MagnusWootton said:
I always thought if the coins did more than just give u an extra life in mario maybe there would be more purpose to collect them.

Good example leading to a common uncertainty i have…

Say we add shops to the world map, and you can buy power ups for the coins.
(That's bad, since you don't need power ups on the world map, and also you would need to watch out count stay below 100, but just for the sake of an example.)

I can imagine two options how this would feel to the player:

Player A feels annoyed from the option. It stresses him out he now needs to decide if he wants lifes or power ups. He's annoyed from the additional complexity, which he thinks isn't needed.

Player B likes the new option. He likes the thrill of the gamble, potentially regretting his decision if it turns out he would better have taken the life instead the power up.

Can we say something like ‘Action game players are mostly of the kind X’?

Probably not. I play Super Mario daily with my wife and i'm clearly A but she is B.

The problem is: Running out of new mechanics, all we can do is extending classic and simple concepts with new mechanics like this. So we increase complexity with time, no matter if we really want or not. E.g. like currently with adding RPG mechanics to pretty much every game.
That's a real dilemma to me. Making a science out of game design does not help, and seemingly makes it only worse.

Nice to hear from you Joe! My first reply!

I guess what I said is pretty thoroughfare for games that come out, just take any action game and u can turn it into an MMO! haha, and it might not be so enticing for some people.

But its always been my dream to make one, John Romero did the big MMO thing, But I never was quite good enough a game programmer to be as professional as that and actually get one finished and running.

But yes, I may find I dont get anyone playing it if I finally finish it and I'll learn another life lesson.

JoeJ said:
The problem is: Running out of new mechanics, all we can do is extending classic and simple concepts with new mechanics like this. So we increase complexity with time, no matter if we really want or not. E.g. like currently with adding RPG mechanics to pretty much every game. That's a real dilemma to me. Making a science out of game design does not help, and seemingly makes it only worse.

That's part of a long-developed part of the science aspect of game design.

Consider two axes, one axis is width, the other is depth. You get wider the more features you add, feature A, feature B, feature C. You get deeper the more you can do with the feature, the more variation it provides.

If you implement a feature and it only does one thing, that's adding width but not increasing depth. You get a single command and a single action. Open an item, that's it. Swing a sword, that's it. You can add depth by making it do multiple things. For example, instead of just “swing a sword” you can get five or ten varieties of swing based on how you take the action.

A good example of this recently was the Zelda Tears of the Kingdom Fuse mechanic. Simple mechanic, fuse two items together. But it's very deep, there are lists with thousands of combinations. Many are boring and seemingly obvious: attach a spear to a spear and you get a spear with extra long reach. Some are a little weird, a fan on a shield can push away enemies, a spring on a shield can bounce them away, which gets interesting on ice. Bee honey on a sword or shield summons bees to help you, weird but whatever. Adding fire, ice, and electric elemental to weapons an parry, that's good. The Ancient Blade can one-shot any enemy, including most boss-tier monsters, but breaks after a single use. Light Dragon parts fused will heal you each time you use the item, plus do extra holy damage. It's a narrow mechanic to fuse any two items but very deep as spreadsheets with a hundred thousand combinations. Creative players can create amazing airships with auto-targeting missile launchers raining death from above using only a few simple mechanics.

Another example is alchemy in Nethack. Dip anything into a potion, or mix potions with other potions. Simple A+B mechanic, narrow, but deep gameplay. Simple is making a stack of water potions and dipping with holy water or unholy water, the entire stack becomes holy or unholy water. Healing + speed = extra healing, extra healing + gain energy = potion of full healing. full healing + gain energy = gain ability, booze + potion of enlightenment = potion of sickness, levitation + enlightenment can give gain level, or random. Dip in oil to waterproof or rustproof certain things, a potion of acid can corrode some thing, can dissolve iron bars blocking the way, can remove grease, and so on.

Portal is a great example. In Portal 1 you have a single mechanic - make a hole between two places. They explore a bunch of uses, a narrow mechanic but very deep in exploration. In Portal 2 you have the portal gun plus 3 types of gel. Splash repulsion gel on a turret at it bounces around. Gel bombs can reach unexpected areas, fling and double-fling to launch places, conversion gel to reach exploitable places. Very narrow and easy to understand, yet at the same time very deep when combined with the rest of the environment.

