All posts


Game development








Hello! I'm Tom. I designed a game called Gunpoint, about rewiring things and punching people, and now I'm working on a new one called Heat Signature, about sneaking aboard randomly generated spaceships. Here's some more info on all the games I've worked on, here's the podcast I do, here are the videos I make on YouTube, here are some of the articles I wrote for PC Gamer, and here are two short stories I wrote for the Machine of Death collections.


By me. Uses Adaptive Images by Matt Wilcox.

Heat Signature’s Launch, And First Player Legend

A Leftfield Solution To An XCOM Disaster

Rewarding Creative Play Styles In Hitman

Postcards From Far Cry Primal

Solving XCOM’s Snowball Problem

Kill Zone And Bladestorm

An Idea For More Flexible Indie Game Awards

Teaching Heat Signature’s Ship Generator To Think In Sectors

What Works And Why: Multiple Routes In Deus Ex

Natural Numbers In Game Design

Naming Drugs Honestly In Big Pharma

Writing vs Programming

Let Me Show You How To Make A Game

New Heat Signature Video: Galaxies, Suction And Wrench-Throwing

What Works And Why: Nonlinear Storytelling In Her Story

My Idea For An ‘Unconventional Weapon’ Game

From Gunpoint To Heat Signature: A Narrative Journey

The Cost Of Simplifying Conversations In Videogames

What Works And Why: Invisible Inc

Our Super Game Jam Episode Is Out

What Works And Why: Sauron’s Army

Showing Heat Signature At Fantastic Arcade And EGX

What I’m Working On And What I’ve Done

The Formula For An Episode Of Murder, She Wrote

Heat Signature Needs An Artist And A Composer

Improving Heat Signature’s Randomly Generated Ships, Inside And Out

Gunpoint Patch: New Engine, Steam Workshop, And More

Distance: A Visual Short Story For The Space Cowboy Game Jam

Raising An Army Of Flying Dogs In The Magic Circle

Floating Point Is Out! And Free! On Steam! Watch A Trailer!

Drawing With Gravity In Floating Point

What’s Your Fault?

The Randomised Tactical Elegance Of Hoplite

Here I Am Being Interviewed By Steve Gaynor For Tone Control

Heat Signature: A Game About Sneaking Aboard Randomly Generated Spaceships

The Grappling Hook Game, Dev Log 6: The Accomplice

A Story Of Heroism In Alien Swarm

One Desperate Battle In FTL

To Hell And Back In Spelunky

Games Vs Story 2

Gunpoint Development Breakdown

Five Things I Learned About Game Criticism In Nine Years At PC Gamer

My Short Story For The Second Machine Of Death Collection

Not Being An Asshole In An Argument

Playing Skyrim With Nothing But Illusion

How Mainstream Games Butchered Themselves, And Why It’s My Fault

A Short Script For An Animated 60s Heist Movie

The Magical Logic Of Dark Messiah’s Boot

Arguing On The Internet

Shopstorm, A Spelunky Story

Why Are Stealth Games Cool?

E3’s Violence Overload, Versus Gaming’s Usual Violence Overload

The Suspicious Developments manifesto

GDC Talk: How To Explain Your Game To An Asshole

Listening To Your Sound Effects For Gunpoint

Understanding Your Brain

What Makes Games Good

A Story Of Plane Seats And Class

Deckard: Blade Runner, Moron

Avoiding Suspicion At The US Embassy

An Idea For A Better Open World Game

A Different Way To Level Up

How I Would Have Ended BioShock

My Script For A Team Fortress 2 Short About The Spy

Team Fortress 2 Unlockable Weapon Ideas

Don’t Make Me Play Football Manager

EVE’s Assassins And The Kill That Shocked A Galaxy

My Galactic Civilizations 2 War Diary

I Played Through Episode Two Holding A Goddamn Gnome

My Short Story For The Machine Of Death Collection

Blood Money And Sex

A Woman’s Life In Search Queries

First Night, Second Life

SWAT 4: The Movie Script

A Different Way To Level Up

Levelling up is pretty much the heart of RPGs, because it does these cool things:

  • Makes you feel like you’re achieving something by playing.
  • Gives you new abilities to try at well-paced intervals.
  • Lets you enjoy feeling more powerful than you used to be.

All this makes repetitive tasks feel worthwhile and even fun, which is particularly useful in a massively multiplayer game, because you don’t want players to get through all your content quickly, get bored and stop paying you a monthly fee.

