"Too Many Clicks!"

Gamasutra has published a six-page article written by Philip Goetz about the very high number of mouse clicks, keystrokes, and mouse scrolls required in computer games. It states that the unit-based user interface (UI) is no longer sufficient and suggests a different way of thinking about UIs. Civilization III's unit-based interface is used throughout the article as an example. Below is an excerpt from the article:

Computers can now animate more units than any player could reasonably want to control, and the number will continue to increase exponentially. This leads to player frustration rather than fun. In a good user interface design, no player should control more than seven game entities. To enable this, the UI may let the player control something more abstract than an on-screen unit. This requires object-oriented developers to think of code objects as abstractions beyond the mere units on the screen. The UI may also give the player a chance to specify behaviors off-line in order to reduce the amount of on-line supervision needed.

Click to expand...

I have to agree with the author. Games along the lines of Civ, as well as most RTS, have gotten a bit out of hand. I, too, dislike the endgame in Civ, not because the game isn't amazing, but because the size of my empire and armies gets overwhelming. This becomes especially true when you develop a navy, which is fairly simple before endgame, and airforce, which is nonexistent before endgame. Shortcuts (like route-to or auto-bombard) can only partially alleviate the tedium and others, like automated works, just don't work that well. What we need is exactly what the author calls for, a new conceptualization of what the player controls.

However, just thinking about creating a distance between the player and individual units inspires fears about losing massive battles based on "luck". Sure, we have all lost a tank to a spearman at some point - but how about losing a massive army based purely on game mechanics that you cannot control. I think it comes down to two competing camps - the "builders" who would welcome a streamlined warfare system stripped of micromanagement, and the "warmongers" who wouldn't play the game without direct control of each and every unit.

Furthermore (am I writing my own article here?), I dismiss the idea of having it both ways - micromanagement for those who want it and automation for those who don't. As any decent Civ player knows, automating workers, except perhaps in endgame, is simply bad game play. Micromanagement is what lets you beat an unfairly advantaged AI. Given the choice, we must choose to micromanage, even if it means we stop playing once the modern era dawns. Which is why the author is on the right track. We don't need more automation options like those in Civ, we need a new conceptualization of how such a complex game can be played.

Any ideas?

lmao, this guy needs to play Starcraft. Or watch a first person VOD of a professional playing. He'll become dizzy and faint if he thinks Civ has too much clicking.

I've been thinking about this while playing lately as well, so this is a timely article for me.

I think Civ4 is about as "big" as the game can get. I think in the next edition they should really go for simplicity (don't just make it an "option".) Significantly reduce the direct control you have, but make the decisions you do make very important.

They should also make the game more linear- give more options at the beginning, less towards the end. Obviously, the game will always get bigger towards the end, but try to reduce the huge variance between the first 50 turns and the last 50 turns.

I have seen wargames similar to what the author suggests, and believe the majority of Civ users will think him silly, if they don 't already, if they saw one of these let-the-computer-run-things stupidity. The key is that when you have bits that the computer controls, either abstract them so the player doesn't really know they are going on, or have the player control them. There's nothing worse than telling a unit to go one place and the wise computer sends them in another (claiming the unit is using it's own mind in the process). We have seen this sort of thing even with the Civ series, where, for example, you tell a settler to go to a spot to found a city, showing the route before he goes there, and then whammo without fail he takes another route, although often the different route isn't a problem. In fact today I had such an instance where the settler for some odd reason actually took an extra step completely out of way towards the goal and directly into the path of a previously seen barbarbian. I'm not talking that the settler had been moving to that objective from a previous turn either. I told the settler to go there with the barb already in view; avoiding the barb was extremely easy. At that point I re-booted my last save, and the exact same thing occurred again. So? The solution was for me to move it manually until it was out of danger altogether. Yeah, like I need more computer help!

Face it. This sort of article is made by people who spend very little time to play games, so they want to dumb them down, and really don't have a grasp on just what the computer can do when unleashed, and the Civ example is very mild compared to a wargame or two I've had the displeasure to experience that were of the automated class. I do like some more ease to some of the play, but i n the right places. Having the view that somehow that unit is thinking on it's own, I've yet to see it work.

