Realism vs Arcade.

Real life is frequently binary. In a war, you're shot, you go from fully combat ready to down, disabled, or dead. Drive through traffic at 90mph and clip a car could wreck your car or kill you. Falling from a 3 story height will kill you more times than you will survive.

This isn't always fun, and it doesn't always lead to the best gameplay. For example, the effectiveness of modern rifles means that modern warfare is much more about communication, spotting the enemy and positioning than aiming and running through incoming fire. Games generally work best with a danger and punishment system - when you do badly, the player is in a position of danger, realises this and can correct it. This is as opposed to a more binary system where there is little or no chance to realise that you are in danger before you are given the heavy punishment.

So in fun games, we see health bars and recharging shields for our infantry, cars and tomb raiders.

Part of this is that real life gets to the punishment by skipping the danger - you could be a soldier off to Afghanistan, on patrol for 4 boring months before you even see a shot fired. The first bullet that goes off in your vicinity could go straight through you. You could die without even realising you were in combat, without even hearing the shot fired.

And seperately, part is because of what the player expects. For instance, in a shooting game, most players won't consider (even if they are forced to witness it first hand over and over) that they are in danger when bullets are flying past them, and thus won't correct for it. Players expect to be able to go at top speed in their Lamborghini, and will be frustrated if they keep crashing and losing their car and the race because of it, even though in real life going at top speed would be terrifyingly dangerous.

This is because players want to feel threat, but the threat they experience is perceived threat, not actual threat. A player as a lone knight charging down an army of men perceives that he is threatened, even if his health bar is ridiculously long. But he expects to be able to kill a peasant with a pitch fork, even if in real life such a peasant would be dangerous. When player's expectations are not met, more often than not they blame the game, not themselves.

So, we want a satisfying game. We want it to be immersive... (I believe that realism will generally lead to immersion. I'm not saying every game should aim for it. But for games where realism is within reach, go for it.) so let's say we go for realism. This means we can't just take the easy way out, up the perceived threat by exaggerating everything (enemy numbers and our own health) out of ordinary life. We need to match actual threat to perceived threat.

Now, actual threat is what we the designers adjust. We can establish that X mechanic is dangerous and will lead to death, but if it is actually dangerous enough that the player perceives it as dangerous, it is usually as it has been taught the hard way. The player has been shot one too many times by a hidden sniper, so now has to check every area before proceeding and isn't having fun, etc. This is where we bring out the ace in our sleve. Make X dangerous, and establish that it is dangerous. And then cheat, by making the player more powerful than he realises. This way, it's not that the player beat the situation because the situation is easy. It's because the player played really well! (or so he thinks!)

Now, I suppose that some younger players will still think that, in the lone knight scenario, they won because they were awesome. But having a huge health bar is just too obvious and isn't going to fool teen and adult audiences.

We can supplement the player's abilities to allow the player to skirt close to but around actual threat. Take Halo for example. In halo, the game subtly adjusts the sensitivity of the controls so that as you move your crosshair towards an enemy, your crosshair moves slightly faster, and as you move away from an enemy (such as going too far as you swing to aim at someone) your crosshair moves slower. In game, it's virtually undetectable, 99% of players will never notice. But it means that pretty much any time you want to do an action, you can actually do it. If you see an enemy squad below, picture jumping down, landing a grenade on a leader and then hosing down the little aliens in his squad, then you can do so first time 9 times out of 10. Rarely will you

So for example, imagine we have a realistic driving game. Perhaps it's part of a Borne Identity game you're making. You've already established with the player that colliding with oncoming traffic will total you, it's a real threat, but the player is likely to do it anyway. So what we could do, is implement the steering system like halo's to guide the player away from incoming cars. And when we go into a power slide, the computer can analyse if we're going to end up right in the path of an oncoming bus, and subtly adjust the friction on the road so that they are more likely to skid by it or pull up short.

Because it's what the player wanted to happen, they're unlikely to question it. They were just 'really awesome'.

The player's caution is fixed relative to the perceived threat. What happens when the caution is less than the actual threat? We can reduce actual threat by cheating.

