Soren Johnson's Sins of Strategy Games

[Chris Nahr]

Soren Johnson (http://www.designer-notes.com/?p=106) on "seven deadly sins" of strategy games. Normally I'd start quoting excerpts here but nearly everything he says is true and right and beautiful, so you should just read the whole thing. Especially if you're making a strategy game.

[Hanzii]

You're right. I agreed with every seingle point allthough I must admit that I have a love of games that has Too Much Stuff.
I know the road from complex to confusing is very short indeed, but I love having lots of units and lots of buildings.

[Kareem]

I agree with most of Soren's points, but being a fan of single player campaigns I have to disagree on the first and last points. If the story mode (with all its requisite scripting) exists as a separate part of the game and doesn't actually encroach on skirmish or multiplayer modes, I don't see what the problem is.

Is it that it takes up development time that can be spent on developing a better overall game instead? Well, alright I can understand that, but I also think that a lot of people play these games for the campaign mode as well, at least judging from the polls we've done here before, so there's definitely some kind of trade-off/opportunity cost.

I guess it makes me a strategy philistine, but I don't spend that much time playing skirmish against the AI, and I don't think I'm good enough to play online. So for me, most of my strategy gaming occurs within campaigns. I haven't played Sins of a Solar Empire yet because it doesn't have a story mode, which is why I'm looking forward to the expansion and its campaigns. I haven't touched any of the Dawn of War expansions because the main game's story was very disappointing to me, since it was simply a series of skirmishes disguised as a campaign. Warcraft 3 is my most played RTS (in campaign and skirmish modes) because of how superb the story mode is.

[jellyfish]

Good article.

[Gordon Cameron]

Warcraft 3 is my most played RTS (in campaign and skirmish modes) because of how superb the story mode is.

I didn't like Warcraft 3's gameplay much at all, but the extent to which I played it (first 2 campaigns) was because of the story. Blizzard is the exception for me on that, however. Apart from Warcraft 3 and Starcraft, every RTS story campaign has invariably left me totally cold.

[lesslucid]

I don't seem to be able to access the article... is it just me?

[Kareem]

I don't seem to be able to access the article... is it just me?

Works fine here.

I didn't like Warcraft 3's gameplay much at all, but the extent to which I played it (first 2 campaigns) was because of the story. Blizzard is the exception for me on that, however. Apart from Warcraft 3 and Starcraft, every RTS story campaign has invariably left me totally cold.

Not only did I replay the full campaign several times, it also made me pick up WoW (which was the only MMO I ever played) and it influenced my race choice in it. But I disagree about other RTS story campaigns. I had fun with the over-the-top cheesiness of most of the C&C series, and I loved Age of Mythology's campaign as well. World in Conflict had a great story mode, and so did Company of Heroes. There are a lot with great story modes, but none were as compelling to me as Warcraft, with all its cliches.

[Tyjenks]

I love reading Soren's stuff. Always he has excellent insights and this one is no different.

[Telefrog]

I agree with most of Soren's points, but being a fan of single player campaigns I have to disagree on the first and last points. If the story mode (with all its requisite scripting) exists as a separate part of the game and doesn't actually encroach on skirmish or multiplayer modes, I don't see what the problem is.


I have to agree here. I've rather liked the single player campaign stories in the Age of Empires, WarCraft, and C&C games. I frankly wouldn't have bothered playing the C&C games at all if it wasn't for the cheesy story. (I've never been all that impressed with the tank rush gameplay in C&C multiplayer.) Company of Heroes is all about the campaign for me since I suck at multiplayer in that game.

I got Sins of a Solar Empire despite the lack of a campaign mode. I think it was an exception for me because the pace of the skirmish games feels like a long campaign mission anyway. I'm a builder/turtler by nature, so a nice long skirmish game (a couple of hours) is just as satisfying to me as a quick multiplayer rush is to a typical RTS gamer. I'm making up my own narative as I go along.

This is why I get so much satisfaction out of the 4X style games, again despite the lack of a story mode. The games are long enough, and involved enough, that I can just write my own epic plot.

[Aeon221]

He's not really harping on about stories in games, at least so far as I understand it. His beef seems to be with bad stories in games, hence the reference to Rise of Legends, which had a story so effing terribad that it sapped my will to live.