When people start putting more effort into the computer making logical decisions, especially for your own side, then and only then is this play a dumbed down game worthy of consideration.

iamke55 said:

lmao, this guy needs to play Starcraft. Or watch a first person VOD of a professional playing. He'll become dizzy and faint if he thinks Civ has too much clicking.

Click to expand...

You know when I saw this, my first reaction was "Finally someone's gonna put those damn SC players in their place". I love RTS games, I love War3 and I hate the SC players who think it sucks because you don't need to get carpal tunnel to play well. I can't believe they actually consider rapid clicking to be a skill worthy of high praise and monthes of dedication.

Charles 22 said:

Face it. This sort of article is made by people who spend very little time to play games, so they want to dumb them down, and really don't have a grasp on just what the computer can do when unleashed, and the Civ example is very mild compared to a wargame or two I've had the displeasure to experience that were of the automated class. I do like some more ease to some of the play, but i n the right places. Having the view that somehow that unit is thinking on it's own, I've yet to see it work.

Click to expand...

I didn't get the impression the article was saying "keep the game the same, but let the computer control more", but then I did skim over some of the more technical speak.

I think the answer is just change the way the games are made- keep the player in total control, but lessen the micromanagement aspect. For example (and this is just off the top of my head, don't take it too seriously), why build every single unit- Why not have army bases that can send out an army based on your tech level and military spending?

I think it's more about re-thinking the game, not just handing control to the AI.

I think the problem is less that lots of mouse clicks and so on are required to play computer games, but that it often increases exponentially as the game progresses. For example, I'd say my "ETA" to 1 AD in a marathon civ4 game is 4 hours...and I'd also say that's how long it takes me to play a golden age from the Taj Mahal, as I'll check every city every turn so as to make sure I don't waste any part of it. I don't mind some micromanagement, but it needs to increase in a more linear manner. While making the UI count go down where reasonably possible would be great, dropping 100 clicks out of 1000 at one point in the game isn't going to change the exponential problem as long as it's dropping 10 out of 100 at the earlier point. Even if I'm having fun with a long game, the realization that it will take me several hours in the endgame to go through as many years as were in one turn at the beginning of the game can become a deterrent to continue playing. Exactly how long later in the game doesn't really matter as much since, while an hour or two saved from better UI is nice if I notice it, what I'm noticing more is how little time it took in the early turns of the game.

I don't, however, by any means wish for any sort of AI to control what I do until it's, well, actually intelligent - while I think it's excellent for AI, if I was going to miss any more than a turn I'd have a hard time choosing between letting the AI do its thing and just putting the workers to sleep and my cities in anarchy. In fact, I'd much rather grab a friend that I know is good at games but has never played civ before, give them enough instructions so that they know enough to play and let them have at it for a few turns.

While there's the military reasoning, I still think that the "rule of seven" seems entirely arbitrary - and far too small. He mentions chess - yes, it has sixteen pieces per player, but it can actually become more complex with fewer pieces! Managing several large stack of units often isn't really that big of a problem - partly because they've been grouped together, but as long as they're active I don't find large numbers too problematic. In fact, I often move every unit individually once I've split up a group even if I could reform it, partly because I enjoy making tactical maneuvers if I'm playing well. I've easily had not just more than sixteen "pieces", but more than sixty on my "chess board" in a game of civ and it's rarely been a problem...and when it has been, the inactive units have been more of one. Different people can manage different numbers of "units" and some certain types better than others - thus giving players better options as to what, how and how much they manage would be better than having some ideal management amount.

What I think could help the problems would be better ways of implementing non-AI delegation, which would probably require better notifications and queues. I don't queue many buildings on civ4, for example, as I prefer to still check the city screen when a building finishes - but it would make for less management if I could have the option, enabling me to skip checking when I know I don't want to and making it faster when I do. Being able to set the number of turns to build something and then switch to something else would be great, too. It would be helpful to have notifications for city growth, an option to inspect the city, and, particularly, a queue for which tile to work next. Go to orders could be queued, too. Other changes and improvements to the general way the interfaces of games work could help streamline some of the management issues.