But it has some problems:

  • It means players who’ve played for different amounts of time can’t play the same content together and still both progress.
  • It makes player-versus-player combat imbalanced unless strictly and artificially organised.
  • It can’t go on forever, and when it stops, even if there’s new content you haven’t seen, your game life feels empty.


All of these bother me, but the first in particular is absolutely ridiculous. If Tim and I are playing World of Warcraft at the same time, I can’t kill level 80 bears with him because I’d get slaughtered, and he can’t kill level 26 bears with me because he’d destroy the challenge and gain nothing in the process. The two activities are functionally almost identical, we don’t mind which of them we do or even if we do something completely different, but the game can provide absolutely no way for us to enjoyably play together. So I hate levelling.

Champions Online and City of Heroes get around this with a cool side-kicking system, where you can bump a friend up to your level for a bit. But it really just demonstrates that levels are meaningless anyway, and suspending them briefly shows how good life is without them. Champions has other level-related problems (I’ve run out of doable quests), and it’s that which got me thinking about what the perfect superhero MMORPG would be. This post is the first of a few about that.

What I’d like to see is a system where content – zones, quests, groups of enemies, bosses – has no level. It would work like this:

  • Whether I’m new or I’ve played for a hundred hours, a single monster or thug of a type I’ve never fought before is a serious threat to me. I can’t easily take on more than one at a time.
  • As I defeat more of this enemy type, I get better at fighting them. I start to do more damage to them, then learn to better dodge or block their modes of attack, then gain the ability to completely evade certain attacks or very quickly kill certain sub-types of enemies.
  • Each enemy group has a series of missions associated with it, usually culminating in defeating their boss. Once I’ve completed that, I can choose a new power.


Firstly, it means me and any friend can go to a zone neither of us have done and be on equal footing. Until between us we’ve done everything the game has to offer, there’ll always be some new challenge we can take on together, have fun and make progress.

Even if we go to one that one of us has made progress through, the newer player can still take on one enemy at a time effectively. Talking to roBurky about this, he suggested the newer player could just generate less ‘Threat’ – so even in a large brawl, only one or two enemies would go for him, the others would concentrate on the more dangerous player.

The second obvious benefit is that you can start anywhere in the world and the challenge will be appropriate to you. As well as the freedom that brings and the ease of joining friends with existing characters, it means that when you start a new character yourself, you can immediately be playing stuff you’ve never played before. Starting again is as fresh an experience as continuing. That’s particularly important in a superhero game, because there’s all sorts of fun stuff we can do with alternate characters made by the same player that I’ll get into.

The third is that all new areas, enemies and quests added to the game after launch are relevant to all players. That spectacularly increases the efficiency of content creation: every little thing you work on makes every player of the game happier and gives them more and more varied stuff to do.

The fourth is that it means anyone can fight anyone in PVP and have a chance. More experienced players will have a wider selection of powers, but late-game powers wouldn’t be outright better than your starting ones, just useful in different circumstances.

Aside from the problems fixed, it also builds on all three of the key reasons levelling is fun:

  • You’re making progress much faster, since a four-hour questline takes you from struggling with one dude to diving into huge crowds of them without fear.
  • Gaining new powers is still carefully paced, but now coincides with a major victory against a formidable opponent and the accomplishment of your quest. Rather than just spontaneously exploding in a sudden jump of progress when the fifteenth Spider Hatchling slain tips you over the edge to level 63.
  • And you’re always seeing how much more powerful you’ve become, because you regularly dive into mobs of enemies that were a problem individually not long ago. In most MMORPGs currently, there’s rarely any reason to take on enemies you used to struggle with.

The biggest potential problem with it is the notion of getting one new power per major questline completed: depending on the number of powers and zones, it might need to alternate between new powers and improvements to old ones. Adding new questlines in free updates seems like it could take as much work as raising the level cap on all classes, but whether that’s significant depends on how the end-game works, and that’s for another post.


Fede: It's an interesting take, but there would still be problems. For example, wouldn't it become a little too easy a bit too soon?

Richard: MMO difficulty really doesn't change that much between levels - you're artificially restricted from taking things on, and you have a few more skills to juggle as you progress, but it's only when you hit raid content that actual tactics usually kick in.