For most games, the campaign is this tacked on series of lame cliches and uninspiring voice acting that could have been left out without any complaints. RoL would have been far superior without that angsty Mary Sue of a main character bitching and moaning after every goddamn mission. RoN's Conquer the World mode had all sorts of nifty scenarios -- cold war style infiltration missions, skirmishes, ally protection, yadda yadda -- that really made the game. The most depressing part of RoL was that it had its own really cool scenarios, but they were locked inside the campaign. La-a-a-a-a-a-a-a-ame.

To pick another example, look at AoE3. The story was fairly forgettable, but most of the missions were incredibly creative scenarios that I greatly enjoyed. But unlike its predecessors (and its expansions) they were locked inside a worthless series of cutscenes that prevented me from picking and choosing scenarios at my leisure.

If developers really truly absolutely want that feeling of progression, why not go with a Majesty style campaign, where you'd start with some missions and get more unlocked as you started winning. You get all the sense of power growth without the annoyance of a rigidly linear campaign.

[Spasticon]

If developers really truly absolutely want that feeling of progression, why not go with a Majesty style campaign, where you'd start with some missions and get more unlocked as you started winning. You get all the sense of power growth without the annoyance of a rigidly linear campaign.

Usually this just ends up feeling like disguised linear progression to me though.

What's worse with the unlocking though is when you're given several to pick from, which then lock out the others in order to keep the story moving forward. I'm never given enough information to make what I feel is a good decision, so the end result is the only way to tell if you did it right is to save & reload many times ... which completely ruins the effect for me.

[scharmers]

Brad could have disposed of all of the campaign scenarios in all of the GalCiv II games and worked on something else instead. Does anyone play those things?

[SwampIrish]

Players must be able to mentally track their in-game options at one time, and putting too many choices on the table makes it impossible to understand the possibility space.

He's talking to you Gas Powered Games.

Edit: at least your strategy games anyway. I'm sure he's just as flummoxed as the rest of us when it comes to your RPGs.

[Quitch]

I did, but mainly because I liked the interesting setups rather than the story.

[wisefool]

Remove one of those Sins to add interface woes, please!

You should be able to right click everything to get reference materials at the same time you can see the board (on a sidebar / transparent overlay).

It's not fun to need a second monitor or print outs to play.

[MattKeil]

His points in #7 about Rise of Legends really struck home for me. I was immensely disappointed that they hadn't expanded and enhanced the Conquer The World mode from RoN and instead had gone for a fairly uninteresting story campaign that had light elements of CtW. I'm sure the reasoning behind it was something along the lines of feeling they had to explain and provide context for the strange fantasy world in the game, but Sins of a Solar Empire may have proven that you don't really need that if the game's fun to play.

[Quitch]

1. Too much scripting

And without scripting, it's skirmish, and people on this board complain endless about a campaign which is naught but skirmish.

I'd like to see more creative setups with the AI to handle it (and co-op campaigns are possibly the first change in over a decade), but developer AIs tend to suck a fat one. Hence scripting.

2. Too much stuff

I disagree, the stuff is there to make the game appear more interesting than it is. Take Total Annihilation, a game a noob can love because it seems you have a million possibilities and different ways to play. You don't, there is a very small set of core units and once you realise this I suspect most people lose interest because it stops being about discovering new ways to play and simply refining the existing ones.

3. Limited play variety

Dear God yes. Wasn't the RTS genre watching the FPS one?

4. Black box mechanics

A pet peeve of mine as regards Sins, or Dawn of War with its useless damage ratings.

7. Putting story in the wrong places

Not so sure about this, the story segments are in effect the reward for winning. Beat the mission get a cutscene. It's simply that most of them are ass, this is the problem.

[Kareem]

Brad could have disposed of all of the campaign scenarios in all of the GalCiv II games and worked on something else instead. Does anyone play those things?

No, it's not like we were talking about campaigns in this very thread or anything.

[scharmers]

No, it's not like we were talking about campaigns in this very thread or anything.

Putting story in the wrong places

http://img.photobucket.com/albums/v222/scharmers/1173277998-1169739965238935.jpg

[IndridCold]

I disagree, the stuff is there to make the game appear more interesting than it is. Take Total Annihilation, a game a noob can love because it seems you have a million possibilities and different ways to play. You don't, there is a very small set of core units and once you realise this I suspect most people lose interest because it stops being about discovering new ways to play and simply refining the existing ones.