As another example, I'd love to be able to organize my inactive units better with a set of "group/select/wake ___ units" - by location, unit type, city/non-city, etc. In the case of the civ3 railroad dilemma, being able to outline an area of the map, select all workers in that area and then pair them up would be delightful, but not unreasonable to expect. Adding in "build railroad from point a to b" might be even more delightful, but I wouldn't consider it reasonable to expect both due to the increased specificity in the nature of the order and the fact that I wouldn't trust the computer to build the railroad quite where I wanted it. The same might go for adding a "go to this city" order to a unit when it's built, but certainly not "go to whichever of city x, y or z will have the fewest units when this would arrive". Again, the general idea would be to delegate tasks to the game through simple ways to issue and change multiple orders rather than having to delegate to an AI.

I don't necessarily expect games to have a "perfect" way to manage everything, though. If a game is designed well it should become more fun as it gets more complex. Games won't sell because they won't get played if they're overly complex. Sure, sometimes a reasonably complex game will get to be a drag, but there's nothing wrong with taking a break from a game and coming back to it. Certainly I don't think civ4 should be dumbed down in any way - nor given more to the AI to control any more than the player wishes, no matter how good the AI is. No game should ever be a matter of speed of or determined by button mashing, but certainly some amount of button mashing should be just fine.

I don't think you quite get what he means when he says that seven entities are the upper bounds of what a human can remember. Based on several experiments by noted psychologists, that number popped up again and again. A psychologist whose name eludes me practiced memorizing long random lists of letters, and found this to be his essential boundary without practice. Same with most other studies.

Its not like this is some out of the blue number he chose; its backed up by robust studies. None of which I can cite right now, but some of which were referenced in my own Psych 111 book. The author is perfectly correct when he references the command and control system of the US Army. Most other armies have a similar assumption.

Obviously, experience with a system opens up shortcuts and enlarges the number of entities that you can command at one time. A student who participated in a long term version of a memorization study was eventually able to memorize absurdly long lists of numbers by linking them to various track times. When asked to do the same with randomized lists of letters, he was utterly in the dark, not having developed a system. Chess masters, when presented with an utterly randomized chess board, are just as inept as chess novices at memorizing the positions of the pieces.

I would personally lust after a UI improvement that allowed me to simply specify that a defensive line be established on my northern border, or that a railroad be build between X and Y. That is what real world leaders do: they issue a general order, and allow subordinates to deal with the specifics. RTS games are, after all, supposed to be about STRATEGY, not tactics.

Also, you ignore a major point made by the author: those 60 some units you control behave in essentially the same manner. Its not as if you are controlling 60 some units which each behave in some radically different manner.

Perhaps, then, my active units would be considered a single entity and my inactive ones another, but then I'm arguing less that the rule of seven is wrong but that the problem is in the computer/player interaction rather than user interface. One of the problems with computers is that there's a seperation between AI and a program that simply executes orders - if the computer does exactly what it was told, it was never "thinking" for itself. Humans, however, can be trustworthy delegates, as they are (usually ) capable of both being intelligent and following orders at the same time.

This is why I argue that making the orders themselves more efficient - rather than more complicated but with better implementation - would be more productive. Not, of course, that better implementation wouldn't be nice - in fact, I think it would be a great improvement to have better UI...but less strain on my fingers and/or shorter turns doesn't make the endgame less complex or unwieldy, it just is less unpleasant. It's similar for more complicated orders - it sounds good, but it simply is not possible to consider any AI trustworthy. A human can reliably understand, interpret and execute a detailed order (such as "go to and build city at a location that will make a great person farm"); a computer has to be told specific details ("go to and build city at a location that can produce at least this amount of food and that amount of production"), at which point one might as well just select the location and make a simpler order ("go to and build city here").

That does bring up another possibility for delegation, though - a team of human players that plays the same civ at the same time.

First of all, I'm very interested in ludology, and so Gamasutra is quite a find for me. Secondly, the writer is dead on. It's absolutely my experience that by the late game, or even by 1000 AD or so if I'm playing a warmongering strategy with lots of units, the game becomes highly unwieldy. It begins to feel like busy work as I direct 30 of the same unit all over the map. In fact, that problem is exactly why I tend to focus on a cultural/diplomatic/space race victory, because it involves me caring for only 5-10 objects (my cities) rather than forcing me into having to care for numerous lower level objects (units).

