Saturday, March 19, 2011

Metaphor for Skill vs. Imbalance.

Introduction
Do you ever find yourself arguing over balance on forums?
Damn right you do - or at least I do occasionally.
One thing I find in every single gaming community is that many people (new to the game and experienced alike) don't understand that when discussing balance, you must eliminate skill as a variable.

Hence, most discussions are like so:
"X is too powerful"
"Just use [unreliable attack] to kill him!"
"But that assumes he just stands there and lets you use it!"

In this situation, someone defends a unit by assuming a player of low skill is playing it.

Or
"X is too powerful"
"I find it to be balanced as when I'm playing, other people kill me with [unreliable attack] all the time!"
"That attack is easy to counter if you are aware of your surroundings, though."

In this situation, people often degrade their own skills, or give examples of extremely strong opponents to defend the unit.

Obviously, all of this argument is invalid, because, your test subjects should be at an equal skill level, or, as I posted in a previous article ( http://zandagamedesign.blogspot.com/2010/02/skill-ceilings-and-floors.html ) - they should be at the skill ceiling of their respective units to accurately judge absolute potential.

Anyway, I'm not going to dive into all that and restate everything - this article is made so that, if you ever do need to explain to a person how skill relates to a unit's potential, you can just refer them to this article.

On with the Metaphor!
To balance two units, you MUST measure their potential at their skill ceiling (or peak human skill if the skill ceiling is higher than the limits of human qualities).

To demonstrate this, I will be discussing the fine art of carpentry.

Firstly, imagine that instead of complex characters, the units in your game are simple tools, that you as a player, utilise to win the game.
This is exactly what they are. You unit in a game is a tool. Nothing more.

So, let's say we have two Hammers, the "Craptacular Nailpusher 101" and the "Super-Gold-Hammer Deluxe 5000". We'll assume that, as their names suggest, the 5000 model is far superior to the 101.

Let's say that "Balanced Hammer Co." wants to measure up the hammers against each other - as they are unsure of whether the Deluxe model really is any better.
So, they pick two carpenters from the local area, Tom and Dave to test out the hammers.
The thing is, Tom has been hammering for an odd 20 years now, and specialises with his Nailpusher 101, while Dave is completely new to the trade, and was frankly dropped a few too many times on his head while he was a baby.

The test begins, and Tom immediately takes the lead. He finishes the task deftly and with prowess, while Dave manages to hit himself in the face more than the nails he's supposed to be hammering.

The company records these findings, but obviously note that Dave was a really bad test subject, and so they discard the reports.

To relate this back to balance briefly, this shows that just because a unit is overpowered, it does not necessarily mean that new and unskilled players will immediately dominate skilled players with it. Even in unbalanced scenarios, a wide skill gap will still favour the more skilled player.
If it doesn't, then the balance issue should be so plainly obvious, it would have to be intentional.

So, back to the hammers.
Balanced Hammer Co. goes back to the drawing board.
What could they possibly use to test the hammers?
They search around, and find another carpenter - Adam, who has the same degree in hammery as Tom, and has been hammering for exactly the same number of days as Tom.
As far we can tell, Tom and Adam are both skilled at carpentry, and they are both about equal in terms of skill.

So, the competition starts off again, and, Tom, who now gets to try out the 5000 model, wins by a mere nail.
The company takes the results, and concludes that their new hammer isn't worth the 200% markup over the 101 model, and agrees that they must add more and more features to the 5000 model.
Seeing as Tom has been a willing participant in two tests, the company give him a free 5000 model as thanks - and guarantee him a discount on the model when they finish superpowering it.

So, Tom goes home, and, decides to use the hammer full time. It did win the competition, so why not?
Anyway, 3 weeks later, he's hammering some nails when he realises that on this particular model, the handle has a flat base that, on every 3rd swing, can be used to strike nails harder than usual.
Conducting his own test, he decides that this feature should be raised with the Hammer Co. company, to help them with their designs.

What does this show us?
Just because players aren't currently bunnyhopping with the Sniper class, headshotting every player they encounter with quickscoping, doesn't mean that the unit cannot do it. Players will often gain more and more skill until a large portion of a unit's potential is realised.
When judging balance, you cannot just use "good" players, you must go further, you must evaluate the unit's absolute potential.

