Flash of Steel header image 2

Guessing Wrong

July 18th, 2010 by Troy Goodfellow · 10 Comments · Design

I spent last week working my way through Spartacus: Blood and Sand, the Starz TV series from earlier this year that is equal parts soapy fun and ridiculously overdone spectacle. The first couple of episodes are terrible, and it never really rises above “good”. But the pacing is excellent with some great “what the hell just happened?” moments.

I talked to a friend about the historical Spartacus and some of the problems he posed for Rome and that he still poses for historians. It led me to think about one of the issues that made the slave rebellion so successful so quickly – Rome underestimated the size of the problem, even though it had faced a rather large revolt in Sicily a generation earlier.

Guessing wrong is one of the great joys of strategy gaming, but one that is difficult to pull off well. In the ideal design, the player would encounter an unforeseen situation, misjudge how to deal with it, face further setbacks, but then be able to come up with a workable solution. The player will have been challenged, suffered near defeat, but learned valuable lessons. A good strategy game should always give you the impression that nothing is certain.

This is hard to do for a number of reasons. First, strategy gamers tend to embrace the Powell Doctrine – use overwhelming force to subdue an enemy. Underestimation is hard to work into a design if players use their biggest and best army to crush the tiniest speck. Second, a lot of strategy gamers have embraced the narrative of eternal forward progress. Suffering a near crippling loss can get in the way of this upward momentum story, so people reload games or quit if things start falling apart. Third, good gamers are much less likely to guess wrong consistently, especially once they’ve learned how to read the system.

Some of my favorite gaming moments have come when my own hubris has gotten in the way. The Civilization game where I was loved by all but one rival, and that tiny war I started still spiraled into a World War because I misjudged my enemy’s diplomatic strength. The Europa Universalis game where my Muscovite empire though a weakened Lithuania would be a pushover, leading to a bloody war of attrition that took me 50 years to recover from – to win the game and teach those Lithuanians a lesson. The wargame where I rushed my armor ahead to seize an objective, but was then forced to extricate them from an encircling enemy because I scouted poorly.

It is important to note that I take responsibility for those errors. That is the fourth and hardest part of designing recovery from error. If the player feels that he/she can blame the game for his/her mistakes or chalk it all up to bad luck, then there isn’t a lot of joy in undoing the damage or in rebounding from folly. Being an idiot and learning from it is a big thing, and the stories you get from those experiences are better than the inevitable march of the Aztecs, or whatever alternative there is.


10 Comments so far ↓

  • Jon Shafer

    Good article. I think THE most challenging issue to tackle is the natural desire players have to win. Everything else the developers has some measure of control over. Even if the designer does everything “right,” some people simply won’t find any sort of bump in the road palatable. I can vouch for this personally – when I was younger I played quite a bit of Panzer General, and any time I’d run into a battle where I’d suffered a bout of bad luck I’d religiously reload a save from earlier in the turn. I don’t prescribe to that particular sort of behavior any more, but for the longest time that was just how I played. So even I can’t deny having a bit of that in me.

    You can also look at movies for another good example. Most (but not all) have some sort of threat or challenge that must be overcome, and although things might get rough at points, the end result tends to be a win for the good guys. Even at this level people are conditioned to expect victory.

    As a designer you always look for that magic of a memorable experience, but ultimately you just can’t fight human psychology. This is why so many games rely on story or a campaign to provide that sense of drama along with accomplishment. Outside of that, I think the best you can hope for is to keep players engaged in other ways long enough to allow for those moments to occur on their own within the framework you’ve created. It may be possible for a game to reliably produce this, but I can’t think of a single one to get there yet.


  • Ginger Yellow

    I think EU3 is a pretty good example of this type of design, though it’s not always quite “guessing wrong”. For instance there’s the mechanic where a change in policy has a chance of creating a revolt (directly or more indirectly through lowering stability), which makes the game much more about dealing with setbacks than, say, Civ. Alternatively there’s the the genuine guessing wrong by inexperienced players – for instance, in my first game as Spain I thought I was doing great with my massive colonial income, but because I had let inflation run at around 10% for a while and not bothered to get the National Bank idea, I suffered an economic crisis which basically crippled me for a hundred years. That’s an extreme example, but the game is filled with events with “hidden” trigger conditions that can punish the player who goes too far out on a limb.

  • Zer0s

    One way of dealing with that kind of oops-reload behavior is making it so that the consequences of the actions and decisions the player makes have impact (not only inmediately but also) long after it is made.

    The Witcher was good in that regard in that some of the decisions you make affect things later in the game, so save scumming isn’t really possible because many things happened in between, thus leading to a “I’ll try the other choice in another playthrough” mindset than a reload one.

    I could see something like that working out on a EU/Paradox kind of game, but it’d be interesting to see it in a RTS or a turn-based strategy game.

  • George Geczy

    This point about player setbacks was covered in some detail by Sid Meier at this year’s GDC keynote, and he admitted the fact that his original leanings toward designs that had a “rise/setback/recover” pattern simply didn’t work well. He thought players that recovered from setbacks would get the greatest satisfaction, but instead learned that the vast majority simply hit reload after the setback and didn’t try to soldier on.

    And I remember when playing Panzer General I would be very tempted to reload after each time one of my core fighters or tanks suffered a loss by running into an unspotted unit, and I imagine most players actually did succumb and reload.