As a series, LoZ has always made particularly good use of “narrow but deep” design, and they look to be doing it again with Echoes of Wisdom.

Other games go wide. Magic the Gathering in particular goes quite wide with about 300 mechanics overall. They do a good job of balance within each set, and mechanics gradually drift to extended sets, then legacy, and a small number were restricted or banned in sanctioned play due to unexpected combinations. Many of them come in families, for example a bunch of “look at the next few cards” like Scry, Surveil, Explore, Hideaway, Clash, and Fateseal, yet each is a variant. The game is also deep because there seem to be endless ways to combine abilities to a wide variety of effects some are comically explosive, like creating thousands of token creatures, near-infinite buffs, but generally relatively balanced, powerful combinations along with counters to each combination which you might or might not have in hand. The explosive combinations usually are a 3-effect combo, giving an opportunity for opponents to dismantle the combo when they see it coming or to overwhelm the opponent before they can trigger it.

I agree with the less is more idea.

Need something that cooks, then dont overdo it with extensions.

frob said:
A good example of this recently was the Zelda Tears of the Kingdom Fuse mechanic.

Yeah, that's good. I don't have a Switch, but looks great.
I guess they came up with it after playing some Garys Mod and Minecraft, and then they said with a big smile in their faces: ‘Hey, let's put this in a real game. It's fun.’

Then the game becomes successful and influential. So we want to analyze and discuss how it works, to adopt some ideas. We talk about systemic environments, creative gamepley, sandboxes, etc.
After that we see similar features in many games and it may become a trend.

At this point the university professors of the game design academy chime in. They pick up the mechanics and terms, analyze how they work together, paint Link on their chalkboard and teach it to their students.

The students are now equipped with all the skills they need. They know all the recipes, and they even know how to merge and combine them, with scientific precision.

But it does not really work. It's never as much fun as it was in the games they had on the chalkboard.
Why? I think the reason is: Fun is no science. There is no point to teach it. Either you have some idea with a big smile on your face, or not.

They better teach how to make ideas work on a computer, but not so much how to develop the ideas.
The fact that game design academies exist really looks like the industry taking themselves too serious. But fun is the opposite of being serious. : )

frob said:
If you implement a feature and it only does one thing, that's adding width but not increasing depth. You get a single command and a single action. Open an item, that's it. Swing a sword, that's it. You can add depth by making it do multiple things. For example, instead of just “swing a sword” you can get five or ten varieties of swing based on how you take the action. A good example of this recently was the Zelda Tears of the Kingdom Fuse mechanic.

I forgot to reply regarding the dilemma itself.
The answer you gave is very good, pointing out what we can do.

But we can not avoid the dilemma. What i mean is, taking the Zelda Fuse as example, we can enable a rich set of options with just one simple mechanic of combining things. Yes.

But to make the mechanic work, we have to address a big problem: We need an interface to assemble two 3D objects in 3D space. So we need to control 6 degrees of freedom using 2D controllers and 2D displays, just to move one object around.
It won't be super intuitive. Players need to learn how the interface works, and likely it feels clumsy, restricted, or complicated to many.
To help out, we may add some snapping anchors, so pieces fit together like Lego. So we make it easier to use by making it even more complex / restricted.
(Really thinking about Garys Mod here and just assuming Zelda is somewhat similar.)

To enable the simple mechanic we want, we do add a lot of complexity to the game, and the game becomes less accessible. The players need to build up and maintain more knowledge about the games functions, making the game less casual. The simplicity of the primary idea does not prevent this, but there is a cost.

Yes, most often there is a cost to development. Done well it is a small addition, such as a single tag added to each item.

In good systems there are typical values and presets so designers don't have to think about it much, just pick a preset from a drop down list. Then they can specialize if desired.

In more complex systems there are enormous data tables that have all the combinations hard coded, in a massive m*n grid.