Back to hammers.
So, Balanced Hammer Co. take this last lesson in their stride, but unfortunately they're still unsure of how they can measure up the two hammers.
But, one of their engineers has been doing some searching, and he finds the contact details of a couple of Romanian Hammer Brothers who are said to have such amazing skill in carpentry that they can immediately take one look at a hammer, work out every single trick with it, and utilise the hammer and all its features to a near perfect degree. The degree at which they can utilise these hammers is at the absolute limit of human hammering ability.

So, the company flies them in, and hands them each hammer.
The brothers take a quick moment to analyse the hammers, and the competition begins.

The result is that the brother with the Deluxe Super-Gold-Hammer 5000 wins by a land slide. He found many more tricks to this new hammer, than the other brother could find with the older one due to the new design's slick form, useful handle, and other features.
The brothers were completely equal in skill - and both were at the peak of human skill in hammering terms - so what happened?

As we started at the beginning, the Deluxe Hammer was simply better.

If we liken this to games, you can have a player at the skill ceiling of a Machine Gun class - but, if the machine gun class only deals 2 damage per second, while the others deal 100 damage per second, all that skill does not matter. The class is inferior to the others.

Conclusion
So, Balanced Hammer Co. choose these results over all others - why?
  • We are not comparing a bad player to a good player.
  • We are not comparing two equal players, that cannot utilise their units properly yet.
We are comparing two players, who use the units to their full (human) potential - and they demonstrate how powerful the units really are in comparison to each other.

This is how you evaluate balance.

I hope everyone enjoyed this post, and I want to try to follow this up with more skill vs. balance articles really soon.

Friday, March 18, 2011

AIE Sydney Game Design Competition!

Okay guys, I've been talking about this for the last year or so, but I've decided to finally get off my butt and set it all up.

Not all details are finalised, I'm just putting this up now so that everyone can get started.

In the spirit of the competition, I've decided to lay this info out as its own game pitch.

So, let's get started:


AIE Sydney Game Design Competition

Basic Concept:

Participants must submit a video pitch for a video game idea to be judged.

The video should be of 3 minutes or under in length.

The design can be any original design, for any style of video game.

The video should detail the concepts of the game, its tone, its gameplay, and etc.

The competition ends on 3/4/11.

The winners will be announced on (to be decided)


Gameplay (Specifics):

Information and updates on the competition will be displayed in my statuses, notes, and on my blog. (link below)

This competition is only for AIE Sydney students (first and second year)

No implementation is required, and any additional material will not be considered during judging.

Any video style is fine, whether it's full video, audio over animation, or a simple powerpoint.


Submission:

Videos must be uploaded to Youtube.

When the video is uploaded, post the link to my wall, or email it to alexander.cetinski@gmail.com

Submissions close on 3/4/11.

I will showcase submissions on my blog, and on my wall.

You can only submit one entry.

Multiple people can work on the same video, but the prizes are not negotiable or multipliable.

Submissions > 3:10 mins will be rejected. The extra :10 is to protect against youtube length issues.

Please include your year in the video description.


Win Conditions:

The competition will be judged by our amazing and wonderful game teachers - Conan, Dan, Epona, Matt and Mike.

The Judging Criteria Are:

- Concept

- Gameplay

- Fun Value

- Sellability


You will get a rating out of 10 for each criteria, which is averaged for your final score for that judge. Then the final scores are averaged for your over all mark.


Rewards:

Overall Awards

1st Prize: "Game Development Essentials: Gameplay Mechanics” - http://www.amazon.com/Game-Development-Essentials-Gameplay-Mechanics/dp/1418052698

2nd Prize: $25 in Steam games

3rd Prize: $15 in Steam games


Prizes are not negotiable or multipied for group submissions.


Year Specific Awards

First, Second, and Third Best Submissions will be announced for each year.

There are no year specific prizes however.


Additional Awards:

Best Original Protagonist

Best Original Antagonist

Best Game World Concept

Best Tone (Comedic, Suspenceful, Dramatic, etc.)


Joke Awards:

Most Likely to get Cease + Desist’d by Nintendo.

Most Likely to get a terrible movie adaption.

Most likely to require a Valve-Time release date.

Most likely to be refused classification in Australia.


-------------------------------------------


An example entry that won me a similar competition 2 years ago (2009):

http://www.youtube.com/watch?v=YL294g3svr0


So yeah, spread the word, and start with the entries!

I want to see lots of them. LOTS OF THEM.