Not sure I completely understand what you mean..maybe I'm misinterpreting you here.

Correct me if I'm wrong, but I always thought it was considered poor game design if a strategy game contains any unit, upgrade or ability which costs some kind of resource that is always the "wrong" thing to do when you are playing at a more competitive level. Lots of units are indeed fun, but (as you say above) once you begin to understand the game and figure out that there are units that you should simply never build because they are always inferior to other units that you must always build - then it makes sense to me that those inferior units should not be in the game at all...

I've never designed a strategy game, but I remember reading this once. Every single choice (unit, building, upgrade, ability etc) should be a viable choice for the player. I think CoH did a good job with this - at least back in the earlier versions - not sure if they nerfed it into a different game through patches.

[Alan Au]

About scripting, note that the complaint is about scripting when it violates player expectations about how the game works, or when it places too many restrictions on player actions. I would suggest that quick-time-events are an example of scripting gone horribly wrong, since they're basically cutscenes with a failure mode.

As for the problem with too much stuff, it sometimes occurs as a response to the complaint about limited play variety. Maybe a good compromise is to feature a small set of basic tools that interact to provide complex options. Personally, I like having more stuff so long as a given subset is manageable during a single play session.

When it comes to transparency of game mechanics, it's useful to have access to the numbers and formulas when developing intuition about how the game works.

I'm in favor of open code/data, especially because developers are sometimes unwilling/unable to work on a product post-release.

I'm going to sidestep the anti-piracy issue for the moment, except to say that I'm happy Soren is calling people out on this one.

Talking about story in the wrong place, I think this is really just the usual "bad story is worse than no story" problem. Speaking of which, I heard about a studio that made the mistake of contracting out their CG work for their intro movie without consulting the writers, the end result being that it contradicted some of the story elements. However, the CG work was pricey enough that instead of redoing the intro, they rewrote the story in an attempt to salvage things.

- Alan

[Jason McCullough]

Yeah, the scripting thing is basically "don't be stupid with it."

[jpinard]

I like Soren tremendously, but as far as RTS's are concerned he's wrong on points #1 & #7 (and even with some TBS's). Red Alert, C&C, Dune 2, and Warcraft are prime examples of compelling stories that make you want to play. For example I could have made the story much more compelling, fun, and convincing in Galactic Civilizations Dark Avatar if I'd had more scripting capabilities... The crux of that game however is randomly generated scenarios not the single player campaign, BUT... IF I'd had the scripting tools (I believe at the time all I could script was start/end conditions) I would have been able to give the player the ability to feel like they were in control of what was going on while adding elements to flesh out the conflict.

So in this respect, with Gal Civ you have an attempt to make such stylized alien races, but you never get to "know" them based on random games since you quickly see their conditional responses.

But what it comes down to, is we should get both -but maybe that's because I'm greedy. I great campaign that doesn't suck or make you wince when it pops up, AND great random scenarios.

[TomChick]

Soren's awesome and that's a great read. But it's funny that I love the new Colonization, partly because it commits about half of the sins Soren mentions. :)

-Tom

[Shadari]

Soren's awesome and that's a great read. But it's funny that I love the new Colonization, partly because it commits about half of the sins Soren mentions. :)
That's because sinning is fun!

[Tyrion]

If Soren is saying that stories don't belong in strategy games (which he might not be, the wording is confusing), I strongly, strongly disagree.

My best counterexample? Sid Meier's Alpha Centauri. The story in that game was brilliant, and improved the experience hugely. The key was that:

1. It was superbly written.
2. It was tightly integrated into gameplay. Little text boxes would pop up for key events in the gameworld (for example, when the player first discovered Mindworms on the map), and the rest of the story was told through descriptions for the various technologies, buildings, characters, etc.

I agree that plopping in cutscenes in between missions, and throwing in lots of heavy-handed story scripting, are bad ideas. But the SMAC approach shows that Soren shouldn't be too quick to dismiss the notion of story-based strategy games.

[Lorini]

But Alpha Centauri wasn't a scripted story. It was a story that you created, they just gave you the framework. Boardgames are going this way too, if you look at a lot of the session reports on www.boardgamegeek.com you'll see people telling 'stories' of their games.

