Gungrounds

[Gungrounds]

Hi guys, I'm a founder of a small game development studio.
It is called Gungrounds, and it is dedicated to making quality 2D shooter games inspired by titles such as Geometry Wars 2, Metal slug, Resogun and Nuclear Throne.


I do most of work in Gungrounds on my own except art and music. This is why I like running things on personal level and pitch my ideas to others in hope for constructive feedback.
I hope it will be interesting to you to join the discussion, and help me in my goal to make good shooter games.



Here are two upcoming games I work on.

ROCKING PILOT In close future, one country's army develops an elite fighting helicopter, featuring sharp rotor blades capable of slicing tanks like butter during breakfast. They also install an artificial intelligence in its mainframe, which learned to fight by watching hundreds of 80's action movies. The player takes control of this flying fortress as it breaks through dozens of levels and faces countless foes, ferociously advancing towards the final challenge.



Rocking Pilot has been approved on Steam Greenlight, after which it was ported to Unity Engine. Together with Mad Head Games, we are improving the game’s graphics for final version.




SPITFIRE INFERNO Have you ever wondered what happens to history’s most evil individuals once they end up in hell? The devil challenges them with hordes of demons, enjoying the shooting mayhem which occurs. The game is imagined as a roguelike shmup, and players are able to choose among several historical villains, including Atilla the Hun and Napoleon.




Recently we started working with company Mad Head Games. They are helping us by upscaling art for Rocking Pilot, to make the game look good on modern HD devices.




Both games were presented on Reboot Infogamer and Reboot Develop Conferences, as well as several other gaming-oriented gatherings. We are planning to release them on Steam and current generation of consoles. Once we complete them, we already have new projects in plan.




If you have any ideas or questions, feel free to ask them. I will post soon some game design questions.


Facebook: https://www.facebook.com/Gungrounds
Web: http://www.gungrounds.com

Rocking Pilot Greenlight Trailer https://www.youtube.com/watch?v=B2AvPv9-UBk

[Gungrounds]

Few days ago, I added some new power-ups and a novel combo system. When enemy is destroyed, it spawns shards that, once collected, raise the combo value. Once ten of these ‘combo-shards’ are collected, a new power-up item appears in the arena.
However, if player fails to collect the shard, or doesn’t deal damage to enemies in specific timeframe, the combo bonus lowers to zero. It’s simple and quite intuitive.

I’m having balance issues, though. Some levels have a lot of enemies, causing too many power-ups to appear in the game. How can this be solved? Any ideas? How is this usually solved in shooter games?

You can see it in action here: https://www.youtube.com/watch?v=vBsOn2CFbTs

[Gungrounds]

Also, it would be interesting to hear which power-ups systems you recommend.

I hope all this doesn't sound like some company doing marketing. I have been making 2D shooter games for last 3-4 years now. This is how I spend most of my days, but I didn't bother posting much before because it wasn't good enough. Though, after I started getting feedback from other devs and players on the conferences, I realized I have made huge mistake, and if I was showing my games earlier, the quality could progress faster.

I would really like to discuss design, share demo versions of the game here, and iterate the game by using your feedback. :thumbs up:

[masterfrog]

I kind of don't get what the game is about.

It looks like you are a helicopter, but also kind of seems like you're a big spinning ninja shuriken? And a bunch of stuff is flying around and you are shooting at it, but what's the story? I care a lot about story so a game with no story at all, or where I can't really see any kind of theme present is unappealing to me. It definitely looks like something I would describe as a generic shooting game, but with flashy graphics and lots of visual effects.

I kind of don't like the non-scrolling background, I thought all games were scrolling these days. I get a feeling I have seen in a lot of games... where you're not sure what exactly the game is all about, but there's a lot of stuff going on and you're quite serious about it all.

Very polished presentation, overall. Maybe playing it would feel different, I am just kind of getting a visual first-impression. Sorry if feels get hurt, since I am the first one to respond.

[Lord Satori]

I love a good story as much as anyone, but there's no need to force one to the forefront.

It'd be nice if there was a little bit of backstory/scenario, but there's nothing wrong with focusing more on gameplay.

That said, you can clearly see in one of the screenshots that there is someone speaking to you in a text box, so it's not like there isn't one at all. It's just not very present in his current posts is all.

After all, what matters most for games like this is gameplay. A compelling story is a great addition, but presentation wise, it isn't something to focus on primarily, at least for this particular genre.

[mystran]

I’m having balance issues, though. Some levels have a lot of enemies, causing too many power-ups to appear in the game. How can this be solved? Any ideas? How is this usually solved in shooter games?

I ain't no professional dev but "theoretical game design" is kinda my thing (I like to design games "in theory"), so I'll share my thoughts anyway:

I think the standard solution to "too many power-ups" (at least in case of weapon powerups) is basically to cap the upgrades to 4-5 levels or something, maybe another 3 for a secondary weapon if there is one. If there are more than that, then the power difference between the higher levels should probably be mostly visual... and I'm pretty sure there are shooters where the visually 10 times more powerful max-power does less than twice damage (although typically gets a lot more coverage) compared to the miserable peashooter without upgrades.