This comp will be useful for getting your skills up to pitch for the major project coming up!

Saturday, December 25, 2010

Risky Design: Plain Sight

Introduction:
All games have design flaws. Some are major, some are minor. Some can be fixed, others can't without changing the game greatly. This is just a reality we all get used to. I doubt any game could have perfect gameplay, pacing, balance, or interface - but all great game designers aim to limit their game's flaw's impact on gameplay as much as they can - so they are barely noticeable.

An example of this would be for example that in many modern games and sports that are objective driven and focus on simultaneous attacking and defending (such as Capture the Flag), once a team has the lead - in some games it may be more effective to defend said lead, rather than attacking and trying to get a larger lead.
This is a flaw, yes, as part of the main idea behind Capture the Flag is balancing your offense and defence - but most games will just allow it - why?

There are very limited ways to actually fix it.
You could separate the offensive and defensive elements so that each team takes turns trying to get each other's flags (kind of how Rugby-styled football games do, or how Counter Strike's CS maps play out), but that's not capture the flag. It would in fact completely remove the balance that we want to achieve.

Another alternative is maybe buffing the losing team slightly so they can break through defences easier - but then you could run into balance issues, and teams being wary to get the first point in fear of a strong counter attack - which, could lead to stalling until the clock is very low and then trying to win.

So, Capture the Flag is flawed - but we all live with it and enjoy it still, because there are few ways in which the game style could be slightly tweaked to fix the issues surrounding it, without completely changing the game type. The problem is also fairly minor, as most teams will continue to pursue the flag even if they have the lead.

Still, there's room for improvement.

Moving On:
The point of this article is to talk about the game "Plain Sight".
Plain Sight is an amazing game developed by Beatnik Games, and the reason I'm writing thisarticle is because Plain Sight takes many features that have very easy to recognise flaws, and actually mashes them together to make for a greatly fun, near-seamless game style.

Before I dive into it, check out Plain Sight here: http://www.plainsightgame.com/
And I greatly recommend you buy it from Steam here: http://store.steampowered.com/app/49900/

Plain Sight uses the following flawed systems:
  • "Big Head" mode from Goldeneye.
  • Progression through death
  • Exponential power growth
  • Transparent player actions

"Big Head" Mechanics.
Explanation: In case you haven't played Big Head mode in any games such as Golden Eye, Unreal Tournament, etc. - it is a supplementary game style where, you still pick "Deathmatch", "Capture the Flag" or the like, but in Big Head mode, when you get a kill, your head gets larger so it's easier for people to headshot and kill you.
Pros for Gameplay: A fun silly mod for people to mess around in.
Cons: People can be reluctant to go for a large spree, and will sometimes commit suicide to make their head smaller.
In Plain Sight: Characters get larger and larger as they get kills, and as they do this, they can be targeted and tracked from a wider distance. Also, players who are larger yield more rewards for those that kill them, making them not only an easier kill, but a more sought after one.

Progression Through Death
Explanation: Progression through death is where a player feels that they gain more from relentlessly pursuing a goal and dying over and over, rather than trying to stay alive. Think of it like a player using a character with high area of effect damage running in a group of enemy players over and over to slow their progression, rather than taking them with skill and planning on as the game designers would intend.
Games can reward progression to death, they can punish it heavily, or any measure in between. It depends on the game.
Pros: The only time when death should be rewarded is where death is the objective - which is extremely rare. So, if "Death" is considered the failure condition, there are no pros to rewarding it.
Cons: Again, if death is the failure condition, it is bad to reward it.
In Plain Sight: Plain Sight is the first game I have played where death is the objective. In Plain Sight, you must gain energy by killing opponents (and getting bigger), which, then, when you commit suicide - gives you the fruits of your labour in your past life. Dying with a lot of energy (being a big fat robot), gives you far more points than dying with no energy.