And I think this kind of storytelling in strategy games is awesome. I want more of this. The scripted stories not so much. My guess is it's a lot easier to script a story than it is to create a framework and have the player script their own story that changes every time they play.

[spiffy]

I agree with the point #1. The only stories I've ever enjoyed in a strategy game were in Myth 1 + 2, everything else has been utter cheese ( especially in Blizzard games. I just don't understand the love .. elves n orcs n babbledeeblah).

Games that let you create your story, like Totalwar or MOO are so much more rewarding.

[Quitch]

It's not the Blizzard stories which are interesting, for the most part it's the characters.

[Qenan]

Parts of Alpha Centauri are completely scripted. It was a great game, but I do get a little tired of the same events over and over.

I liked Soren's essay, but I'm not even convinced these are the biggest sins.

[malkav11]

I mostly disagree with points 1 and 7 as some people have. The more an RTS game focuses on multiplayer and AI skirmishes, the less time I have for it. If there's no story-driven single-player campaign at all, I won't play it. But I can definitely think of some turn-based strategy games that are either best without campaigns or should never have focused on them. In particular, I've found that the Age of Wonders games, which of all the fantasy 4Xes I've played come closest to Master of Magic, miss by a country mile specifically because they have campaigns and designed maps instead of a random map generator like MoM had.

On the other other hand, I really love tactical RPGs like Fire Emblem and Final Fantasy Tactics and I don't think I'd love them nearly as much without story.

[salwon]

Medieval: Total War had the best story in any strategy game I've ever played. But that's only because I never actually conquered the whole world in MTW2.

[SorenJohnson]

For the record, I'm not anti-story or anti-scripting, per se. I just feel that they are being used in a rote way in many games.

I loved, for example, the _mechanics_ of the campaign in Rise of Legends. I didn't necessarily hate the story (didn't really care one way or another...), but I did hate the fact that the existence of a story meant that I could only interact with the meta-game in one fixed way. The opportunity to create interesting variations on the campaign mechanics, with random maps or random goals or whatever, was lost for the sake of a story. Everything comes at a cost, and I don't think developers often consider the costs of having an inflexible story. (On the other hand, I think having flexible story elements in games - like in Alpha Centauri or in Pirates! or the event system in Beyond the Sword/Fall from Heaven - is awesome.)

As for scripting, I also think that developers are not considering the costs of violating the rules of their own games by having scripting control the progress of scenarios. When I play a game, I want to know what part of the rules matter and which ones don't. And I don't want that to change half-way through. You can still make great scenarios without hard-coding the events and futzing around with the core rules.

[Quitch]

I think having a campaign is like having multiplayer, you include it to check the review boxes. How many reviews knocked Sins for a lack of campaign? I can't think of many games less suited to one, but there the reviews were, complaining it was missing.

In the FPS world you see games knocked for missing multiplayer. I bet Deus Ex got roasted for it.

[Alan Au]

In the FPS world you see games knocked for missing multiplayer. I bet Deus Ex got roasted for it.Multiplayer-only games are also risky, because they only work if there are enough players online. This is less of an issue for MMOGs which include solo activities, so really I'm talking about games like Quake III Arena, Fireteam, or Command & Conquer: Sole Survivor. That isn't to say the multiplayer-only games are doomed to failure. Consider successes like Counterstrike, Battlefield 1942, and Team Fortress 2. However, note that CS started as a mod of a popular singleplayer game, as did the original Team Fortress. Besides which, TF2 was bundled in with the Orange Box. BF1942 is a special case--it had that amazing demo, and I wonder if the game would have succeeded without it.

- Alan

[Fugitive]

It doesn't necessarily have to be multiplayer-only, though. You can still have hand-crafted scenarios and randomly-generated AI skirmishes; it's just that some people think it's pointless to arrange them into some specific, forced order and call it a campaign.

Likewise you can still have scripting that doesn't violate the game rules by having it control the AI's decision-making process, rather than spawning stuff out of nowhere or making them invincible until a certain trigger. A scenario where you have to stop Faction A from overrunning Camp X can be done by a scripted AI that decides to build units and send them to a certain area, but still has to adhere to the usual base-building and production rules.

(Unless I'm misunderstanding and Soren is objecting to even that level of scripting.)