In games where you can switch between different types of weapons the power-ups then serve for the purpose of switching, otherwise they typically become just another form of score items. Then you balance the power-up gains in the early game, and late game you assume player is at max power (or should power up fast if he's not there yet) and dropping a lot of score items is fine.

The other typical solution, quite common in arena shooters (and others, isn't this what Metal Slug series did?) is to basically just have one power-level, but drop special weapons that are limited ammo (or time, but limited ammo feels more "fair" as it won't run out when there's an idle segment, since you can stop shooting). Then getting more power-ups doesn't increase your power (beyond having some special weapon, which are better than default), it simply lets you chain the weapons more reliably (pickup another as soon as one runs out of ammo), so you ideally never have to fall back to your backup gun.

A curious variant is found in Aero Fighters / Sonic Wings series (at least 2 and 3; there's probably other games that do this, but I can't think of any right now) where power-ups up to the second highest level are permanent, where as max-power is ammo(?) limited. So once you reach the max-power point, the power-ups become resources you can (try to) time to parts where you want a little extra firepower or chain them by spending one while letting the next one bounce around on the screen before picking it up... or you can pick them up in groups to get some score, and in bonus levels they just spam them at you like score items, because that's what they are if you're already maxed.

Basically, the problematic combination is when you have a deep linear upgrade path and highly variable power gain. If you keep the power-growth in short, finite bounds, then too many power-ups shouldn't cause any long term damage to your game balance.

While not directly related, it might be also helpful to think of how RPGs usually make the deep "linear" path work by instead controlling the power gain. First they let you slowdown and grind for additional XP/items if you fall behind. This leads to the opposite problem where grinding too long might make the player too powerful, so usually you either see exponential XP requirements (getting levels from low level mobs becomes infeasible and if you're behind a lot you'll quickly get a few levels to get you closer to intended power), no XP at all for stuff too much below your level (time to move on), or most of the power shifted to items (so you must progress to get a chance for better loot). All these essentially do the same thing, namely "soft" cap the maximum power at any given point through the game (ie. many short upgrade chains glued together), relative to whatever content the player is going through at the time.

YMMV, but think about the actual power-growth, it doesn't need to progress directly in relation to the number of power-ups picked up. You could balance the number of power-ups, but you could also balance the power-up system so the number of power-ups isn't too important.

[Gungrounds]

Thanks for the responses.

I wanted to first write a short introduction of our projects.

Narrative - we don't have much finished in this area, that is why I didn't go into details, but there is gonna be a cool story centered around the AI that learned to fight by watching 80's action movies. This AI ( player ) controls the advanced helicopter. This helicopter has special blades which are capable to slash soldiers, tanks, and planes into pieces, and in some situation make him GODlike. The nation that constructed him to destroy enemy nations wants to shut it down because it is not suitable to live in our "normal" world. You can expect lots of comedy and over the top situations.

Like Lord Satori said, the focus in both games is gameplay.

[Gungrounds]

@ mystran

Thanks a lot, really interesting post! Currently, the power-ups make the player stronger ( increase his attacking power by 2x ) for the short amount of time. If player is not shooting, this time is not reduced. More precise description of problem is this: if there is double amount of enemies on the screen, they will drop double more power-ups, and the power of power-ups is so big that the player will not need to collect them all to clear the screen which results in player having power-ups all the time. ( even though they disappear after few seconds if not collected ) .. On the other hand, if I make them too weak, this is a problem in levels where there is not many enemies because it is harder to hit the target, so you don't gain so much from it. With 1 power-up you don't kill as many enemies. So, maybe I should just tweak more the damage factor to make this work.



I will try to explain the basic mechanics of "Rocking Pilot". All this is still work in progress.
If any of you would like to give it a scan, give me any feedback on how I could make it better, I'd love to get your input.




0) PROPELLERS

A) NORMAL MODE
- they kill instantly any enemy soldier in range
- they inflict damage to other enemy units ( tanks, planes etc. ), but player must be careful to not come too close and crash ( lose life )

B) GOD MODE
While player holds the special button, propellers extend and switch into GOD mode.

Benefits of this mode:
- larger close combat radius ( any enemies in range will receive amount of DMG per second )
- immortality - player cannot be damaged
- increased movement speed - player can change his position to gain the advantage
- deflects all enemy bullets back into enemies
- deflects gas ( poison, smoke, fire etc. )

Basically in this mode player can win in any situation.

Limitation:
- it lasts only certain amount of time, few seconds
- it can be activated only if the helicopter temperature is below 25%

While the propellers are spread, he is immortal, and he inflicts huge damage to all enemies in range. While this mode is active, the helicopter overheats, and once the temperature is too big, the attack mode is switched off.

Player must reduce that heat by shooting it out through his weapon.
Because of this player needs to switch very often between this two modes.



1) SCORE

Player increases score each time he kills an enemy. This score is multiplied by current combo amount.