Besides, it's not a problem for other games. The Mario that beats Bowser at the end of the game is functionally the same one that gets killed by a Goomba on World 1-1 - the difference is in player skill. The Garrett that gets spotted in the first level of Thief is the same one that does a flawless run a couple of months later. There's no core reason MMOs can't be similar, except that historically they've been designed around making numbers get bigger.

Finalfrog: I'd love to see this system merged with some sort of reincarnation system, ala Kingdom on Loathing. So even when things become too easy because you've mastered all the current content, you can start over, gaining something you didn't have last time in the process. This would at least serve to mitigate the problem of diminishing difficulty Fede suggests.

Ideally this gain would not have a large effect on the combat so that things would be fresh again, but perhaps it would open up new areas or allow you travel between the existing ones in a more efficient manner.

Phill Cameron: This is a somewhat similar approach to what the Secret World seems to be doing; providing players with a flat rate of damage, etc. and just allowing them to broaden their repertoire when they get more powerful. So a beginning character is just as powerful as those who've played for months, it's just they don't have the same versatilitity. It's going to be a pretty interesting one.

Tom Lawrence: Once you hit level 20 in Guild Wars, which in the Factions and Nightfall versions is really just like saying "once you have completed a somewhat extended tutorial", it's a bit like this - the primary means of advancement beyond that point is in attaining new skills, which (broadly speaking) are alternatives rather than improvements.

Phydaux: I can imagine it working really well with WoW's phasing system too. As (in your system) a vet who's done an area, playing with a newbie, would have the same problem as you describe with the level 80 killing the level 26 bears.

Jace: This is an interesting concept - as with all Pentadact's ideas! - and were it ever to come into being, I would certainly be interested to try it. The aspect of not being able to play with your friends if they've been playing a bit longer has always bothered me about every MMO I've played.

SenatorPalpatine: I'm not a big RPG or MMORPG person , but with some of these ideas implemented I would definitely give them another try.

Also, Google Reader cuts off the post after "particularly useful in" right at the beginning. Is there some way I can fix this or is it a problem on your end? Not that I don't come here to read the comments half the time anyway, but I do like the whole post on Google Reader.

Simpunzle: Very interesting.

I would say instead of learning new powers after completing quests (unless that quest was to go learn a new power) as you kill thing your powers become more effective against that particular mob, just like you become more effective at killing it and resisting it's attacks.

This would also be interesting in PVP. The more times you meet an enemy player in battle the better you are at hurting them, BUT it should also work the other way. The more you get beaten by a player the more resistant you are to their powers.

I can see this working very well if executed correctly.

Simpunzle: It could also be interesting to have you suck at your abilities when you first learn them (maybe incorporate a bit of Eve's persistent learning while your offline kinda thing). Not only do you master the ability as you use it, making it more powerful, you also kill that specific mob more easily the more you fight them.

I could also see this working very well with an infamy system and factions with different characters.

Say I have my character and have been in this particular area fighting and slaughtering these mobs. They know me. They avoid me and only a large group or the tough guys will attack. Now I have a friend with who hasn't been here before. Fresh meat. Lowly mobs will attack the friend but the majority of them will attack me because they hate me now and I'm a much greater threat.

I could see an instance where you have generated enough threat/hatred from a given faction that as soon as one sees you and runs away to get help you would have the entire faction/dungeon/instance coming down on your head, all gunning for you.

It would truly be an epic fight that would require enough friends to keep you alive and distract some of the mob away from you.

Noc: Phill Cameron: It's a little different from a horizontal progression system, actually. This looks like it's still a very traditional, linear progression that's just been localized to specific areas and questlines.

It's a good idea, but as long as we're revamping the system, I figure we might as well go a bit further. I was pondering alternatives to this the other day, and I came up with something. It's a bit more suited for a broad, persistent PvP setting (as opposed to the PvE environment the article's written about), but I think it's still relevant:

. . .

First, scrap levels entirely and replace it with a skills-based system. You've got a collection of skills, each of which gives you one or more abilities. These abilities let you take advantage of situations you come across: your rank in the skill limits how much of that situation you can take advantage of.

So, suppose you've got a "Backstabby" skill that gives you a damage bonus if your target's embroiled in fighting multiple enemies. Your extra damage is determined by the number of distractions, with more distraction letting you get in more damage. But instead of multiple ranks in the skill increasing the bonus you get from distractions, your rank would just limit how many distractions you can take advantage of. So everyone would get the same bonus from one or three or five or ten distractions, but someone with Backstabby II might only be able to take advantage from two distractions, while Backstabby IV might let you take advantage of six.

