Game Theory Discussion - Today's topic: skill ceiling

[YCobb]

You just pasted a definition that applies to a strategy, not even a quality of a game. I'm hardly even sure what you're arguing anymore because you can't seem to stay consistent on whether or not trial and error is a thing players do or a thing games do. ((hint: it's the former))

It's a useless term, you're absolutely right. You're also absolutely right that it applies in bodaciously every situation where a player makes mistakes. You don't fix that by changing its meaning, you fix that by finding a better term. ((Hint: 'poor game design'))

If games require players to employ trial and error, they're not necessarily bad. You can't throw around absolutes like that in a thread about something as nuanced as game theory and expect to be taken seriously.

[wordNumber]

Please stop arguing about semantics. I know I did it too but it's been a half of a page now. It's time to move on.

If we need a new topic, what do you guys think of, I dunno, hub-worlds. What makes a good hub-world? Do you guys think hub-worlds should be should be like levels but without an objective, or straightforward safe zones? Or do you think hub-worlds should just be forgone entirely?

[Exeres]

I am dumb, please explain to me what a hubworld game is.

[Shad]

Think Mario 64's castle or Mario Galaxy's observatory

[wordNumber]

You know the castle from Super Mario 64?
There you go.

[Exeres]

I have been outed as one of those people that never played Super Mario 64.

Eternal shame.

[YCobb]

It's probably just because it was my first game, but I think that Sonic Aventure handled hub worlds pretty well. If they're going to be in a game, I prefer that they fit the game.

Stage selects work as well, though, and so do pure progressions, so it doesn't matter too much.

I don't think the area between stages matters until it becomes an unfun hindrance to stage entry, like in Sonic '06.

[wordNumber]

Well I dunno, I kinda like when Hubs have neat little secrets in them, and encourage exploration. I mentioned the SM64 castle already but It really is one of my favorite hubs, because there's so much extra shit hidden in places you might not look. Granted, I didn't realize Rainbow Ride existed for years and that's kind of a problem because it's a major stage and it probably shouldn't be that hidden, but I still like when you can explore a hub and find things.

Hubs are neat because you can add a bit of exploration to a more linear game. You can also change the hub around to reflect story progression and stuff. If a hub isn't much more then just a room to connect levels that's fine but there is a lot you can do with it to make it matter rather then the baseline being "just don't suck."

[YCobb]

I agree completely. I meant to add something about how Sonic Adventure's hubs were a bit boring, but I got distracted with my digression about alternatives.

I hate to stick with Sonic games here, but Unleashed fixed this (but then took several steps back) by packing some of the areas (especially Spagonia) with rails and pathways on buildings.
Too bad it ruined the benefit by making you A] choose your country/area/region B] pass through a hub world C] into another hub world D] with pretty unimaginative challenges before you can finally play the stage.

You could skip this later with an alternative set of menus after the first menu, but that's still more menus than I like.

[Ablu2]

Hub worlds are cool, I like Donkey Kong 64s hubworld, it's kinda the same style as the Mario 64 hub world except it had a little more stuff to it than the mario 64 hub world.

Are we talking about basically a way to choose which level to go to without the use of a menu, or does the definition expand to fit a place where a lot of gameplay occurs too, like in the Jak or Sly Cooper series?

[Shoolis]

I think hubworlds can be done well or terrible, either by introducing difficulty into an area designed to help you relax or by being just generally boring

the hubworld in the Portal 2 multiplayer was really good in my opinion because it was an area to unwind in between the crazy shit in that game, but gives you a chance to warm up when going to the next level by having you solve a simple puzzle before starting a level set.

I wouldn't call the Banjo Kazooie hub world that great though because there are Things That Hurt You in between levels and just a general blandness in the design in my opinion

[Galaxy Man]

Hub worlds should either be very big with a lot to see and do, but with a way to go between each point very quickly, or really tiny.
Super Mario 64's hubworld is kind of a poor example of a hubworld. There isn't a whole lot to do, and while there's a lot of stuff packed into it, it's too time consuming to get around. Galaxy kind of had the same problem but I do remember a way to get between every observatory easily.

Depending on the game, you could probably get away with large areas with not a lot in them. Shadow of the Colossus did that superbly. The entire world, the whole landmass, is just completely empty and devoid of anything important but the bosses, but since it's all entirely and completely optional to explore and see, with no rewards, it's suddenly very interesting and well made. If you were forced to explore every inch of the world, it'd probably be hated.

[Shoolis]

I think the Paper Mario series did hub towns pretty well as well. I would say the first game did the hub town the best and then petered off as time went on (Flipside + Flopside = I want to die) but Toad Town was very effectively laid out and had multiple attractions that made consecutive visits more substantial than "okay let me restock on items and get back to the story." Final Fantasy has that problem quite a bit, though I have less experience with that series than others.

[Brekkjern]

4 pages of discussion and I didn't see this thread before now

A few things I have noticed playing games featuring AIs is that the "I" is lacking. And honestly, it doesn't matter that the I is lacking, more so that even the perception of intelligence is lacking. Through browsing around Reddit and other places, discussing game design, I have noticed that people hold FEAR and Halo as shining examples of good AI design. I have played through FEAR and I Halo a whole lot. I can say the AI isn't all that intelligent, but they manage to create the illusion of it. In FEAR, the enemies will shout to their allies and ask for help, order them to flank or just generally comment on the situation. They are much more than mindless goons that get in your way like the ones in Call of Duty.

The same goes for the Halo AI. The grunts, elites and brutes have very different behaviour. They even react to situations differently. If you manage to sneak up on a grunt it will jump and scream when it notices you. The brutes and elites will usually shout. Grunts run away when the elites fall, while the elites and brutes stay their ground. The enemies spread out wide to attack you, but they stay in tight clusters working together.

Yet this seems to be a fairly uncommon thing. I can only come up with those two examples off the top of my head. Why isn't this more common? It can't be all that difficult to make some sorts of personalities for the AIs and make some behavioural patterns that differ from all the other games created the last 20 years?

[DoNotDelete]

I like it when you're provided a representation of a geometrically accurate six-sided dice that can somehow randomly generate numbers between one and ten (as in some of the Mario Party games) - of course this is impossible in real-world terms.

I've often wondered about the randomly decided/generated shapes present in games like Tetris - I would guess those games are probably dependent upon user input to generate the random numbers/shapes but I can't be sure.