2) PROGRESSION

At end of the level, player is rewarded with variable number of stars. Number of stars is proportional to score.
For example, in first level, player needs to score 5000 pts to receive 1 star, 15000 pts to receive 2 stars, and 30000 pts to receive 3 stars.
In each world there is around 5-8 levels. To unlock next world, player needs to have certain number of stars.

Score is a requirement to progress. Thus, it is super important.



3) COMBO SHARDS

When enemy is killed, it will drop few collectible shards ( pickups ). For each shard collected, the combo will raise by 1.
If not collected, combo shards disappear from the screen after x seconds.

Challenge: how fast can player kill the enemies, how much time he needs to spend to move his position in range of shards, and most important - in how much risk he gets involved by collecting those shards.

Neither of this is currently visible enough.



4) RISK & COMBO

I think the most interesting balance is : "to increase the combo, you need to risk" which translates into "to get the higher score, you need to risk, which rewards you with progress ( new levels and power-ups )".

Few possible solutions:
1) Enemies come in groups so combo shards will always be close to the other enemies.
2) Most often, spawned shards are near the edges of the arena. New spawned enemies will attack this area.

On paper this sounds great, but it doesn't work so well in practice.



5) COMBO TIMEOUT DROP

If player doesn't collect combo shard in x seconds, the combo will drop to 1. This time can be extended by dealing damage to enemies, either by shooting or propellers.



6) POWER-UP

For each X combo shards collected, new power-up is spawned, 300px away from the player, towards the center of the arena.

When collected, player deals 2x more damage. It lasts around 1-2 seconds and is reduced only while player is shooting.
If player collects power-up while another power-up is already active, that power-up will increase in power by 20%.

There are some other benefits and variations, but it would be too much for this post.

Purpose: reducing enemy count + magical moment that feels good.



7) ENEMY SPAWNING ALGORITHM

To make the power-ups and combo system work I think it is crucial that this is done really well.
Since I just added power-ups and combo system recently, I don't have much done in this area at the moment.

I am also not really satisfied with the current spawning system, so it would help a lot to hear suggestions.

To make the power-ups really work, I think it is required that there is almost no cap on the max number of enemies. This way, player would use power-ups to reduce that count to something he can manage with his normal weapon.


OTHER STUFF

Weapon has an infinite amount of ammo.

Player has 3 lives, one life is reduced when player receives damage. There is no common way to recover those lives, they are reseted on level restart.

Controls : twin stick or WASD + mouse.

[mystran]

4) RISK & COMBO

I think the most interesting balance is : "to increase the combo, you need to risk" which translates into "to get the higher score, you need to risk, which rewards you with progress ( new levels and power-ups )".

Few possible solutions:
1) Enemies come in groups so combo shards will always be close to the other enemies.
2) Most often, spawned shards are near the edges of the arena. New spawned enemies will attack this area.

On paper this sounds great, but it doesn't work so well in practice.

The "in practice" is probably something that really kinda would require playing to really understand why it doesn't feel right. Also do you want aggressive "kill everything fast" style or "slow and steady to keep the chain" type gameplay?

For aggressive style, you could always award more value for players close to the target on kill, either by dropping higher value shards if player is close, or by having the shards drop in value if not collected fast enough (eg. Futari "glowing gems"). These two encourage slightly different game play (in terms of how they encourage the player to move), but the basic idea is the same.

You could also just limit the number of shards on screen (either oldest one disappears if you spawn more or can't spawn more once the limit is reached) to encourage the player to pick them up without keeping too many around. Or spawn lower value shards if there is already too many high value ones on screen.

Or you could put new enemies on a timer, encouraging the player to try to keep the old ones around as long as possible to try to keep the combo going which would then discourage aggression and encourage careful kill control.

5) COMBO TIMEOUT DROP

If player doesn't collect combo shard in x seconds, the combo will drop to 1. This time can be extended by dealing damage to enemies, either by shooting or propellers.

With combo systems I'd make sure you are aware exactly where and when the player is and isn't going to be able to carry the chain and what does that do to your game play, eg. does it encourage stalling kills, does it involve luck, or should you shoot stuff as fast as possible to get bonus enemies or something?

Especially "hard" combos (that reset all the way down) can lead to very unforgiving system if you allow long chains, but make them tight since dropping a combo once can essentially ruin your score for the rest of the run, so make sure that's what you want if you go there. Obviously if you force a combo drop at regular intervals, then it limits the damage from dropping the combo to the end of the next forced drop, but if you do allow very long chains, you might want to also consider just making the combo roll down at moderately fast rate once "dropped" if you want something more forgiving... or whatever.

Especially if you have any significant amount of randomization in your enemy spawns, I'd personally make sure to limit the penalty from a chain drop to something reasonable (either by forcing short chains or making the drop "soft"), unless you want your players to grind for that "perfect chain" (which combined with RNG can be really frustrating experience).

6) POWER-UP

For each X combo shards collected, new power-up is spawned, 300px away from the player, towards the center of the arena.

When collected, player deals 2x more damage. It lasts around 1-2 seconds and is reduced only while player is shooting.
If player collects power-up while another power-up is already active, that power-up will increase in power by 20%.