Skills get increasingly more expensive to raise as they get higher, but a very high skill is only useful in very specific circumstances. Players can snag a few ranks in the skill without too much trouble, which gives them a useful utility. Specializing, though, lets them be very effective, but only in the right circumstances - so team play would be about setting up those situations.

An old character who's out of their element, and denied circumstances they can take advantage of, is just as vulnerable as a new character without any skills at all.

. . .

Secondly: you raise skills by using them, Elder Scrolls style. This seems problematic and grindy, since no one wants to sit there using the same skill over and over and over again to improve it. But there're some twists that make this easier.

You improve a skill by using it. You improve the skill faster by using it in combat, under fire. You improve the skill faster by using it against something or someone better at defending against it. You improve the skill faster by doing things where your skill caps out; Backstabby with too many distractions, or doing really difficult shots with Snipey, or such.

You ALSO improve a skill by seeing other people use it. You improve the skill faster by seeing someone who's very good at it use it. You improve the skill faster by seeing other people use it in a combat situation. You improve the skill faster by having it used on YOU.

The benefit you gain from using a skill (or watching it be used by) specific characters or enemy types slacks off after a bit, so you improve the skill faster by using it against different opponents.

All of these multipliers stack which each other. So the quickest way to improve your skills is to go out and find a bunch of folks with a similar playing style and go take on some challenges. Since early ranks of skills train pretty quickly, you can become useful to the team (or competent on your own) pretty quickly, and still have longer term goals to work for.

There are a few other advantages, too, such as that if you're consistently getting crushed in PvP by some jerks spamming the same tactic, you'll quickly start to acquire the tools to remove their advantage.

. . .

So for our established goals:

- It makes you feel like you're achieving something by playing, both in a character improvement way and a player skill way as you start to get better at your chosen tactics set.

- It doesn't give you wholly new abilities at intervals, but acquiring competence levels in skills (i.e. reaching the point where the skill is high enough to take advantage of the most common situations) does open up new gameplay strategies.

- Lets you enjoy feeling more powerful than you used to. (See goal #1)

- It lets - and even encourages - players to pursue the same content regardless of character skill levels. Underskilled players are still useful against tougher enemies, and will quickly reach a point where they can help out significantly.

- There's likely always going to be a lot to improve. The nature of the system means that you'll reach a point where a given skill covers most of what you need it to and will take a long time to improve to the next rank. So once you reach competence, if you're having fun you can keep doing that, and get even better at it, or you can experiment with different play styles that rely on different skills.

Obviously the gameplay itself has to be interesting to support this: since skill progression is a matter of utility instead of game progress, it's going to have to hook you in by being fun instead of pulling the "Constantly dangling carrot" trick. But I think the idea has potential.

Dave_C: Hmmm. ''Mathematics: The MMO''

Cptn.Average: I think it's interesting how this could lead to world events that everyone can take part in, not just top level players.

Devenger: Intriguing. Whether players would really feel a sense of progress when they are largely discarding prior superiority whenever they enter a new area is unclear. That might be countered by an interesting selection of powers, especially if they promote teamplay, but that's a challenge.

I've greatly enjoyed MMOs like PlanetSide that had only subtle elements of characters getting explicitly better, as opposed to just gaining new options. I tried the Champions Online beta having not played a more traditional MMO since the City of Heroes beta (so it's not a field I pretend to know well), and just found levels frustrating. Often, I knew a few rocket jumps away was an area I couldn't fight in; this made my lower-level objectives feel less worthwhile and very artificial. Teamplay was fun, but teams seemed few and far between (even compared to the CoH beta) partially because of not great numbers of players in one area being the same level. And don't get me started on how terrible other levelled features seemed to be (crafting, I'm looking at you).

So, yeah, I think we do need something clever like you're saying in this post.

Pod!: Patent this idea.

Lack_26: I'm still a fan of the GW approach, getting new skills and trying new builds interested me far more than just levelling up in that game. And two players at different levels (till your about 7 levels distant) could easily join together and take a area between the two players level and still have some fun.

Joe Snuffy: I think where Guild Wars went right was the shift from being incredibly high level to having the right balance of skills. Since the max level is reached relatively easy, you spent the rest of the game moving in and out skills and around to counter the threat.