Exponential Power Growth
Explanation: Exponential power growth is the term used to describe when players are rewarded for their achievements in the game, and so they are more effective, and are rewarded more. The cycle repeats until the game is unfair because one player (or team) is completely dominating with overwhelming power due to a couple of early achievements.
Pros: None. You never really want exponential power gain - though, in any game where you level up, it's a given. As a player gets stronger, they'll naturally be more capable when dealing with the same threats. Most games balance this out by making it so that over the scope of a 'round' or 'match', a player will never gain permanent advantages that an 'unlevelled' player cannot overcome with skill. For example, in Monday Night Combat, a non-upgraded player can still kill a completely upgraded one with skill.
Cons: An unbalanced game.
In Plain Sight: In Plain Sight, as a player gets kills and energy, and gets bigger, they also get stronger, they can jump higher, acquire targets at a larger range, their suicide detonation is bigger, etc. That said, they do not gain defence, so while they become a killing machine, it still only takes the regular dash attack to kill them. Also, as a player detonates with more energy, the points they gain is increased greatly. So, detonating after one kill yields next to nothing, while over 10 kills is a real game changer.

Transparent Player Actions:
Explanation: When a game makes player actions transparent, it means you can see and understand what the enemies are doing easily, and there is no real element of surprise.
Pros: Forces players to play in the open, without worrying about flanking, camping, stealth, etc. Depending on the style of the game, this can be good or bad.
Cons: In most games, removes strategic depth. Like it or not, camping, flanking, and sneaking are all valid strategies.
In Plain Sight: Players emit energy trails, that other players can follow to the player. Similarly, when players step on a part of the level, the ground lights up so that other players can see where the action is. The trails and ground light up more as you gain energy - so it's far easier to find more valuable targets.

How These Features All Come Together
So, Plain Sight implements all of these pretty flawed concepts, all in the one game - but it comes out great! Why?
They all cover each other's weaknesses.

The Big Head Mechanic and Exponential Power Gain mechanic work together to make for a very risky game style.
As you get more kills, you get more powerful, more deadly, but - you also get easier to kill.
Progression through Death allows the person to capitalize on the risks they have taken if they wish - or continue on, as the risk builds up higher and higher. As the risk builds, so does their reward if they are successful in committing suicide.
The Transparency of Player Actions is important as it increases the risk. An energy-filled player cannot hide and camp, they can only run and kill.

Conclusion:
The concepts above all work together perfectly to make a game style that is all about risk, greed, and chasing the most valuable players. And I love it.
It's great fun, it's an adrenaline rush whenever you have a long streak going and - I love how the game's gameplay almost seems to mirror its design. Beatnik Games took a gamble on mashing these features together, but it completely paid off.

I think this article is just saying that, well, first off, Plain Sight is designed masterfully - but also, that with enough thinking and creativity, you can find solutions to design problems anywhere, and even mash them together to fix each other.

Who knows? You could create a better Capture the Flag one day!

Friday, November 26, 2010

Lord of the Rings Online vs. World of Warcraft

Introduction:
Well, I finally got some inspiration for a post. I have been playing Lord of the Rings Online recently - well, for the last couple of days, and it was certainly an experience.
It's very similar to WoW in some ways, but in other ways, not as much. The main difference being that WoW is far better with pacing, and that's why I'm writing this post.

My thoughts on Lord of the Rings Online:
I originally had here a full run through of what happened during my playing experience, and how I felt about it - but it ended up as being extremely long, so I'll just make a simple list.

Things that surprised me in a good way:
  • It's a free to play game, with solid graphics, and stable gameplay.
  • Each character has a very unique play style, and the ones that I tried were all very strong when you got their core abilities.
  • A strong crafting system.
  • A lot of quests and instances.
  • The game gives you a quick taste of power from the start by giving you some Novice items, and a free Rental Mount (that lasts for 24 hours)
Things that surprised me in a bad way:
  • A lot of awful quests, with very few fun and rewarding ones.
  • In fact, too many quests full stop - leading to a grinding game style.
  • Sprawling zones full of lots of little towns, with lots of quests - it's impossible to decide what to do first, or at all. It's overwhelming.
  • Too much running, (often) not enough action.
  • The game has you completely strapped for cash in the beginning, as your expenses for gaining new class skills gets too high for your level.
  • An unstable GUI.
  • Many limits on even basic features without using Turbine currency - which can be paid for with real life currency. (Fair enough, the team has to make money, right? Still... it annoyed me a bit)

The main things I want to talk about ais the Pacing. So yes, expect me to be criticising LotRO a lot in this post - but, before I do, I will clarify that when it comes to free MMOs (that are legal - not private servers), Lord of the Rings Online is probably the best I have played. And I don't mean - "No monthly subscription", I mean, free all over - you don't even need to buy a serial key.
It's a great game, so if you don't want to be forced to pay for an MMO - I'd go straight for it.