Analysis of Supreme Commander

I recently reinstalled the game Supreme Commander. I have played it for around 20 hours and feel like doing a game play and design analysis of it. I’ll do more than criticise, I’ll suggest what the designers could have done to change any issues. After 20 hours, you can say that you fairly well understand a game. Perhaps you’re not a decent competitive player yet, but you can feel that any criticism you make is valid as if it requires more than 20 hours to learn how to avoid, that in itself is an issue with the game.

Supreme Commander is a game of exponentials. I’ll describe that as it’s an important concept. An exponential increase is caused when the rate of increase of a number is proportional to the number itself. It forms a graph like this, which rises very steeply. The resources you get in Supreme Commander to spend on tanks etc, all things being equal, tend to go as exponentials. Being only a minute behind the other player can mean you are getting half as much resources.

It’s unfair to criticise Supreme Commander for this directly, this is what it is- it means you go from small 3 v 4 battle to, a few minutes later, 20 v 20 battles, and at the end of the game 500 vs 500 battles. This is its chief strength, that the game play works all the way up, leading to incredibly epic battles. It’s what the game is. However, it comes with some drawbacks.

Chiefly, the key battle occurs over the middle 20% of the battlefield. If you win this 20%, you generally have a big enough economic advantage to get yourself well defended there and secure yourself the win. When playing against 4 other AIs, I generally kill one early on, take the resources that are in their base and from then on I have more resources than any other player in the game can hope to match. To an extent this makes the fight for the middle 20% all the more exciting, I’m sure against humans, if this can switch hands several times, then you have a game that nobody knows who is going to win until the very end. However, this requires evenly skill matched players. Even then, you’re not guaranteed that one player won’t dominate the middle 20% early on and get an easy win, but with uneven skills on each team, this means games are decided early on and from then on, not much fun.

This is an issue we had with Vietnam: Glory Obscured, though to a slightly lesser extent than Supreme Commander. I came up with a concept called the golden zone of game play. This is whenever both teams have similar resources; both can throw lots of units at one another and have plenty of options because they have enough money to be experimental. Both players are having fun. It’s more fun for the winner too, because not only is the game still a challenge and the losing player a threat, but the losing player will be able to produce decent armies to throw at the winning player’s even bigger army. It’s pretty lame when you get yourself the biggest, most expensive unit in the game and find that there’s nothing left for it to fight.

In Vietnam: Glory Obscured, the sole way to get resources was the villages. On a three village map, with a village near each player and a central village, the player with the central village would have two villages to the opponent’s one village. He would get twice the other person’s income and could afford to buy two tanks for every one of the enemy’s. By spending all of his money on defending that central village, as he could produce more than the enemy could hope to destroy, the game was decided early on. What’s worse, the game isn’t even over. The loser is obliged to continue playing, despite knowing that he has lost. Some players just quit at this point – the best players would always quit at this point. Some didn’t believe it was sportsmanlike to quit, so just had a miserable time.

I proposed that we halve the money given out by the villages. We then work out how much money, with this new system, the players would get from controlling half of the map. We give this money to both players for free. This sounds complicated, but it’s not. When both players hold 50% of the map, they get the same quantity of money as before. However, if you lose an area, then you do lose the actual money from that village, but the free money for owning that village keeps coming. As the person that just won the area, you get half as much money from your newly captured village. You get rewarded as before, but you’re hopefully still in the golden zone.

Sorry for the spreadsheet, but it should make it easier to see the effect:

This also means that: games are played not just over the central 20% of territory on the map, but right from base to base. Games are not always decided before they are finally over- it’s possible to have almost nothing left and then come back to do well. It means that the skill floor is raised, and importantly that new players which come and lose every game they play at least have a bit more fun doing so.

What could supreme commander do to make territory control less important, so the game is still in the balance right up to the end?