There are some other benefits and variations, but it would be too much for this post.

Purpose: reducing enemy count + magical moment that feels good.

Instead of stacking power-ups, you could also just reset the timer on new pickup, but if you do want to make them stack, you could mess with the "X" somehow, eg. have a counter that increases by 1 for every shard if the player is NOT powered and only by 0.5 or something if the player already has a power-up. You could then reduce it further for additional levels (eg 2-stack gives only 0.25 or 0.33 or whatever), making it harder and harder to accumulate additional power-ups while at high power. You could also stall the accumulation completely if there is a power-up already on screen.. or you could do something completely different (maybe cap the power-ups at some point, don't accumulate counter with a power-up on screen, so player would be encourage to try to maintain max-power by picking up power-ups then always quickly accumulating enough shards.. or whatever)..

In general though, try to think outside the box, you can tweak a lot more things than what might be immediately obvious.

[Gungrounds]

4) RISK & COMBO

I think the most interesting balance is : "to increase the combo, you need to risk" which translates into "to get the higher score, you need to risk, which rewards you with progress ( new levels and power-ups )".

Few possible solutions:
1) Enemies come in groups so combo shards will always be close to the other enemies.
2) Most often, spawned shards are near the edges of the arena. New spawned enemies will attack this area.

On paper this sounds great, but it doesn't work so well in practice.

The "in practice" is probably something that really kinda would require playing to really understand why it doesn't feel right. Also do you want aggressive "kill everything fast" style or "slow and steady to keep the chain" type gameplay?

For aggressive style, you could always award more value for players close to the target on kill, either by dropping higher value shards if player is close, or by having the shards drop in value if not collected fast enough (eg. Futari "glowing gems"). These two encourage slightly different game play (in terms of how they encourage the player to move), but the basic idea is the same.

You could also just limit the number of shards on screen (either oldest one disappears if you spawn more or can't spawn more once the limit is reached) to encourage the player to pick them up without keeping too many around. Or spawn lower value shards if there is already too many high value ones on screen.

Or you could put new enemies on a timer, encouraging the player to try to keep the old ones around as long as possible to try to keep the combo going which would then discourage aggression and encourage careful kill control.

5) COMBO TIMEOUT DROP

If player doesn't collect combo shard in x seconds, the combo will drop to 1. This time can be extended by dealing damage to enemies, either by shooting or propellers.

With combo systems I'd make sure you are aware exactly where and when the player is and isn't going to be able to carry the chain and what does that do to your game play, eg. does it encourage stalling kills, does it involve luck, or should you shoot stuff as fast as possible to get bonus enemies or something?

Especially "hard" combos (that reset all the way down) can lead to very unforgiving system if you allow long chains, but make them tight since dropping a combo once can essentially ruin your score for the rest of the run, so make sure that's what you want if you go there. Obviously if you force a combo drop at regular intervals, then it limits the damage from dropping the combo to the end of the next forced drop, but if you do allow very long chains, you might want to also consider just making the combo roll down at moderately fast rate once "dropped" if you want something more forgiving... or whatever.

Especially if you have any significant amount of randomization in your enemy spawns, I'd personally make sure to limit the penalty from a chain drop to something reasonable (either by forcing short chains or making the drop "soft"), unless you want your players to grind for that "perfect chain" (which combined with RNG can be really frustrating experience).

6) POWER-UP

For each X combo shards collected, new power-up is spawned, 300px away from the player, towards the center of the arena.

When collected, player deals 2x more damage. It lasts around 1-2 seconds and is reduced only while player is shooting.
If player collects power-up while another power-up is already active, that power-up will increase in power by 20%.

There are some other benefits and variations, but it would be too much for this post.

Purpose: reducing enemy count + magical moment that feels good.

Instead of stacking power-ups, you could also just reset the timer on new pickup, but if you do want to make them stack, you could mess with the "X" somehow, eg. have a counter that increases by 1 for every shard if the player is NOT powered and only by 0.5 or something if the player already has a power-up. You could then reduce it further for additional levels (eg 2-stack gives only 0.25 or 0.33 or whatever), making it harder and harder to accumulate additional power-ups while at high power. You could also stall the accumulation completely if there is a power-up already on screen.. or you could do something completely different (maybe cap the power-ups at some point, don't accumulate counter with a power-up on screen, so player would be encourage to try to maintain max-power by picking up power-ups then always quickly accumulating enough shards.. or whatever)..

In general though, try to think outside the box, you can tweak a lot more things than what might be immediately obvious.

Woah. This is super useful, thanks a lot! You gave me new ideas, and probably some solutions as well.
I will write you a more detailed reply once I test this suggestions!

As for the "thinking outside of the box", I usually create a mind map in XMind.
In this mind map, I define all variables, conditions, variations, and possible interactions. I see it all clearly in front of me, and this usually leads to some interesting discoveries.

I hope I am not asking too much, I am also interested to hear what kinda enemy spawning system would you suggest for this kinda system?