3D models already often have a concept of a mount point or attachment point, which modelers use to say which point and orientation connects when an object is held or worn. Adding points for the new mechanic is the same task, a tag on every item, which has a cost but is not overall too burdensome. The manipulation code can find the nearest matching points on the model and use them for whatever binding or attaching or slotting that needs to happen.

frob said:
Yes, most often there is a cost to development.

Sure, but that's solvable with doing the work.
I talk about the player experience. Some example:

Playing Portal or a similar game, attempting to build a staircase from a crate and a barrel.
It's fickely and clumsy, because we can not rotate the object in our hand.
Eventually the avatar holds the barrel at a 45 degrees angle, but i want it straight upwards.
I often try to push the barrel against a wall to make it pointing upwards.
It works, but as soon as i move it away from the wall, the dump game rotates it back to the unpractical 45 degrees angle it had when i've picked it up.

Arrrgh - barrel constantly rolls down the box - fury and rage quit. Playing Super Mario again. >:(
This happens really often to me.

The problem is: There is no intuitive way to control 3D orientation, so games rather try to use a default orientation / up vector for each object. Or they keep it in the initial orientation it had when picked up.
To deal with the limitations, players have to do clumsy maneuvers. Which would look totally silly in the real world, so it works only in first person if at all. Mount and attachment is some band aid, but we restrict player freedom, and likely we tone down the feature as much as possible so the issues do not dominate.
That's why there is no crafting and assembling in the 3D game world, but it happens in GUI pages, turning it into an abstract and bolt on mechanic.

The dilemma is: We could implement all this cool 3D crafting stuff in our simulations.
But we can not expose it to the player with a convenient and intuitive interface, so we better don't do it at all.
(I'd really need to look into Zelda to see how they did it exactly, but i just assume it's too fickely / restricted to become a mainstream feature.)

We can not solve this with software. We would need better controllers, e.g. like VR controllers.
But VR controllers are not really good for couch / desktop. You constantly need to lift your hands up so they are in empty space. You can not rest your hand conveniently on the table. So this won't become mainstream.

Still - there should be more innovation regarding controllers. Consoles need at least a way to make Mouse Look work.
But sadly player habits / adoption seems a big risk and problem, so Nintendos innovations mostly remain a gimmick, although they were highly succesful.

You know, SSDs and raytracing does not really allow us to come up with new games, so we're stuck.
But new controllers would.

JoeJ said:

there should be more innovation regarding controllers. Consoles need at least a way to make Mouse Look work.

Don't the analogue sticks surfice for a mouse? - I know it takes a little getting used to, but I bet alot of people are adept at it by now.

MagnusWootton said:
Don't the analogue sticks surfice for a mouse? - I know it takes a little getting used to, but I bet alot of people are adept at it by now.

No, analog sticks can not replace a mouse.

Technically it's because the stick has limited range, so you can not turn the camera around for multiple revolutions if you want. The hack to solve this is degrading the stick into a button once you reach it's boundary, so you'll keep turning at constant velocity after that point, lacking control over angular speed.

Ergonomically - and that's the real problem - it's because a stick applies a force to go back to center. So if you want to adjust it to some off center coordinates, you constantly have to counteract the force. Thus your' hands are not relaxed but tensed, which is discomfort.

The mouse avoids all this. Just move it to the target coordinates and let go. No tension, no problem with keeping the exact coordinates.
It also has unlimited range, as we can just lift the mouse up, move back, then putting down and keep moving in the same direction to infinity. And we keep full control over speed at any time other than the lifting.

Notice: Because mouse is so comfortable to use, the need to lift it up is not a problem. We even do it subconsciously and automatically.
Operating a analogue stick requires constant focus and force, and if precision matters (they mostly make sure it doesn't), it is a problem and task requiring conscious effort. Which then reminds you constantly that you currently play a game, so you can not feel like being in the game. Gamepeds kill immersion very effectively.
I think this proofs a mouse is a extremely good input device.

Hehe, there are two evils within our rows, attempting to destroy video games:

1. Establishing a need for 2000$ GPUs.
2. Gamepads.

The latter is actually worse. The former is only more fun to rant about.
PC players migrating to gamepads must be stopped at any cost!
Console players might not care so much. Their screen is too far away, so they don't feel like in the game anyway.