My first actual gripe with Supreme commander is that it is 90% resource management sim (warning, this next paragraph is boring). You need to balance Mass income and expenditure and Energy income and expenditure. If you run out of either, your progress is stalled, and running out of energy means some units fail. You need to manage the storage limits of both. You are constantly thinking: do I increase my income now or spend it on tanks? Do I upgrade these mass extractors now or wait? That, by the way, is a whole new dimension to the resource simulator in that upgrading the mass extractors requires mass and energy, and so if you are running out of mass and need more, you can’t just upgrade as you can’t afford it. You have to ride out the low period you are in, THEN upgrade. It’s painless if you remember to do it when you have lots of resources, but if you spend frequently, you’ll end up running out of one or the other resource and be in financial limbo.

Generally I’m so busy working on this that I don’t get to look at any battles. I rarely get to micromanage my troops, as I am forced to micromanage my economy. Another very real issue this brings up is that as it’s such a skill, it has a very high skill ceiling. Each match I played, I got so much better at it that I would thoroughly dominate any earlier version of myself. And I wasn’t even near approaching the most efficient I could be, and that’s after 20 hours of play.

How could this be improved? I would make it much simpler, so it takes a total back seat. Make it so you are getting resources at the most efficient rate all the way through, without any micromanagement, so I can think about what is actually fun: Spending the money, scouting the enemy and of course the combat.

What I would do is make the mass extractors free. As soon as you conquer the territory of the mass extractors and manage to get an engineer out to the position, you can have a free, quick to build mass extractor. These mass extractors would very slowly, for free, upgrade themselves. Assigning engineers to mass extractors does not increase the rate at which they upgrade. You know, as far as mass is concerned, that you are being as efficient as you can be.

Mass fabricators would be relegated to tech 3, the late game stage. This is where, in supreme commander, all the extractors in the game aren’t enough to get you the income you sometimes need, and they are necessary then.

I understand the advantage of having a resource storage limit. It forces commanders to constantly spend their resources, which encourages them to build reasonable and decent armies throughout the game. Without it, it would be too easy for players to go into “economic” mode, where they just increase their economy for ages until they’ve amassed a giant pot of money, and then to spend it all. The limit forces commanders to juggle, which really is a fast introduction to a better standard of gameplay. One does learn, eventually, to set one’s cheap engineers to build long lines of storage, but the whole process is boring micro-management and all to frustrating when you get it wrong and end up wasting lots of resources.

A Solution: Storage capacity would not be purchased, it would be a fraction of the gross income.

Another issue with the exponential nature of supreme commander is that as your units increase exponentially, so does the amount of work you need to put in. In the first few minutes, there are frequently spells where there is just nothing to do. No units to micro around, just watching units build. But the real issue is later on. The number of things you need to be doing across the map increases beyond what you can manage, and you find yourself not just balancing your economy and unit output and logistics but fighting on multiple fronts. The experience is akin to plate spinning. What this means is that where corners can be cut, they are. First most, strategy and positioning with units goes out the window. Forget putting the artillery in the back, fast units at the flanks, targeting specific units – a swirling mass of crap will do fine. The only solution I could propose is that the economics is made simple, so that the player has time to spend on his units.

An issue that this issue necessarily puts in the game is that if you can’t pay attention to the battles, the location that you fight from cannot matter. And when location doesn’t matter, when you’ve seen one battle you’ve seen them all. It makes for far less interesting gameplay, as each encounter is unmodified. Sorry to plug V:GO again, but this is one area that we’ve got nailed. Height gives a defensive bonus against projectile units like tanks, trees give infantry a huge advantage to infantry and change some dynamics completely – where AA infantry used to fear attack helicopters, now attack helicopters must stay away. Villages punish those who bomb them back to the stone age, so you have to select units less likely to cause incidents of collateral damage. By micro managing your troops into good positions, a small group can fight off a large one. It’s not about the kind of “skill” micromanagement, where one unit can take out a tank battalion in the middle of the desert just by scooting around, it’s about positional micromanagement. This only works because you have enough time to micro your troops, you attention isn’t spread elsewhere. Encounters between groups are rarely the same twice, and this keeps things fresh and fun.