Pacing Problems:
Lord of the Rings Online has some pacing issues as I described before - mainly that you're often running around aimlessly just trying to complete the crazy number of quests - and most of your quests are based on talking to people across the map. Then when you find them, you find 20 more quests.
It's overwhelming - badly structured, and as you run from place to place - somewhat boring. Towards the end of my recent play time, I spent most of my time with Auto-Run on, just avoiding monsters as I gazed up from MSN and saw I was running into them.

Similarly, every quest you get is often a part of a quest line - so it's hard to just skip one because you may have done 3/6 parts of the line, but you have not yet gotten your final reward. Similarly, it can be kind of infuriating to be finishing off a quest that has had you running around for an hour, and then talk to who you think is the final NPC only to have them send you back to the other side of the map again.

A map of what my time in "The Shire' probably looked like is as so:

Most quests are in red, the particularly annoying quests are in other colours.

Light Blue: You run around as a chicken, you cannot fight or anything - you simply have to run slowly from waypoint to waypoint. You have to do this twice-3x, with no difference really except for the waypoints.

Indigo: You have to talk to the man in the lower left, then the woman in the lower right, then the man in the top right, then the woman in the lower right, then the man in the top right, then go to the top left to find an item. Show it to the man in the top right, show it to the woman in the lower left, go back to the man in the top right, go with him to an instance in the top left, beat the instance, go to the man in the lower left, talk to the woman in the lower right, talk to the man in the top right.

And I'm not joking.

Yellow: You have to carry pies from all over the region to this one woman while avoiding particular hobbits (who steal pies). You usually have to go a long route, through enemy infested territory, and, while you're holding the pie, you can't attack or ride your temporary horse.

Green: Similar to the Pie-Carrying, you have to carry mail all around the region and avoid a different kind of hobbit that steal mail. Again, you cannot fight/ride a mount.

Pink: A crazy quest like Indigo - a lot of talking and running, with a tiny bit of fighting in the middle.

And as said, the Red are all the other quests - which, heavily outnumber the other colours. I think, if I had have truly represented Red, it would have been more of a smog than a bunch of lines.

Now, this would be... somewhat okay... if you could run from one end of the map to the other in about a minute. Well, unfortunately I timed it, and it took me nearly 10. So you can see how I spent about 15-20 hours in the last few days on this game, and only finished this area.

Now, you certainly have the option of doing the good quests - and then leaving the rest, but as a newbie - you cannot make the distinction, and, there are a lot of rewards for finishing a lot of quests in a region - so you feel that it's what you should do, while my friends have had a lot more success just picking and choosing.
It's also a lot hard to compartmentalise the quests when so many have nearly 10, or even more steps in them.

How World of Warcraft does it:
This is where WoW comes in. WoW has similar areas (where you start out as a newbie, get your professions, and leave the Novice level, choosing where you go next)
But, their areas are always (or at least, almost always) far better designed.
Let's take a look at how you'll traverse the starting area of "Durotar".


Your spawn point is the Purple X, then the green is your starting zone - for levels 1-6.
After that, you are asked to go to a little camp just outside the starting zone where you start getting some more quests - after you finish the key ones, (you don't have to finish them all to progress), you're asked to go to Razor Hill (Red), where the Yellow X denotes that you can set your teleporting (Hearth) stone to teleport there - which is useful when you've just finished many quests in the one area.
Once you've finished the key quests there - you're asked to go to Orgrimmar (Purple) again, you can set your Hearth Stone.

The key things to note is that all areas are somewhat self contained. The Starting area only ventures out when you've finished it. The Camp deals with the areas outside the starting zone and the Echo Islands. Razor Hill, being the main object of levelling inside the area covers the most ground - but even so, most quests are roundabout in the same area (meaning, you can complete them at the same time). Orgrimmar deals with the North.

Also note that all quests expand out, and then return to the single 'Town' in which you got them. They rarely go between towns - and all of your valid quests in the one area, meaning you can see them all rather than finding new ones everywhere.

I always like to praise WoW on pacing, but this is why - it's so easy to understand, it's well encapsulated - if you know the quests, you can do many of them at once to save on time. As the quests don't roam the country side, and aren't all quest chains 10 long, you can just do a couple, and then go to another area if you want - which is used a lot for speed levelling (example, if you hate the small camp outside the starting zone, you can travel to another starting zone, and finish a few quests there before returning).