[mystran]

I hope I am not asking too much, I am also interested to hear what kinda enemy spawning system would you suggest for this kinda system?

To be honest, I've never really tried to design any "twin-stick" shooter, so no idea how the enemy spawning dynamics work with a fixed screen and flexible shot direction. For scrolling shooters, you basically have a couple of options and a handful of variations, using camera position to activate enemies obviously won't work without scrolling, but (quite popular) timed-scripts would work, spawning waves when all the previous enemies (or some specific "big" mob) are dead would work (variations are to use these for bonus enemies if the next timed event is not reached yet or freeze the script until some midboss is dead, or use timer to force the next wave anyway if the player doesn't kill everything within some limit time, or make some timed events conditional on some previous enemies already having died) .. or if you want RNG, then choosing from a pool of hand-crafted waves seems like a reasonable idea (and placing those waves in a "deck" that is shuffled, then drawing from that deck should help give you more even distribution that won't get stuck in choosing the same wave all the time, even if the "deck" is large enough that not all of it is used during the level).

Other than that, no idea really. I'm a programmer and like most programmers I like the idea of spawning mobs completely "by algorithm" but realistically placing stuff at least semi-manually is almost certainly going to get you the most predictable quality, with more interesting challenges that you can balance more carefully. The way I see it, with any algorithm you have to play a little bit conservative (which typically translates to "little bit boring") to make sure it doesn't get out of control and always gives you at least reasonably playable content, where as with manual design you can go totally crazy and design "impossible challenges" and if it plays fine and turns out to be possible after all (and within whatever skill level you expect from your player), then you keep it, if it doesn't you discard it... plus this makes your scores more predictable and allows players to substitute some memorisation for "twitch skills" if it ever seems necessary (which might or might not be what you want, but it might not be a bad thing if you require certain scores for progress).

Speaking of which, the tried and true algorithm for quality level content (manually) is the "try random stuff" algorithm:

Step 0: Try to minimize iteration time if you can, hotkey-reload of scripts is good, simple (as in "thrown together in 2 hours") in-game editor is better
Step 1: Place some enemies more or less randomly
Step 2: Test how it plays:

• if it's too easy, add some enemies (or use harder enemies) and try again
• if it's too hard, remove some enemies (or replace them with easier ones) and try again
• if it feels a little off (too crowded, or too much dead air, or some enemies feel unfair, etc), tweak stuff and see if it improves
• if it feels bad or gets stuck in the "feels a little off" state, throw it away and go back to step 1
• if it feels good, keep it and go back to step 1 for the next wave/segment/whatever
• if it feels too much of the same, try something crazy or design a new type of enemy, or both

After a few rounds of trial and error you'll get a feel for what works and what doesn't (for your game) and it gets easier to make something reasonably "good" on the first try, at least until you get to the "too much of the same" problem where the most reliable solution I can think of is adding more enemy types (or even just different versions of the same enemies, like make them shoot slightly different patterns or move slightly differently or whatever).

That said, I do admit I'm not a huge fan of randomized content personally (well, except I do kinda like rogue-likes, the old-school ASCII type), yet I do realise that there are plenty of popular games that are more or less completely driven by algorithms... so I don't know. I'd probably build some basic facilities for manual enemy placement into the game anyway even if I was going for purely algorithmic placement though, simply because it'll make it easier to try what works and what doesn't.

Ultimately I think the solution is to try something, if it doesn't work (ie. game isn't fun) then try something else, then repeat until something fun emerges. Convergence can probably be improved by analyzing why the different iterations failed, but ultimately you will probably need some iteration either way.

.. but like I said, never tried to make a twin-stick, not even any kind of prototype hack, so this is mostly speculation and extrapolation from other types of games and YMMV quite widely.

[mystran]

Oh .. and I forgot to say that from the trailer video I do get the impression that you have a pretty good idea of exactly what makes this type of game fun.. so it's entirely possible that you are actually way more competent that me when it comes to design.

[Gungrounds]

It's alright. I am not looking for final solutions. Fresh ideas and new perspectives help a lot.

I was working more on this today, and this is what I got so far :

1) combo shards fade away fast - player must be close to the enemy to pick them up
- before player needed to move to that position anyway, but because he didn't need to do it fast, he usually was far away from the enemy
- the problem was : it was really boring to travel such a long distance in straight line, and collect this pick-ups
- new solution reduces that traveling length, and increases the tension
- also it increases the tension even more because it forces the player to collect combo shards right away even if there is an enemy approaching that location

I love games where pickups are protected by nearby enemies. Though, it is a bit hard to achieve it in this game. Any idea?

2) to keep the combo from dropping down to 0, player can damage an enemy with bullets or propeller or collect combo shard or deflect bullets
- by doing that, he will reset the combo timeout timer always to 0.75s ( it doesn't depend on how big is the combo, it's always hard )
- max combo is unlimited for now, before it was 99 which required around 33 kills in a row ( I love the feeling of breaking my combo record )
- combo timeout doesn't count down while there is no enemies on the screen
- when there is no alive enemies, new enemy will spawn right away. Thus, there is always a target, no rest.