Inception though design

Using different textures for different areas of the map can do more than just describe the thing they paint (cliff, beach). We can also direct the eye and put ideas into a player's head, just as lighting for the stage and cinema can, with textures. This post describes work I've done mapping for the real time strategy game Vietnam: Glory Obscured

Valley of the Ethereal was a difficult map. I was given it in a very rough state. It had a strange texture pallet that I wouldn't have picked and a labyrinthine layout. It had the basic textures thrown across the map. However, it had a nice set piece which the initial author had spent much time on. Also, I was feeling in the mood for a challenge with the texture pallet, so I kept it.

I did my best to fill out the rest of the map and put stuff for the players in appropriate spots. Using the other person's initial texturing and working around it both looked strange and took much longer than usual.

Several months later, I decided to go through all the maps with a friend and bring them all up to the best quality they could be. This map in particular needed attention.

Players did not like the map. We decided:

- They didn't understand the layout. They didn't know where they were supposed to be guarding or why, as they had no idea which areas were valuable and worth keeping.

- The map was ugly. The textures didn't go well together.

The map is huge, completely changing the

texturing would have been an nightmare. In addition, the AI waypointing, which for a 4 player map like this takes perhaps 3 - 4 hours was already done.

Two important changes were made:

The layout was made very much more symmetrical: A pocket village (pocket meaning very near to the player's starting position, so easy to take; villages in VGO are the players' source of constant funds) counter-clockwise for each player, and two symmetrically placed ones in the middle of the map.

More importantly, I changed all of the ground level texturing away from the river. I used a light grassy texture set to contrast with the dark mountainous textures which had previously been across the map and were now exiled to just the raised areas. Now, immediately when looking at any area of the map, the players could see where the high vantage points were, where the low flat areas were. The lighter textures also contrasted better with the grey/green cliff textures, so the shape of the map was easier to work out.

Below: After and before the changes. Think: is bottom right low and flat or high or what?

Explicit statement of the stakes

The driving force pushing the narrative along in games is usually the protagonists preventing some bad thing. Usually some big scheme or plot, often (sometimes exclusively) just trying to survive. The player feels a need to progress. The more the player understands and is threatened by this bad thing, the stronger this need. Where games succeed or fail is often to push this onto the player.

Let's consider the death of the player in gameplay. Sometimes, you just don't care. Playing [Half life:] Opposing Force recently, I found that if I'd just saved then I wouldn't care if I died, indeed I was reckless often just running out of cover guns blazing, expecting to die. A couple of minutes later, I found myself having survived a couple of very hard encounters. I was hidden beneath a staircase reloading a shotgun and listening to the pattering of a female assassin's feet around me. The assassin could appear and fire off her gun in a fraction of a second, killing me. I managed to reload two shells just before she appeared and as she ran across me and crouched into a firing position, got her with both barrels. It was one of the most tense moments of any game I've played recently. Opposing Force has a quick save system, that at any point you can save and return to that spot if you die. Of course, I could have saved while under the stairs, but that would have made the encounter trivial and destroyed any tension.

Saving in the middle of a fight removes any threat of losing something, in this case losing progression made. Thus when playing Opposing force, I choose not to use the quick save.

You can give a constant punishment for dying- perhaps your game continually saves, and dying sends you back to where you were 5 minutes ago. Perhaps you lose a life, or money, or your character becomes uglier each time they die. This should keep a constant level of tension. An example of this would be the gameboy game Mario, where you are constantly at threat of losing a life, right from the start.

You can use a checkpoint system- reach a destination and you automatically save. This is nice as it provides a sawtooth shape graph of tension against time- as players get towards a checkpoint, they have more and more to lose, so each encounter becomes more threatening. At the checkpoint, the player can then relax. This is less mentally tiring on a player and gives a broader range of emotions for the player to experience. The highs are higher because he has lows to contras with. A checkpoint system trivialises the game just after the save, but this also lets the player mess around. With too many checkpoints, as in mass effect 2, you rarely accumulate any progress to worry about losing.

I don't care about dying for another reason. If it's too easy.Each time I die I only feel "meh, if I'd really been trying that wouldn't have happened. Also, the game got lucky there." A game will never be an exciting fight for survival if the player does not have to engage his brain.

