Tidalis Beta 1.017: Improvements To Combo Scoring At Game Over

[LaughingThesaurus]

Ahh, I see, so by lightweight you actually meant in terms of the workload? That's just elegant, there. If you're a developer capable of adding a lot of gameplay with little work and still have it be fun, varied, and challenging, that's pretty darn elegant. All of those game modes were really what I was looking at when I referred to all of the content. I remember Tidalis being one of those games that was particularly hard to learn for me, I remember the modes and stuff, but I never got good enough to get into the campaign or puzzles. There's a point where the campaign just like locks you out of doing anything until you're like MLG level at match-3-with-a-twist.

But yeah, the thing I guess is that the gamemodes didn't have all of the complexity that monsters do. Each monster has a lot of art assets, careful tuning of animations and attacks, don't even mention fighting with balance. Game modes are much more abstract. The blocks themselves don't really attack either. They're more like rules that move around and flow with the gameplay. Or, in the case of turning on the items, basically convert the game into Super Smash Bros Brawl.

Regardless, that kind of stuff is really cool, and adding that sort of element is something that I would want to do in any game I had a part of. I'm referring to some more abstract things that change the way that the game works, like the game modes in Tidalis. I only particularly see advantages in doing that. Is there any real disadvantage to adding replay value in that way? I guess, maybe for a team that isn't as great at game design, it might end up being harder.

[x4000]

Yep, a lot of the game modes were added near the end of the development cycle.  And easily half of the 20ish modes took me under an hour each to code.  A couple were 20 minutes or less.  Whereas a typical monster would individually take me 3-5 hours, plus the time for the artists.  Whoo, expensive!

In terms of variety by adding small tweaks and so forth, the main big one that we've found is that it is overwhelming to new players.  Players were overwhelmed by both AI War and Tidalis in terms of the number of options.  It didn't harm AI War at all because of its target audience.  But for both Skyward Collapse and Exodus, we're going more with an unlockable route -- as you play, you unlock more options which can make it more complex.  Also, Skyward Collapse in particular has some ridiculously complex things you can do if you're in to really metagaming the game systems, but these are things that would not even occur to (or be needed for) a novice player.  I love that sort of thing in particular, because it uses minimal interface options to still allow for tons of flexibility.

Anyhow, yeah -- to keep using the monsters analogy, we'd also generally with these games do more like 20 monsters instead of 125, but then make each of them way more involved and interesting.  Not that that's directly applicable to either of the games we're doing now, but that's the general metaphorical idea.

Anyway, thanks.   We're pretty stoked about how this is going, too, and are excited to share more soon.

[LaughingThesaurus]

I think I would approach new players with nice presets instead. I've had some ideas for games cooked up in my mind for a while, and for people who don't like mucking around in menus, that's often a decent option. You've got the premade classes in shooters, you kind of have startup scripts in AI War, stuff like that. The optimize command in older RPGs is a good example as well. I don't think I could live without creating a great degree of options.

As far as the monster thing goes actually, were you comparing the monsters to the gamemodes? Because, just thinking really quick about Valley, you have the 20-30ish monsters in this game-- not many, for those who want visual variety after playing for 40+ hours like I have. But, you also can give each monster tons of attacks and little variations, which can make the same monster function differently if you encounter it at a different time or on a different world. Seems to be more in line with what you're talking about, like the gamemodes that interact with each other in unique ways.

Also I would totally pay for a gamemode that worked more like Meteos, because I rock at that game. Alas, it probably won't happen.

[x4000]

In terms of presets, yeah that's another thing.  And that probably will be something we do more of.  But basically, when it comes to the first playthrough, that has to be kind of carefully guarded unless you want players picking something odd and having a terrible time.  If the option is there, some people will do it.  And possibly some reviewers.

That said, having a cheat code or similar to unlock the content if you lose your data or something isn't something I'm opposed to.  The point isn't gating players arbitrarily, it's providing a sensible progression for someone playing the first time.  If they don't want that and want to enter some sort of code to bypass it... well, that's probably a power player to begin with, and that's fine.

The main problem with even presets before a first game is that you're asking players to make decisions about mechanics they don't even understand yet.  Once they have at least a basic understanding with how the game plays, that really makes a big difference in how they approach future options.

For Skyward Collapse in particular, those games are meant to be shorter, on the order of 40ish minutes by default (though there are options for less and more).  But there are a bunch of carrots for repeated play and mastering the finer points, etc.  So that helps some with the replay thing, when it's not "play 13 hours first."

[LaughingThesaurus]

Cheat code is definitely a really good option... and you do bring up a good point. There are people who click the horrible buttons anyway-- I've seen it. Only example sticking out in my mind is actually with Terraria. Players opted for the "Mediumcore" setting which is certainly the middle-most option but is certainly not the way the game was meant to be played. So, I've had to steer people away from that before.

I also am thinking about Bastion as a good example of what you mentioned about "making decisions about mechanics the player doesn't understand". The game does not have difficulty settings-- it offers a few gods to invoke that make the game harder by giving enemies extra abilities. This option is only given to you several levels through the game. You do, however, get it in full immediately on a new game + or score attack mode save. That's okay, because those are unlocked after completely finishing the game. However, even when you first gain access to the shrine, you can't use any idol except the "Doubles enemy attack and movement speed" one... which is totally the easiest to understand aside from the "enemies have more health" one. So, I guess the unlockable route is the Bastion-like route, and I'm actually okay with that.
I can see how too many options is overwhelming. I generally can just look past them until I understand them, but there are people that just don't want to or decide not to (like the poor souls who start games on max difficulty). I'm already thinking of ways to incorporate that kind of idea in any of my game ideas.
...admittedly, not an easy thing to do when one of the core ideas is complete customization of what your projectile attacks do.