Short and static length of timeout, makes it more clear how the combo system works. It is always the same, so I am always focused, and I know how I need to play to keep it high.

Approx length of each level is around 30-60 seconds. Survival levels are infinite.

3) With level progression the health of enemies increases to make it more mandatory that player must collect combo shards and use power-ups to kill them fast.

4) Power-ups also fade away fast. So in levels where there is lots of enemies, player doesn't manage to get so much advantage because they disappear while he is using the current one. That is why it is more useful to collect them few in row ( since they are close ), and increases the power-up level. Example: standard laser, piercing laser, spread piercing laser.


Perfect play would be something like this.
Try to stay close to enemies you are destroying.
If the target is far, shoot at it while approaching. Always look for the new target.
If there is multiple targets, try to distribute the damage among them to freeze the combo timeout.
Use special mode to increase you speed and reach faster combo shards in the distance.
Plan your moving route by taking in account combo shards, possible combo shards from enemies, and power-ups, while avoiding the danger.
Collecting power-ups must be a part of your route because it enables you to destroy enemies faster and get more combo shards.
The best power-up to keep the combo up is laser because it instantly hits the enemy.


That's it for now. Any thoughts?

[Gungrounds]

Today I iterated more the combo system while hoping to improve it further.

COMBO SYSTEM
- while helicopter is warm, combo is active
- combo increases by collecting combo shards OR by deflecting bullets
- helicopter warms up while player is in GOD mode or while he touches the enemies or grinds the edges of the screen
- helicopter cools down while he is shooting that heat out


To gain more score player needs to:
- activate GOD mode more often, use it for attacking instead of defending
- allow enemies that shoot bullets to stay in the arena so he can get higher combo from that
- destroy the enemies asap
- collect combo shards
- dont allow that the helicopter cools down completely ( he must enter god mode or damage enemies with propellers or grind the edges )

This is interesting because it rewards the player that risks more.


But this is still not good system because:
- to progress the game, player must get higher scores, which means I am forcing the player to use this risky play style
- it doesn't feel so interesting because player cannot decide so much when he wants to do what, he is forced to use propellers so that helicopter doesn't cool down and break the combo


All combo systems I tried so far were fun at start, but after a while they became a bit boring. I wanted to play the game in my own way, but game didn't allow me to.

[mystran]

But this is still not good system because:
- to progress the game, player must get higher scores, which means I am forcing the player to use this risky play style

I think this is a bold move in general. Some games are fun to play for score, some games are fun to play for survival ... but what is fun for score or survival for any given game is somewhat personal... not to mention I'd imagine most people would normally want to play for survival first, then gradually shift to scoring more aggressively as they get used to the game. If someone is struggling to simply stay alive, it might be quite frustrating if they are denied progress unless they score... and if your scoring system is fairly "strict" to begin with, I think making score play mandatory is pretty risky.

- it doesn't feel so interesting because player cannot decide so much when he wants to do what, he is forced to use propellers so that helicopter doesn't cool down and break the combo

My two cents would be to try implementing a completely different scoring system (well, at least a prototype of one), or radically tweak the one you have so it plays completely differently, then see how it makes the game feel. You could keep the "god" mechanic, but make it so that there is no combo, rather your score is directly tied to how close you kill enemies (forcing one to navigate around their bullets to get in range, possibly holding fire), or make the combos very short, such that the ideal approach is to burst many enemies together (probably needs some more thinking).. or make it so that shooting is not required to "cool down" and ideal scoring is to never shoot anything (and the risk comes from having to navigate around while the copter cools down, you can shoot stuff to make it easier, but you lose score).... or make it so that whenever the "god" mode ends, it clears the combo (but you get big points by going "god" with a high combo), maybe with a combo on how far you can build the combo if you want the player to cash out often.

Even if you go back to what you had originally, trying different prototypes might give you more perspective of how you want the game to ultimately play out.

All combo systems I tried so far were fun at start, but after a while they became a bit boring. I wanted to play the game in my own way, but game didn't allow me to.

This is why I don't really personally like most combo-systems (unless they are short), because just like you say they tend to force a very specific style where you basically need to figure out the "right solution" and then execute that perfectly and there's little room for improvisation.

From traditional shmups, I like the one in Ketsui (pointblank something to get a 5 chip, then chain with the lock-shot; I'm terrible at the game though, but the system itself is really fun) because it's short and simple (well, unless you're going for "credible high scores") and dropping the chain (which happens all the time) just means you need to find a way to pick it up again (eg. quickly use shot on a close-by pop-corn) or use the "empty lock" or something.

[Gungrounds]

Good points.

I tried short combos, but problem is that levels are not time limited so the optimal strategy is to let the enemies come in so you can kill as many as possible at once. This is a bit boring because there are periods of time when player is just avoiding them. How would you make this work?


I like the highscore hunting in survival ( infinite ) levels.
There is a way that I make that every level is infinite - after initial goal is finished I could spawn some item that player needs to collect to end the level. This item can be deflected around by using your spread propellers ( GOD mode ) so it is possible to avoid it.