In further iterations of the game, I hope that the design team will construct a UI that involves delegation. This delegation could be managed far better than merely "Automate workers." Rather, there could be varying AI's that represent different advisors/senators/generals/etc. which could be hired, fired, appointed, elected, could be loyal, disloyal, could mutiny, share information with one another, or try to maintain control of their "own realm," could come up with their own strategies that stray from (improve upon?) your plan.

I do understand that this runs the risk of moving Civilization towards too much of a "Political Simulation" rather than what it is right now, but I do believe it would be an interesting road to take, and one that is much more realisticially representative of "ruling a civilization."

If one is scared of tainting the game with political simulations, maybe what the designers ought to do is realize that, at its most basic level, Civlization is a board game. It even looks like one, with every part of the map being not realistic, but rather representative. Giant units walking throught tiny forest squares. Cities with 12 citizens. Working tiles of land. It's all a board game, right down to the core. The problem is that there is simply no way that Civilization 4 could come even close to existing as an actual board game. It would be far too complex, and not just in the numbers sense. There would be too many pieces to keep track of, too much to handle, too much busy work. Maybe the best idea is for the designers to think about handling the game in a more board-game-like manner, and reduce complexity organically with that rule as a starting point.


EDIT: Kudos to the poster above me. Having a Civilization run by many humans would be a brilliant plan, creating truly intelligent delegation and creating a more social game environment. It actually sounds like a very fun idea.

Charles 22 said:

Face it. This sort of article is made by people who spend very little time to play games

Click to expand...

That's wrong. I've been playing CIV since the first version, even almost every side version. Because the game topic is actually fascinating and addictive. And I share every single point of the author. I'm sure I spent more hours playing Civ in my life than you.

Though since CIV II or CIV III I actually experience this MABS (Modern Age Burnout Syndrom ). This means that in the late game if I don't have the time because it's over midnight but you want actually something to happen before you go to bed. But half an hour later only 4 years have passed. As a result for huge empires you don't take the time anymore to think about every city but just click on the next building to be built (and don't ponder about it anymore for long). Often enough with CIV IV I simply let an unhealthy city just starve away ignoring the why because my domination victory is just 20 turns away (meaning still more than one hour to play).

Your CIV AI experiences are there because no CIV designer actually tought of any of such game abstraction concepts. No settler would run into the lion if you issued the abstract order "put a city (specialized on science) there". The AI would simply build or summon an archer (if you are in war or the "barb threat" is really serious a little army) and a settler. Ok, you might say, I would send the settler alone. But that is your risk. If the settler would be safe on its own the AI could do it as well. This is such a typical scenario named by the author you could choose. Protect the settler or take a risk...

Anyway, I for myself, consider this article being like a distant dream. That's not because I don't think it to be possible. But because I think that the game industry would never support such a revolutionary game development. It's to risky for them. Because they can't know today if they would ever get any interested customer for such a game. Because it's difficult to estimate how much research and development efforts such a revolution would cost. Because it's difficult to estimate the system requirements for such a game. It's rather something for some little company or even open-source game development community. Because they would develop for the fun and not for their net profit So don't expect it from Sid Meier.

If you plan abstractions from current CIV game concepts like this you could even make the AI truly challenging. Because the difficulty the AI has to keep up with the player through the lack of abstraction. Yes, the AI suffers from the same game concept that we do. The difference is that the game developers have to program the whole AI before the customers play the game and develop better strategies that the AI could ever have. A human brain is simply more powerful that a computers (preprogrammed). But a human can get bored, while a computer cannot.
If there were abstract game mechanisms like 5 generals that present options to and receive orders from you, the AI would have it easier as well. Because it could use the same simple concepts. But as I said, it won't complain anyway. It cannot get bored like I can

Thalatta said:

That does bring up another possibility for delegation, though - a team of human players that plays the same civ at the same time.

Click to expand...

Imagine that as a multiplayer - the ultimate team game.

1 person in charge of army building.
1 Person in charge of finances.
1 person in charge of infrastructure.
Have the team decide the strategy as in real government/civics.

I think you get the idea, but, it is one way of lowering the UI level and complexity and raising the multiplayer level.

For what it´s worth, I have to say I think the author is spot on.

I love CIV, but the modern age can become a long winded grind (especially if you have to get up to go to work in the morning!).