Note:
At this point I'd like to recognise that a couple of weeks ago, The Shattering happening in World of Warcraft - and many zones were changed. And, Durotar is even faster, and designed better than the picture shown, if you would believe me.

Conclusion:
In closing, both World of Warcraft and Lord of the Rings Online are great MMOs - but, in order to be more competitive with World of Warcraft, LotRO has to work on its quest pacing and over all progression.
The most important factor with MMOs is getting players to play for long periods of time, so that their characters are strong enough to utilize end-game content, and when it comes down to it, I feel that for most players, LotRO is missing that gripping factor.

To improve, it should:
  • Cut down on the number of individual quests in each zone, so players are less bored by the end.
  • Shorten some quest chains, and split some up into individual quests so players can pick and choose easier.
  • Restrict quests' objectives, so, for the most part quests are all in a small area. If they do move to another area - prohibit them from moving backwards. Players should be able to draw a line from questing hotspot to hotspot.

This means that a player will likely feel like they are progressing faster, and they likely will be - as they won't be wasting time running from on end of the map to the other, and doing boring quests.

To put this into perspective, in 16 hours I can:
  • Get a character from 1 to 15 in Lord of the Rings Online, or
  • Get a character from 1 to 25 in World of Warcraft.

Which would players likely prefer?

Wednesday, October 27, 2010

If you've been wondering where I've been...


That explains it,
Been doing lots of work on Pokemon Universe!

I will make a real blog post soon!

Saturday, October 2, 2010

Power Creeping and Power Shifting

Introduction:
Well, I haven't posted in a couple of weeks - and I've been feeling guilty. That said, I've got a splitting headache, and I'm a bit angry... so we'll see how this goes.

Anyway, this post is about balancing to a certain level of power - if you have one character who is much better than everyone, do you nerf him? Or buff everyone else?
If you have half of your units being too strong for the other half - what then?

Finding a power level, then rebalancing your units to that power level is called Power Shifting.

What is a Power Creep?
"Power Creep" is the term used to describe one solution to the problem of balancing units - where idea is to buff all units to the best unit's level of power, meaning that the average power of units in the game increases (thus, 'creeping'). The term is usually used in a negative tone - to give the idea that the changes were poorly planned - though a Power Creep isn't necessarily a bad thing.

How should units be balanced?
The thing is, this depends entirely on the game, and the current balance situation.
It can't even be generalised, for example - say you had a roster of 34 characters, where one was generally more powerful than the rest - then you had maybe 5 or 6 under it, that could stand a chance against it. Assume that the rest of the cast are unusable against these 7 characters.

What should you do?
Most people would logically say "Nerf the 7 units, and leave the other 27 as they are", assuming the other 27 are on par with each other - but, this is not always the best solution.
What if the other 27 characters are ridiculously bland, and no fun to play because they have limited options and strategies in battle?
What if the 7 characters up the top of the list are the game's only saving grace, and, the existing metagame and matchups between these characters are amazingly fun to play?
Nerfing them would be a mistake - you're taking away all the fun.

So what's the secret?
Always have fun in mind when choosing what power level you're going to balance to.
There's no use nerfing all your characters to a boring, pathetic level - and similarly you shouldn't be buffing all of your characters to such a height that the game loses strategic depth in favour of button-mashing, camping to punish, sheer luck, etc.

With each game individually - you must think of which characters have the balance of reliability, power, and fun - then move characters up and down to their level. Hopefully, you'll have a good group of characters at this level who are balanced with each other - but, other times, you have to form the group unit by unit.

Monday, September 13, 2010

Speed is Opportunity.

Speed is Opportunity
If I were to ask you what's the most important stat in your favourite game, I'd get a range of answers depending on the game, and the playstyle of the person at hand.
But, chances are, after a bit of investigating, maths, and some analysis of the competitive format of the game - I'm 95% sure that with any given game, the correct answer would be "Speed".

Why Speed?
Because Speed is Opportunity.

Let's take the most common forms of speed, "Movement Speed" (mobility), and "Attack Speed".

Movement Speed
In every game, Movement Speed, and general mobility are extremely important.
Mobility is Control.

In an RTS, if you outmaneuver your opponent, it's often called Flanking.
In a Fighting game, it's Spacing.
In an FPS, it's Zoning. (Or gaining weapon advantage, flanking, spacing, etc.)