Additional idea.
If you lose all lives before you collect this item, you don't get any points, so once you have only 1 life left, you start to gamble. But maybe even more interesting would be if number of lives you have left is the score multiplier, so it is a gamble from the very start.
The number of stars you got in certain level is reseted every time you start that level.


Another idea. What if there was a special hittable trigger / enemy. When you hit it, you force the next wave of enemies to spawn right away.
It could be special enemy with special behavior ( maybe pacifist ) or it could be any enemy possessed by this trigger ( it would be visible visually ).
They do something similar in tower defense games. This way, the player that is hunting high scores could trigger this switches very often and make the level much harder.



Though, the game is more fun when there is a score multiplier, and you can see bigger and bigger numbers flying just from every single kill, so I would like to keep this in some way.

[Gungrounds]

Here is few more gambling ideas.

1) When you kill some enemy, the score is added to stack. Player can do some action to transfer those points into bank.
If he gets hit or something else happens while the points are not in the bank, he loses all points on stack.
But more points he has on stack, the bigger combo he gets ( combo count = enemy chain kill count ).

2) The stack is an item in the arena, to transfer the pts into bank, player must hit it with propellers.
If player gets hit, this item breaks.
Once the regular goal is completed, you must hit it to finish the level.



3) Player combo raises while he is not using GOD mode. Once you use GOD mode and spread propellers, all points you got so far are transferred into bank. So even though this broke your combo, it is a really satisfying moment. Of course, while not using GOD mode you are risking to get hit, and if you do - you also lose all points and current combo falls to 1.

[mystran]

3) Player combo raises while he is not using GOD mode. Once you use GOD mode and spread propellers, all points you got so far are transferred into bank. So even though this broke your combo, it is a really satisfying moment. Of course, while not using GOD mode you are risking to get hit, and if you do - you also lose all points and current combo falls to 1.

Well, if I was you, I'd prototype this and see how it works, it sounds fairly straight-forward and workable on paper.

Here's some possible refinements: make it so that you always collect score normally (from killing enemies, picking up shards, whatever you want to give score for), but also build a multiplier (eg. from the shards) that stacks into a counter and becomes active for the duration of your next "god" burst and reset afterwards (or split in half, or whatever). Then it's not a simple "point transfer" but rather the player has some incentive to try to "cash in" the multiplier when there is a lot of scoring potential as opposed to using the god-mode defensively (which you could still do, but then the multiplier would be largely wasted).

If it seems like such an multiplier has too much of an effect of dominating all the score, you can always make it increment in smaller steps (or equivalently, raise the base multiplier.. like if you have base multiplier of 100, then +100 on the combo counter will only mean double score) and you could always cap it some value (eg. 99 or 999 depending on how fast it builds) to make the "cash in once at the very end" strategy ineffective.

One more idea (if you're feeling evil) that could make it fun would be to use the current combo counter to control enemy aggression (ie. "rank"): make enemies shoot more often and/or faster bullets and/or spread their bullets more if the combo counter is high. Then it'd not just be about "should I cash in or wait" anymore, rather it becomes "can I build it a little higher before the difficulty becomes unmanageable." If you do this and it feels like cashing in makes it too easy again (unless you keep half the combo counter even after burst, which could play nicely with a dynamic rank), just move some part of the combo counter to the permanent rank so it doesn't drop quite all the way down (and then only decrease that "permanent" rank on player death).

[Gungrounds]

Geometry Wars 3 had this progression system. It did frustrate me a bit when I didn't understand the mechanics good enough.

I tested most of the suggestions you gave me. Thanks again, it was really fun, and it definitely helped.

1) "For aggressive style, you could always award more value for players close to the target on kill"
This was really fun at start, but after some time it became a bit repetitive. I think this works good in standard shmups because enemies come in groups. When you kill one enemy just next to it there is another one attacking you. So it is really risky to approach them. I think this is really fun, but it cannot work in this game. It is implemented in other game we are making, Spitfire.

2) "if you force a combo drop at regular intervals, then it limits the damage from dropping the combo to the end of the next forced drop"
This is one of changes I added. Right now the combo resets every 10 seconds, and that is the only way it will reset. Player still needs to play aggressively. It is also possible to leave some combo shards for the next round, so you can have bigger combo for the whole 10 seconds. I am gonna visually represent this with burning rope that is connected to combo box. Once the fire reaches the combo box, the combo box will explode. During satisfying animation game will reward the player for few seconds with static combo, effects, and infinite power-up. Magical moment. I am still figuring out can I do more with it.

3) "combo raises only by killing enemies with propellers"
This was so fun at start, but again, it became repetitive after a while.
There are already situations in the game where player can destroy enemies faster and reach more score by using propellers.
Just one of them is that you can kill soldiers and damage other enemies in range while you are shooting at some other enemies at the same time.

4) "control kills"
Enemies have low health so the only way to do is this by waiting, and game is not so fun when you are avoiding enemies without shooting. Maybe this would work with more interesting enemy movement / behavior.