With CIV IV, I tend only to play for cultural wins simply because you don´t have to micro-manage vast armies over large distances.

Someone suggested that the game designers won´t take on the radical ideas suggested in the article. Whilst they might not go the whole hog, I don´t see why some of the ideas shouldn´t be taken on board. The notion of being able to specifying what you want to build where and letting the workers just get on with it seems eminently do-able. Identifying borders to be defended, and just letting the troops take up position seems do-able too.

Moreover, even if there is still too much micro-management in CIV IV for my taste, the designers did make a decent fist of reducing the amount involved from previous versions. I don´t see why they shouldn´t move further in this direction in CIV V, CIV VI...

darkman-perth-a said:

Imagine that as a multiplayer - the ultimate team game.

1 person in charge of army building.
1 Person in charge of finances.
1 person in charge of infrastructure.
Have the team decide the strategy as in real government/civics.

I think you get the idea, but, it is one way of lowering the UI level and complexity and raising the multiplayer level.

Click to expand...

This is actually possible in Starcraft, one of the many reasons I love it despite it being ancient. The problem is, you need friends to do this. You have to play with people you trust, though it can make some really fun games. Playing with a griefer would just make it .

Here's a crazy idea: play an RTS in first-person. You play the role of a king or general giving orders in first person. I could see that being really fun.

If this guy was correct then Savage would have been a huge hit.

About the too many clicks I have somenthing different to say.
When I used to play a few days in a row for 12 hours a day (good old days) my finger and my hand where really hurting. I had to stop playing for at least a day. The worse is when you have to pick a destination for a unit. You have to hold the button and is really hard on the tendon of your finger.
I don't know if any other player had the same problem, but it's somenthing that still effect me.

Regarding the endgame, I must say that I didn't finish many games just because it was taking tooooooooooooooo long to move everything around. It's when you know you already know you won and all is left to do is run over the remaining civs.The game is not a challenge anymore and it becomes less fun. It doesn't really make sense spend more long hours on a game already won.

I also agree with the author, but I would fear to loose to much control when I want it, I don't want the AI take over control for me. So I think the developers need to make options availible where the AI should take over, or what to do when. And you need to be able to do everything yourself but when you're bored with doing something you should be able to give the command away to the AI. And when you want to stop the command wich was given to the AI you should be able to do so straight away.
And late game is boring I almost never see the late game due to early dropout because I became bored and started over.

But still when I read all this back, I doubt if it will be fun enough to give some commandm way to the AI... It's a though dilemma...

I think that the system used in Civ (and most games) is good. You have to choose between managing everything (and destroy your fingers), or let the computer do the job (and destroy your game ).

When I see guys playing wargames (you know, these games in which you move colored squares on hexagons ) with hundreds of units to manage individually, I'm not sure that players really complain about the amount of clicks in games

Author makes some very valid points, no doubt about that. Designing a user interface is one of the hardest things about creating games, and the Civilization games are no exception to that rule. I found it interesting that the author chose to use Civ3 for his paper, and specifically two extremely tedious game elements (moving hundreds of workers and haggling over pennies on the diplomatic screen) which were eliminated for Civ4. The most recent Civ game was actually designed to remove the problem of needing hundreds of workers, and with a single click you can figure out exactly what the AI is willing to give you in diplomatic negotiations.

That doesn't take away from the validity of the argument, but it's a bit of a strawman tactic to focus on interface problems which WERE adressed in a game's sequel.

The article is spot-on at the theoretical level, but I start having my doubts about the actual implementation of some of these ideas. The author seems to be suggesting that games should be streamlined by removing direct control over units, cities, etc. Instead he appears to want the player to work through governor mediators, who would deal with the nuts and bolts details of the game. Well - that's a noble goal, and maybe a designer will be able to implement such a vision successfully one day, but I'm a bit leery of the overall idea. This very concept was attempted by the game Master of Orion III, and it was one of the biggest duds ever. You literally had to fight the game's automation to issue any orders, and if you weren't willing to do that, the game would play ITSELF with no input, building fleets and settling new colonies on its own. It was possible to click "next turn" 200 times and win the game without doing anything. All in all, it was one of the least-entertaining games of all time.

So again, it may be sound in theory, but when you start taking away control of the game from the player, you may not be left with very much stuff that would constitute fun!