David Sahlin: Well said. Grats on the critical distance.

Peterd102: This is great for the very occasionaly players wanting to join with the players who have played for a while. But not from the other way around. To sit there having done the same style 'quest' many times and have a large array of skills. But in raw data, still doing the same dmg as someone who has just joined AND having to deal with more challenges if the threat idea was implemented too. It would be disheartening to them. You cant get a sense of progression, and have a fair playing feild between the highest levels and those who have just started.

You would allow the long players and new people to play together, but would lose the long players altogether.

Jason L: Dealing with 'more challenges' is a reward. Actually, the reward. Being more awesome and crucial than other people is not disheartening. I'll tell you what's disheartening: .5% loot drops. Those are pretty disheartening.

Dagda: Some other alternatives that can address the issuess raised, off the top of my head:
-Benefits are applied to a different level of gameplay so as to not significantly affect current level's odds of victory. PvP battles with cutlasses and pistols yield gold for kills, which can only be used to upgrade your ship for the naval skirmishes.
-Your abilities cost points which must be accrued during the course of a scene. Higher-powered abilities cost more and accrue points at a much slower rate. Spend 5 killpoints and you spawn with a rocket launcher, which only earns killpoints when enemies are "splattered" by direct hits.
-Abilities can be "looted" by enemies following some eventual failure on the wielder's part (death or failing to guard vs a Disarm attack), causing a trickle-down balancing effect. The aforementioned rocket launcher gets 10 shots, drops on death, and has the ammo count go up by 2 whenever someone picks it up for the first time. Alternatively, items abilities can be freely lent to team members.
-Based on comparative level, you are either "stronger", "weaker", or "even" with regards to others; gameplay type is changed based on these three states (say, "stronger" enemy attacks deplete your guard but incur less penalty for losing) but ignores the degree to which you're stronger/weaker.
-Players each contribute a small number of abilities to a pool usable by the entire group.
-Abilities improve performance vs NPCs only.
-Abilities improve performance of NPC allies, who will always retain vulnerabilities that allow a PC to take them down in large numbers.

Combine as needed.

Peterd102: If your more awesome and crucial than others, then yes it would be a reward to have more challenges. But then you avoid the whole point of changing the system to allow new players an equal footing, there just not. Additionally .5% loot drops are only disheartening when there the only thing that could drop. When theres a normal reward to go with it, you forget the .5% drop and if it does drop, its fantastic.

My main point is it may be impossible to have progression for those who put time and/or effort into doing something (gaining more power to take on greater challenges) and have new players on an equal footing with the long players.

DoctorDisaster: I thought a contextual leveling system like this might work the first time I heard about Warhammer Online's Tome of Pwn (or whatever it's called). Having almost everything you do work toward "achievements" that confer in-game bonuses is a leveling system in itself, if you properly tie the bonuses to the tasks required to get them.

I think the key is that you shouldn't just perform better here or against this mob; there should be a whole array of different stacking bonuses that you can accrue. Your character gets to be familiar with certain environments, mobs, and weapons, but also learns to defend against certain attack types, use particular combat styles, and so on.

Some of these buffs ought to nest within each other. You'd have a weak general buff for defending from melee attacks, a more specific and slightly stronger bonus against attack types (slashing, piercing, etc), and then a very strong focused bonus against particular weapons or weapon types.

Eventually, your effectiveness would be plotted on a big Venn diagram of layered combat bonuses. So when you venture off to new areas, it's not like you're a complete novice: you'll still get bonuses from equipment you choose and slight buffs against attacks you're familiar with. You just won't stack up well against someone very familiar with the terrain.

This also provides a big social incentive. When you enter a new area, you'll want to make friends with a "native" who can handle the challenges there. If you already have a friend who's been playing, you'll want to visit them on their home turf so that you can check out some high-"level" content earlier on.

This would also make PvP much more of a mind game, as you try to guess based on a character's race, class, and equipment which types of attacks they would be less familiar with. It's not just a matter of "well I'm a rogue so I better use a dagger" -- you'll be trying to judge what weapons, magic elements, etc you're familiar with and your opponent isn't.

Jason L: Questing/beating up the right NPCs for intel, aptitude or both on a nemesis, perhaps? A PC nemesis, even?