[malkav11]

As for scripting, I also think that developers are not considering the costs of violating the rules of their own games by having scripting control the progress of scenarios. When I play a game, I want to know what part of the rules matter and which ones don't. And I don't want that to change half-way through. You can still make great scenarios without hard-coding the events and futzing around with the core rules.

Give some examples, please? I can't really think of any scenarios I'd consider "great" that didn't hard code events and futz with core rules, at least in RTSes.

[Alan Au]

Give some examples, please? I can't really think of any scenarios I'd consider "great" that didn't hard code events and futz with core rules, at least in RTSes.I thought Soren did a good job of providing examples of broken scenarios in the original piece. This includes situations where the AI suddenly acquires units/abilities/behaviors in response to a scripted event, like the destruction of a story-specific buildling or the player reaching a specific point on the map.

Are you looking for examples of how to do this properly? That's a bit trickier. Some missions suggest actions that lead to in-game consequences, like destroying an enemy's lightly-guarded powerplants and causing the automated defenses to shut down. However, these sort of missions can still feel a bit too much like puzzles and not enough like strategy. At the same time, linear missions and constrained solutions can sometime make for better storytelling, because the developers can weave a story around actions they know the player must perform.

- Alan

[Tyrion]

Give some examples, please? I can't really think of any scenarios I'd consider "great" that didn't hard code events and futz with core rules, at least in RTSes.
I'd say the campaign levels of Myth 1/2 are these sorts of scenarios. Ok, they do have some hard-coded events (mainly for story purposes), but aside from those, interesting scenarios rise naturally from 1) the specific units you control, 2) the specific groups of enemy units you fight, and 3) the layout/terrain of the levels. The core mechanics are compelling and flexible enough that they don't need to be manually altered by the developers to make each scenario different--instead, every new configuration of variables automatically generates interesting situations for the player.

[jpinard]

For the record, I'm not anti-story or anti-scripting, per se. I just feel that they are being used in a rote way in many games.

I loved, for example, the _mechanics_ of the campaign in Rise of Legends. I didn't necessarily hate the story (didn't really care one way or another...), but I did hate the fact that the existence of a story meant that I could only interact with the meta-game in one fixed way. The opportunity to create interesting variations on the campaign mechanics, with random maps or random goals or whatever, was lost for the sake of a story. Everything comes at a cost, and I don't think developers often consider the costs of having an inflexible story. (On the other hand, I think having flexible story elements in games - like in Alpha Centauri or in Pirates! or the event system in Beyond the Sword/Fall from Heaven - is awesome.)

As for scripting, I also think that developers are not considering the costs of violating the rules of their own games by having scripting control the progress of scenarios. When I play a game, I want to know what part of the rules matter and which ones don't. And I don't want that to change half-way through. You can still make great scenarios without hard-coding the events and futzing around with the core rules.

That's a very good elaboration. Did you happen to play "Hearts of Iron II:DD"
If you did, I wonder if you felt their scripted meta-events througout the war violated or complented the rules. Just curious.

[RightWrong]

I liked the background story of Alpha Centauri enough because it gave additional flavor to the factions and their differences, but its flexibility sapped any interest in it once the gameplay was underway.

As far as Rise of Nations, I'll never understand the allure of that campaign. Reviewers can chant "flexible" and "free-form" as much as they like; I just had one generic skirmish battle after another. After the first couple of turns, the missions always ended up being capitol sieges.

[Aeon221]

I liked the background story of Alpha Centauri enough because it gave additional flavor to the factions and their differences, but its flexibility sapped any interest in it once the gameplay was underway.

As far as Rise of Nations, I'll never understand the allure of that campaign. Reviewers can chant "flexible" and "free-form" as much as they like; I just had one generic skirmish battle after another. After the first couple of turns, the missions always ended up being capitol sieges.


I think my biggest problem with the campaign part of the game (other than the story) was the scale. If I'm creating cities, why am I doing in such incredibly local scenes. It makes no sense! Give me continents to fight over, not oases!

But it did have a lot of cool scenarios, especially in the second area. Unfortunately, it started to feel grindy, mostly because you were racing against three different AI armies, all focused on you, so you'd end up refighting the same maps rather too frequently.

[Kareem]

In the FPS world you see games knocked for missing multiplayer.

Bioshock disagrees.