Whatever the tactic is called in each game, your ability to utilise the geometry of the playing area, and use the distance between you and your opponent is a domination tactic. It's controlling. It's fighting on your own terms.

Now, usually I resent people quoting Sun Tzu with regards to video games (it's overdone), I love his work otherwise - but really, the best way to express myself is through his quote:
"And therefore those skilled in war bring the enemy to the field of battle and are not brought there by him."

Which is precisely what I'm getting at.

Superior Mobility offers:
  • Control over the terms of the encounter.
  • Control over your enemy's options.
  • Greater ability to scout the enemy's strategy.
  • Superior options and opportunity to counter the enemy's attacks.
Attack Speed
Attack Speed is less important as mobility in most games, but don't get me wrong - it sure helps.
When you think about it, in most games, no-one is attacking without pause. Guns reload, characters guard, and attention is given to other priorities - so, nobody is really attacking at their highest speed, they are choosing to attack at designated times. With a higher attack speed, it's far easier to slot your attack into these times.

So, when you think about it, a character whose standard attack speed is a single attack in 0.2 seconds, will be able to shield until the time is right, and be shielded again after 0.2 seconds - giving his opponent a small window to attack him - in which case the opponent will usually have to attack through a hitbox.
Contrasted to a character who has an attack that takes an entire second to deploy - its opponent will have a much larger window to seize the opportunity to attack.

In this very simple case, it's obvious to see that the 0.2 second attack is superior, and, if these characters were to fight each other, the first could likely read the opponent's attack, and hit it 3-4 times in this window - which in most games means that the opponent's attack would be interrupted.

Think of it this way - every attack animation has foreswing, and backswing.
Foreswing is the lead-up to the attack, whether it be the charging of a strong attack, or the time between the input of the command, and the attack's full potential being reached.
Backswing is often the recoil from the attack. The sword going back into the sheath, the gun being prepared for the next shot.

The addition of these is what gives you the speed, or time of the attack. For example, the speed of an attack that has 0.5s Fore and Backswing, is the same as one that has 0.25 Fore, and 0.75 Back.

A lower Foreswing is mainly important for preventing your opponent from guarding against your attacks - if the punch comes out in less than 0.2 seconds, the opponent will generally have to input the guard command before they know that you've punched. In the same light, a move with a large foreswing (such as 0.8 seconds) can often be recognised, and shielded before the attack is even close to hitting.

A low Backswing is important to prevent the attack from being punished.
The act of punishing an attack is where an attack is blocked, or dodged, and the opponent seeks to attack back before the original attacker can regain control of their character. This is heavily aided by mobility, but simply having an attack with a lower Foreswing than the opponent's Backswing will do the trick.

So, a High Attack Speed:
  • Gives you the ability to punish attacks.
  • Gives your opponents little chance to punish yours.

So, what are you getting at?
You know the slow and heavy, but powerful units in games?
Time and analysis has shown that across many games, across many genres - they rank lower than faster units on average.

This is poor balance, carried by the widely-spread, flawed notion that "Slow but Powerful is equal to Fast but Weaker".
Due to the Control afforded by Mobility, and the Opportunities afforded by Attack Speed, this will likely never be true unless paired with higher resistance to all consequences of being attacked (such as flinching) and, at least some attacks that break the mold.

It seems that many developers are starting to see this, and that's great - but, they are still falling into the biggest trap, which, is continuing to give the units slow Movement Speed to compliment their slow Attack Speed.
To be blunt, this can rarely work, and is unnecessary.

To start with, have all your units start with the same movement speed, and then alter them individually as needed.

Summing Up:
  • Movement Speed gives one the ability to control the terms of engagement.
  • Attack Speed gives the opportunity to attack with less chance of being countered.
  • A larger Foreswing increases the chance of an attack being guarded.
  • A larger Backswing increases the chance of an attack being punished.
  • "Slow but Powerful" vs. "Faster but Weaker" is a false balance notion.
  • If a unit is slow in attacking, they will usually need to be compensated with Resistance to attack.
  • All units should start at the same Movement Speed, and they should be altered after a lot of testing and analysing.
  • If a unit has a slow Movement Speed, they will usually have fewer options against faster opponents.
  • Units with slower Movement Speed often have a lower skill ceiling, and lose the competitive edge in skilled play due to the previous point.