Tom Francis: Cheers for all the thoughts. To respond to a couple of paraphrased points:

If the powers you gain are worth gaining, they'll make you more effective and the game will get easier: this is basically true. If you earn a big area-of-effect attack, even if its average damage-per-second isn't more than your blasting power, either it makes dealing with crowds easier or it doesn't. So to a small and subtle extent, enemies need to pile on more experienced heroes in greater numbers. Running to fetch help is a good example - I found out recently that Champions baddies do this if you're in a group of three or more. Here it could be as simple as small, early mobs only attacking newbies one or two at a time, but all pounding on more experienced players at once.

I'm not totally on board with Doc's idea of accumulating stacking buffs, though, I think that could end up with too large a power difference between new and experienced players, and I might start feeling useless when I pair up with my eight-hours-a-day hardcore friend.

If you're not getting outright better, you won't feel a sense of progress: Partly addressed above, in that inevitably you will get slightly better. Partly fair, on some level no amount of trickery is going to give an RPG like this the same progression feel that current RPGs have. But I don't entirely want it to, I don't want the player to have to buy into this idea that a level 80 unarmoured mage has three hundred times the hitpoints of a level 1 warrior. But I would want to make the player feel a sense of achievement for completing a new area. A genuinely useful new power is a huge reward already, but in a superhero game I'd also want a jail where you can visit your vanquished heroes, random missions unique to players who've completed certain zones, badges that mark them out as a 'Thug Killer' or something, unique costume pieces related to the gangs they've defeated, cameos by characters they've saved in future quests - basically everything short of a straight stat boost. I'm not averse to giving them levels, either: a hero who's unlocked ten powers can be called a Level 10 hero to make him feel like he's progressed. You just have to be careful not to give the enemies a level, or it'd start to get confusing.

Noc: Your system has a nice and natural way to level up. I'd want to know about more of the skills on offer and how they'd progress. It sounds like the benefits of levelling them up could get so niche as to be not very desirable. If you level them up fastest by using them on people most resistant to them, it sort of encourages you to spend the game using all your powers wrong.

Some of the ideas here come up in more depth in the other two posts I'm half-writing, I'll try to get one up this week.

Cunzy1 1: The main problem with the systems you suggest is the emphasis on skill (beyond being able to press keys in a relatively quick and specific sequence) over time invested in the game. This small change means that MMORPGs would probably become less accessible to vast swathes of the current audience.

Tom Francis: There's less emphasis on time invested, but no more on skill. Just like current MMORPGs, you pick your difficulty level by how quickly you choose to power through a set of quests: slowly to improve your abilities before taking on the hard stuff if you're finding it challenging, or quickly if it's easy for you. In fact, you can adjust that more quickly and easily than in normal MMORPGs, because you improve against an enemy type with each one you kill, rather than having to grind hundreds before you level up.

Noc: A bit late coming back to this, but:

@Pentadact: The powers definitely become more niche at higher ranks, since they'll necessarily reach a point where the benefit from that extra rank outweighs the ever-growing cost of leveling them up.

I don't think this is a problem, though. This system's intended to complement genuinely entertaining gameplay; the idea is that once you've sort of settled into a niche you like, you'll keep using it because it's fun and effective, or because it's useful in the pursuit of other gameplay goals. But as you keep playing in that niche the skill will train quietly in the background, and at periodic intervals you'll be surprised with that extra little rank.

The "using the skill on people most resistant to them" bit was probably a little badly worded. The idea is that you'll learn more by swordfighting master swordsmen, or by taking harder shots with your ranged weapon, or such. Not that you'll be able to level up your Marksmanship skill really fast by repeatedly shooting at a tank with your slingshot.

. . .

And the nature of the skills themselves would totally make a huge difference in how this plays out! I actually had something of a rules set in mind for this - the above plan came from working on a (much smaller scale) RPG/wargame thing with EVE in the background, and pondering how that project would translate into a massively multiplayer environment.

It's squad and point-buy based, but the cost/benefit balancing issues still amount to something similar. I don't want to babble all about it here, but for an idea of the sort of things I had in mind:

- Marksmanship skills, where you get an accuracy bonus the longer you spend aiming at a target, with your rank in the skill capping the maximum bonus. But you have to stay still and looking at your target while you're aiming, and are vulnerable for the duration.