Gameplay aside, consider the narrative of the player's life. Ignoring supposed-to-lose fights, the player is rarely sent on an impossible mission. If his mission is to kill the Russian president, then all of the Russian army, then blow up the moon, in an hour or so the player will probably have done so. Being told that the player is a supreme bad ass when there was no other option carries little weight, and can smack of macho posturing (there was no other option, so it's not surprising you managed it). Similarly, a protagonist can expect to survive a cut-scene, suicide mission or falling building. While it's comfortable to have a game roll along like a sitcom, everyone back to normal at the end of each episode, we sacrifice both threat and character development - by which I mean that many good narratives have characters who change as the story unfolds. Grand theft auto has this for most missions.

By contrast, consider the end of the game Mass Effect 2 (Spoiler Warning!). Here, the charecters who have previously led a charmed life die a variety of random, sometimes unfair deaths. The first person in my party to die was crushed in his own quarters by a girder! Not a heroic end for someone I'd come to know so well. Once my party member were falling off, I started to really fear what my happen next. You know the cliche where someone falls off a cliff and is grabbed by the hand at the last second? I've seen this a dozen times in films and it never fails to make me yawn. You guess, sub consciously, whether the character is supposed to die at that point, and work out the result. Well, the same exact thing happens in Mass Effect 2. Two characters tumble towards the brink and I'm surprised to find myself scared for them. Consider the film Jackie Brown, directed by Tarantino. In the finale, one really does not know what could happen. It's a Tarantino film, so the antagonist could easily shoot everyone in the building and walked out with the money. What could be a spectacular but never tense end to another film, by creating the notion in the viewer that the normal rules of cinema do not apply, Tarantino is able to make the finale genuinely tense.

I have experienced enough games and films to understand that if one does play by unnatural rules about who can die and when, it requires immersion breaking suspension of disbelief and removes any tension caused by threat of the negative consequence.

Finally, lets look at explicit statement of the stakes. By this, I mean that you tell the player what they are risking. This brings to the forefront of the player's mind this threat that compels them to play. However, it only works if the player understands that the threat is real. Being told something a player knows cannot occur because of unsaid rules will only cause eye rolling. However, if the player believes the threat to be real, then this is a powerful tool. Consider Ordell (Samuel L Jackson) in Jackie Brown, at the finale:

"I go walkin' in

there and that nigga Winston or

anybody else is in there, you're the

first man shot, understand what I'm

sayin'?"

This man has killed several men already. We know he is capable. We know the director could allow this to happen. It's very effective, and means in the scene that follows we know exactly what to be worried about. Without this dialogue, as the men including Ordell enter the building, were it actually a trap, Ordell killing this man would be completely out of the blue. There would have been no tension as though we know all the information laid out at the top of the paragraph, we are not thinking of the threat.

This is of particular importance in historical games. Battles which occurred in the past are less relate-able, and it's easy to feel that as it's done and decided, there was not the same passion and fear that we experience now. It's hard to imagine the emotions running through a footsoldier in say, the year 1415, being expected to engage in a fight to the death with swords and spears. I've yet to play a game that really communicated this.

Games are more engaging when there is a realistic threat that the player is aware of, and we most make every effort not to remove this threat with unsaid rules about who is allowed to die and how.

I'm re-reading Sirlin's three excellent pieces on good writing:

http://www.sirlin.net/articles/writing-well-part-1-sensibilities.html

"I figured it was like ballet dancing; dancers strive to be the best they can at their craft for its own sake, as well as to impress the judges—that small group who can actually detect the nuances between two different performances.

Writing is not like this.That’s all wrong. Writing isn’t for English teachers or judges of essay contests—it’s for everyone. It is our most pervasive tool for communicating ideas. You should care about writing not for its own sake, but because you care about ideas. You care about clear thinking and the clear and honest expression of that thinking. "

Even if you write only the occasional email, it's very useful.

I think about design a lot

One of the worst feelings I know is when you know you have forgotten something. Especially when this is an idea, it is like losing an item that is no longer in stock, leaving a book on a train and realising it's no longer in print. This blog is making me very happy at the moment, as it's telling me that I never need to forget an idea again.