opponents there is this immersive feeling of forcing your opponent to submit plus the tactical advantage of causing your opponent to waste their mental energy as their decisions become harder and yours become easier which can cause them to play suboptimally in subsequent rounds/games/ matches.

    I'm guessing that most players who gravitate towards playing control decks containing the bare minimum number of wincon cards (typically associated with Esper, Bant, Grixis, Azorious, & Dimir control decks) like this part of the game.

  43. It feels like you're trying to hard to make this part of the game interesting to the players who normally don't find this part of the game interesting. Why not just go the simpler path of just introduce alternate win conditions that are capable of quickly closing out games but require the player to already have a meaning advantage over their opponents in order to gain access to them and to maintain that advantage to complete them? That way these alternate win conditions won't become unbalanced relative to the slower primary win conditions while offering a means to end games that dragging out.

  44. My Civilisations V example: I was going to go to war with Warsaw, when suddenly, a lot of other city-states declared war.

  45. 5:50 Such as Age of Wonders III's "half casting points per turn for 20 turns" objective.

    When I have half my cities protected by summoned creatures, then just lose half of my mana income completely, causing all those cities to become undefended targets for evil empires to immediately take over and migrate… that stings.

  46. This happened to me while playing civ 6

    A dif nation went for a object I've, and I did something so they stopped and that went on for Abit.
    Other civ: i hate you.

  47. Actually, I used to love this stage of the game in singleplayer: The battlefield turns into a sandbox and you can experiment with the units. Warcraft 3 and Red Alert 2 really rewarded this, with Red Alert 2 actually allowing you to create entirely unique units sometimes.

  48. A small idea i recently had for games with technical progression, strongly inspired by paradox games

    One of the major issues I find with games like civ is that the snowballing effect of science accumulation can effectively make you better at everything than everyone else; hence min maxing for research tends to be quite effective. As an idea, if you're racing ahead in technology there might be political crisis if things are unbalanced, ex the transition from tribalism from feudalism (ex Civ6 classical era), feudalism to more concentrated monarchy (classical to renaissance), monarchy to modern era (renaissance to modern era), technological disruption (modern onward). Thus being the one country in the modern era in civ might mean you have do deal with uprisings before you can go ahead and wipe everyone else out.

Leave a Reply

Your email address will not be published. Required fields are marked *