- Teamwork skills, where you get bonuses to attack or defense or such based on the number of adjacent people doing the same thing, with your rank in the skill capping the maximum bonus. The downside is that you're necessarily running about in the open in a tightly clustered blob, which leaves you all collectively vulnerable to people firing into the crowd or lobbing AOEs or similar.

- Dueling skills, making you more effective against single targets but leaving you extremely vulnerable to being blindsided by other opponents.

. . . and similar. So more ranks in the skill let you take more time to aim, and run around in larger formations, and batter more on more isolated targets . . . at the cost of having to take more time to aim, and being stuck in a larger blob, and needing your targets to be more isolated.

And stuff like that.

DoctorDisaster: That's a good point about disparity between a new character and an experienced one in its element. I hadn't really thought through the social end of things as well as I should have. I still think it would work, but I didn't explain it well at all.

You'd start by joining your experienced friend in an area where he's got all his stacking buffs active and you've got nothing, and since you're a total newb, that gives you a little leeway to figure out the basics. But then the two of you would venture outside that comfort zone he'd established. As he wandered into new areas and came up against new baddies and so forth, his experience bonuses would go away layer by layer and the new guy would take on more responsibilities and challenges.

This allows for a ramp-up in how much a total newbie would have to carry his weight. It also means there wouldn't be a stark dividing line between areas of expertise and areas of incompetence for experienced characters. It has a few side bonuses, too, like making exploration a social activity rather than just a timewaster for isolated completionists.

Also, I should have mentioned that when I talk about buffs and bonuses, I'm talking about something completely different from the usual disproportionate nonsense that separates players by level. Maybe someone in a completely unfamiliar situation does one third of their potential damage and suffers full health loss, while someone at their absolute peak reverses those proportions. Enough to where there's a definite difference, sure; but a handful of neophytes can be just as effective as a single expert.

KaBob799: I think that one way to fix the usual problem with levels would be to decrease the difference between levels. Lets say you had a game where you start out at level 1, and the maximum level was 100. At level 1, you would be able to do 1-2 damage, but 99% of the time you would do 1 damage. At level 100, you would still only be able to do 1-2 damage, but 99% of the time you would do 2 damage. Obviously this is just a simplification of the idea, but the point is that the newest player would only be half as strong as the strongest player possible. Any medium level player could easily do content with a high level player because the differences between them would be so small that only an extremely powerful boss creature would cause problems. Newbies would be able to do high level content, but would not be able to do it as long or as effectively as a high level.

Tom Francis: That certainly gets around it. But if it takes a few hundred hours to become twice as good as you were, you'd probably sacrifice that feeling of constant improvement, and the pleasure of getting more powerful than you used to be.

snowyowl: I have a question. What of the sense of achievement players get for passing certain milestones? I play a lot of single-player games, so perhaps I'm not the right person to talk about this, but it seems that a considerable amount of satisfaction comes from passing major milestones that are marked in-game. In WoW for example, this would be reaching level 80. In Team Fortress 2, getting a rare hat (it shows that you have played for a few hundred hours). Even RuneScape has achievements delimited into Easy, Medium and Hard, where completing any section gives you a new ability.
My point is that, to make such a game be fun, the player must have an overarching goal he wants to achieve; a reason to keep playing the game. Rather like a story needs a climax to build up to.

It could be: "complete all the quests". But once all the quests are gone (which could take only a few weeks), what does the player do then? It could be: "gain every ability". But if this were possible, it would defeat the entire purpose of your game, not to mention involving a lot of grinding.
Giving out minor rewards (that might be visible to other players) for maxing out a given skill could be a good compromise. But once you have maxed out your best skill, you are then forced to play with skills that you enjoy less.

My point is that levels are a quantifiable way of measuring a player's skill and progress within the game, and that the player gets a lot of satisfaction from increasing these numbers. Unless you find some way to replace this, the player will have no goal in-game.

PS: Some sort of achievement list, perhaps? Achievements for discovering secret areas, using a certain power in a certain context, for playing the game in all possible modes, for killing/collecting/befriending/casting/using X of a certian item, for displays of remarkable skill, etc...
Having little similarity between achievements would make them fun to get.

Tom Francis: You'd still have a notional 'level', it just wouldn't translate to a massive power differential over someone with a lower figure. Max level would just be the number of content chunks in the world at launch. No objection to achievements, but the ultimate goal is always going to be getting every power your character can get.

Trevor: Maybe the different powers you can gain will be balanced out with others?

All powers have upsides, all powers have downsides?