    In Supreme Ruler we tend to encounter more of what you call the Powell Doctrine – players will simply build up to get the overwhelming quantity and quality superiority so that they can assure themselves of victory, instead of trying the more nuanced approach of using superior strategy to accomplish their goals. And if it didn’t work, reload, build up some more, and try again…

    It creates challenges in strategy game design that I think do not have good solutions, due to the fact that a player’s desire for “winning” seems to outweigh their desire to try to overcome adversity…

  • Ginger Yellow

    “One way of dealing with that kind of oops-reload behavior is making it so that the consequences of the actions and decisions the player makes have impact (not only inmediately but also) long after it is made.”

    This is a very good point, and you see it quite a bit in RPGs. One consequence of which is players moaning when you can’t respec – justifiable if the game is needlessly opaque about the consequences of an action (eg Dragon Age) – but it also results in some very powerful moments. For instance the bit just before the climax of Mass Effect 2, which cunningly used the player’s desire not to suffer a setback against the player (well, OK, me). Because I was determined not to lose any companions, I dallied after some of my crew were captured and completed all the sidequests. I saved all the main characters, but I lost Yeoman Chambers as a direct result. It’s true it didn’t have any gameplay impact, though.

    You see it less often in strategy games, but it’s still there. Choosing a doctrine/tier too early in Company of Heroes, for instance, only for you to realise 5-10 minutes later your opponent has picked the perfect counter. Or pushing early down the religion branch in Civ IV, only to be beaten to founding multiple religions by the AI.

    From what little I’ve played of it,Elemental looks like it could have this dynamic, what with the semi-random tech system and the need to specialise cities and capitalise on scarce resources. You might gamble on a tech being present/available and invest a lot of time going for it, only to find out it’s not there or have it unlock much later than you planned.

  • Zer0s

    @Ginger: ME2 could have worked better in that regard if they hadn’t made it blatantly obvious that the hub structure of recruiting and sidequesting was done for once you cross that certain line. Dramatically, it might have been best if, like you it seems, had to weigh getting the crew ready by doing their sidequests versus hurrying to save the crew.

    Also ME2 is not a very good example (or innovative) because as you say there’s no gameplay impact for losing the crew. After the ME1 -> 2 transition, I’m not sure the extent of the (gameplay) consequences for ME3.

    The big problem in strategy games is that they don’t follow a (story) line like RPGs do, so you can’t script these decision crossroads. I too am looking forwards to Elemental and seeing how they deal with this kind of stuff.

  • Ginger Yellow

    Yeah, I’m not saying it’s the best example in RPGs, but it’s one that struck me recently. And yes, it wasn’t exactly subtle, but it did catch me off guard in that there were actually two points of no return when I though there was only one – if I’d finished all the sidequests before the first one, I wouldn’t have lost anyone. Conversely, if I’d not done all the sidequests but went straight from the first point to the second, I would have saved Chambers but lost some main characters.

  • JonathanStrange

    If you find yourself in a fair fight, you’ve done something wrong. Or so we’re told -that’s one reason why strategy gamers may favor “killer stacks.” If the game’s mechanics allow one to use that strategy successfully, I would find it fairly tempting to play that way.

    I don’t mind getting my behind kick if the AI’s play is better. Hell, I’d find it amusing and enlightening. I think many (most?) do too. Yeah, people reload but they also enjoy when the AI teaches them a lesson if its perceived as being fair: I didn’t scout. I neglected my defenses. My research wasn’t optimal. I failed to deploy my forces effectively. I didn’t garrison my town adequately. If the AI is thought to be winning because it’s been given bonuses or no fog-of-war or better combat odds, then I can see more people reloading.

    I believe more people would have interesting and memorable gaming experiences when they choose to play out difficult uphill struggles. It would be exciting, win or lose. That desperate Civ game where seven nations try to crush you. The GalCiv2 empire that you manage to shepherd to transcendence despite losing planet after planet. The final battle between your Imperial Guard and the invaders’ mob. I think those are more worthy of remembrance than an inevitable execution where the AI can’t avoid checkmate because the player keeps taking back his bad moves.

  • Ginger Yellow

    “The GalCiv2 empire that you manage to shepherd to transcendence despite losing planet after planet”

    Have you read Tom Francis’s (mostly) non-violent GalCiv 2 diary? It’s superb, even if he does eventually break down and resort to violence, albeit in a hilarious way.

  • Rower 41

    For me, the best strategy games create uncertainty by forcing trade-off decisions. These decisions can mimic real life, and make the game more rich. There is a reason why the “Powell Doctrine” shows up so infrequently in history; despite being an old idea (it goes back to Sun Tsu at least). The reason is often the real life trade-offs between providing “guns vs. butter” or the “cost vs. benefit.” In 1941 for example, Japan got the “cost vs. benefit” part horribly wrong by calculating that the US would not want an all-out war. Had Japan known that the US would only stop fighting with Japan’s unconditional surrender, the Japanese would probably have chosen a different course.

    A game like CivIV or EUIII are feature rich enough that these trade-offs create the most uncertainty and are the best part of gameplay for me. If your CivIV neighbor attacks you, do you pull out all stops and go for all-out war? Knowing that all-out-war halts development and research. Or like the ancient Romans, do you swallow your pride, sue for for peace, and give Brennus his 1000 pounds of gold?