[Matt Perkins]

For the record, I'm not anti-story or anti-scripting, per se. I just feel that they are being used in a rote way in many games.
Thank you for clarifying so I can go back to loving everything you do (http://www.whatwouldmattdo.com/2008/10/01/60-of-the-time-follow-these-game-creation-rules-100-of-the-time/). These were my exact complaints with your seven rules, that just because people hadn't figured it out correctly in most games, doesn't mean that story and scripting should be thrown out. I did get the impression from your posted sins and I'm glad to see it wasn't the case.

I do agree completely with everything you've said above. I know you were worried about that. :)

[Ben Sones]

Give some examples, please?

One idea would be to make scripted events a part of the game mechanics, rather than something that overrides them. For instance, you could have a scenario where if players manage to accomplish A and B, it will trigger a scripted event in which an AI-controlled army arrives on the map and attacks your opponent on a second front. The key is that the player would know about this mechanic at the outset (it could be explained to them at the start of the scenario, perhaps)--that way, it's not a game-changing surprise, but rather another element of your strategy. And it would maybe be balanced against other potential scripted events that you could trigger by achieving other (perhaps conflicting) goals, thus allowing the player to make choices that can take the game in different directions.

[SorenJohnson]

Give some examples, please? I can't really think of any scenarios I'd consider "great" that didn't hard code events and futz with core rules, at least in RTSes.

I should be more specific about what I mean because, of course, lots of good scenarios can have alternate rulesets and unique events. For example, the old Starcraft scenario where you defend your base from a wave of Zerglings triggered by a timer is fine, especially since the game goes out of its way to warn you up front about this upcoming event (I think there was even a clock on-screen?). Same goes for rule changes - just tell me explicitly about what is different, and we're good. What I don't like is when it's unclear which or your decisions matters and which ones don't.

Strategy games with good scenarios that didn't violate the game's own rules include HOMM, Majesty's campaign, AoE map scripts, Sins pre-made maps, and so on. My favorite example would actually be the meta-gaming site Realms Beyond which creates alternate Civ3/4 scenarios by imposing extra limitations on the base game (such as "You must declare war on everyone you meet" or "You cannot build defensive units"). Here's a link to their scenarios: http://realmsbeyond.net/civ/civ4-tournamentgamelist.html. I wish we had had the resources to do this type of stuff in the shipped game.

[malkav11]

Hmm. Okay, aside from HOMM (a game - well, series really - where the base mechanics are more than deep and complex enough to keep me interested even without the variations the campaign provides) I disagree with the rest of your examples. They just play like slightly tweaked skirmishes to me, and playing with exactly the same tools and rules repeatedly only works for me if we're talking Civ-level complexity.

Except for Majesty. I loved Majesty's campaign missions. But as far as I can tell, they *do* substantially differ from the way the base game plays. They're more than documented enough, of course, but I wouldn't call them unscripted. And as per usual, I found the undirected play of skirmish games to be pretty dull.

[unbongwah]

In the FPS world you see games knocked for missing multiplayer. I bet Deus Ex got roasted for it.
91% rating (http://www.gamerankings.com/htmlpages2/250533.asp) on Gamerankings. Bioshock got 95% (http://www.gamerankings.com/htmlpages2/924919.asp). Man, those critics were ruthless.

[SorenJohnson]

Hmm. Okay, aside from HOMM (a game - well, series really - where the base mechanics are more than deep and complex enough to keep me interested even without the variations the campaign provides) I disagree with the rest of your examples. They just play like slightly tweaked skirmishes to me, and playing with exactly the same tools and rules repeatedly only works for me if we're talking Civ-level complexity.

Except for Majesty. I loved Majesty's campaign missions. But as far as I can tell, they *do* substantially differ from the way the base game plays. They're more than documented enough, of course, but I wouldn't call them unscripted. And as per usual, I found the undirected play of skirmish games to be pretty dull.

I remember the Majesty missions being essentially variations on the main game, with the changes specified up front, which is exactly the type of thing I support. The alternate scenarios in CivRev would be another example along that front.

Ultimately, what I am trying to avoid is the feeling I've gotten in a number of RTS campaigns (AoE3, World in Conflict, C&C3) where after playing for 2 or 3 hours, I still have yet to make an actual decision outside of the preset path the scenario has laid out for me.

[Wobbo]

99% of what he said applies to all genres of games, not just strategy games.

Absolutely excellent article.