[x4000]

That sounds like a pretty cool idea, though.

And it really depends on target audience.  The crazy complex AI War interface isn't a problem because it's aimed at grognards and grognards only.  "I am the hardest of the hardcore" is the immediate message, and that is a happy message for them, not a scary one.  It has a niche and it works.  It sounds like your RPG idea is also coming into that sort of niche.  AI War is our best-earning game, by the way, so niche does not equal bad by any stretch.

Regarding Skyward Collapse and Exodus, they both have aspects that can cater to the grognards, but also aspects that make them more surface-accessible.  That's one of our philosophical shifts here, really.  Tidalis was too far in the extreme of being casual-friendly, but it wasn't really casual-friendly (was still hardcore and overwhelming), even though it looked casual-friendly (thus turning off the hardcore puzzle fans if they didn't know what it was already).  We're really trying to hit more of the middle ground, which actually I think both of the Valley games and to some extent Shattered Haven all did.

[LaughingThesaurus]

Ooh, how could you tell it was a bit of an RPG?
The plan I really had was 'there's a prebuilt class or three with loads of attacks that just get unlocked as you get stronger rather than you putting together your own as you get stronger'. That way, you could totally go the full on custom way, which is really the route it was meant to go... or could experience the game with a 'just give me stuff' button. I'd love to work on this actually, and get it going. I have ideas on implementation and everything, the problem is that there's potentially a whole lot of multiplayer focus. Without being able to handle netcode yet, it's not exactly a viable idea to go bringing in a co-op intensive game. Let's just say I am having a hard time exercising self-control over how awesome the multiplayer plan hopefully is.

What I've actually noticed, is that the middle ground you're hitting almost looks like the SNES style. They're like, games that are really easy to just start playing, but really hard to master all of the little complexities. In fact, that ties back in with the ideas of unlockables. Sometimes, you didn't just start with hard mode. Zelda introduces one dungeon item at a time. Contra welcomes you to try as much as you want, but you need mastery to succeed and unlock the harder levels with cooler guns.
Tidalis basically sounds like Bejeweled; Overwhelming Edition, right? Looks and sounds so happy and fun and simple and "Wait what are streams? What do you mean THAT'S how you combo?" I can tell you for a fact that the combo system was met with that reaction in my head.

[x4000]

Cheers. And yes, the NES and SNES, and to a lesser extent Atari and 80s/90s PC gaming, were my biggest childhood influences. I enjoy later stuff, but not with the childlike sense of wonder I had before.

In terms of my intuiting rpg, I'm pretty sure you said that! Sounds like a fun idea. Netcode is generally very hard, but if you have sufficiently few agents you could use something off the shelf like Unity 3D's built in netcode. They handle all the internals for you, but we've found that having thousands of entities is not the best thing for our particular pipeline. The way we draw graphics and handle game logic is very engine-independent by design, but that creates some bridge/adapter problems for us. If you're just going to use vanilla unity plus probably one of the 2d packages that are in the unity asset store, then you'd have an easier time.

[LaughingThesaurus]

I actually don't remember saying RPG, so maybe I was more excited than I thought. Although, often times the easiest way to intuit something is for it to be outright said to you.
The netcode has to take into account the basic idea of 'projectiles enhance each other if they come into contact'. So, that may involve some pretty precise detection, potentially a lot of entities. I'm not actually very good with the technical side of things at all, I was hoping for a lot of enemies to bash through rather than a few very tanky ones, but the 'lots of enemies' thing is a sacrifice that can be made for the sake of making it actually work better.
This is a bit of a different topic for another day, but I wonder if games these days really can evoke that same sense of wonder. For some, probably not. It's almost as if more games are geared towards more adults rather than for 'all age audiences'... and also stick with more down to earth concepts rather than 'man in green tights only hope for seven maidens fights giant pig'. There are still magical games, but the market's in a different place.

[x4000]

It's entirely possible that I just heard a few details about character rolls and went "oh rpg."

Regarding the netcode and the spell reactivity, the best way I can think of to handle that is via something like the ai war netcode. Ie, you can't directly move your guy around, all you can slow point and click like diablo, and there is a tiny (subsecond) delay. You could have oodles of enemies and spell with the local gpu and CPU being you only limitation in that case. Bandwidth would only be dependent on the number of orders given by players, which would be trivial. The ai could be deterministically simulated on each client with fixed seed random number generators.

Of course I'd you wanted something more immediate than diablo, where you can directly walk around, you need more traditional-action-game netcode. That is more of a bear, to me. It's an exercise in reconciliation of constant desyncs rather than trying to stay in sync. All fps games work this way, as does any other action game where you directly move around with no sub-second hesitations.

Anyhow, yeah: as the number of agents goes up so too does the challenge of making it, but obviously I have a lot of handy proof sitting around that it can be done.

Hope that's useful.

[LaughingThesaurus]

That actually sounds like something doable. It'd almost function a bit more like Magicka, keep the keyboard open for lots of shortcuts and stuff.
AI War works by the host handling pretty much all of the calculation and the clients all are fed what happens right? Or, was that just how the AI worked?

[x4000]

That's just how the AI works, and that's just because it's so intensive for the higher-level functioning of the AI.  Basically, its "conscious thought."  All of the instincts and low-level decision making on both sides (what ships do when not given explicit orders by you or the AI player, etc) is handled in lock-step on each client.