My conclusion is that the system shouldn't force the player to play the game in certain way. It is more fun if player has a range of effective & fun options, and he needs to figure it out by himself. When I see some enemy in distance, I want to be able to shoot it down and enjoy the explosion without being sad that it could have give me higher combo if I have approached it first.


This is how the system works right now:
- combo resets every 10 seconds, otherwise it is not limited
- when you kill an enemy, if there is less than X enemies on the screen, new enemy will spawn right away - more skilled player will destroy more enemies
- combo raises by collecting combo shards which are dropped by killed enemies
- combo shards stay on the field long enough so that you don't have to approach them right away, but also don't for too long so that you cannot save them for high combo in the next round ( it is not fun to avoid them so I wanted to eliminate that player abuses this too much even though he still can get good benefit... saving few combo shards is not a problem, problem is when you stack them for too long )
- power-ups last 1 second, and if you collect few of them in row, you get higher power
- you get score from : destroying enemies and deflecting bullets
- you get 3x more score if you destroy enemies with power-ups ( I wanted to boost even more the magical moment )
- rebalanced damage done by propellers, god propellers, standard weapon, and power-ups - this made a huge difference

I reward the player for everything he does, while he has freedom in how he wants to do it.


What I got from testing this different combo systems is that I learned more about this game. Some things in this combo systems was fun, and I will try to use this to create some new interesting levels. For example, there is gonna be a volcano world, where player cool downs with slower rate, so he cannot use GOD mode much. This is super fun once you master the game because most of the time you are open.

I will upload new video really soon.

[Gungrounds]

3) Player combo raises while he is not using GOD mode. Once you use GOD mode and spread propellers, all points you got so far are transferred into bank. So even though this broke your combo, it is a really satisfying moment. Of course, while not using GOD mode you are risking to get hit, and if you do - you also lose all points and current combo falls to 1.

Well, if I was you, I'd prototype this and see how it works, it sounds fairly straight-forward and workable on paper.

Here's some possible refinements: make it so that you always collect score normally (from killing enemies, picking up shards, whatever you want to give score for), but also build a multiplier (eg. from the shards) that stacks into a counter and becomes active for the duration of your next "god" burst and reset afterwards (or split in half, or whatever). Then it's not a simple "point transfer" but rather the player has some incentive to try to "cash in" the multiplier when there is a lot of scoring potential as opposed to using the god-mode defensively (which you could still do, but then the multiplier would be largely wasted).

If it seems like such an multiplier has too much of an effect of dominating all the score, you can always make it increment in smaller steps (or equivalently, raise the base multiplier.. like if you have base multiplier of 100, then +100 on the combo counter will only mean double score) and you could always cap it some value (eg. 99 or 999 depending on how fast it builds) to make the "cash in once at the very end" strategy ineffective.

One more idea (if you're feeling evil) that could make it fun would be to use the current combo counter to control enemy aggression (ie. "rank"): make enemies shoot more often and/or faster bullets and/or spread their bullets more if the combo counter is high. Then it'd not just be about "should I cash in or wait" anymore, rather it becomes "can I build it a little higher before the difficulty becomes unmanageable." If you do this and it feels like cashing in makes it too easy again (unless you keep half the combo counter even after burst, which could play nicely with a dynamic rank), just move some part of the combo counter to the permanent rank so it doesn't drop quite all the way down (and then only decrease that "permanent" rank on player death).

After lots of testing, my conclusion is that "reset combo every x seconds" doesn't really work.
It really doesn't matter enough did you collect combo shards before or after since the score multiplication is linear ( instead of exponential ).
But if I make it exponential, I also don't like that player will earn more score if he waits and if he saves combo shards for the next combo bar reset.

So I am planning to test again the system where combo resets when player uses GOD mode. I didn't like this before, but the reason was that earlier levels are not complicated enough. Later in the game, you are forced to use GOD mode often, though it really depends on how skillful you are & how much you risk - which is perfect for this system.


This is the current plan, I would like to hear your feedback. The additions are bolded.

1) COMBO RESET - when you EXIT God mode - which means that during God mode you keep the combo high and earn even more points.
2) PROPELLER KILLS - give you x10 more score. ( not sure about this yet, but it rewards you for risky soldiers / enemies kills + using god mode )
3) POWER-UP collecting - raises combo by 5. ( earlier you pick power-up(s) - more score you make )
4) GETTING HIT - resets combo, but you don't lose the score ( score is always added right away ).


- when you kill an enemy, if there is less than X enemies on the screen, new enemy will spawn right away - more skilled player will destroy more enemies
- combo raises by collecting combo shards which are dropped by killed enemies
- combo shards stay on the field long enough so that you don't have to approach them right away, but also don't for too long so that you cannot save them for high combo in the next round ( it is not fun to avoid them so I wanted to eliminate that player abuses this too much even though he still can get good benefit... saving few combo shards is not a problem, problem is when you stack them for too long )
- power-ups last 1 second, and if you collect few of them in row, you get higher power
- you get score from : destroying enemies and deflecting bullets
- you get 3x more score if you destroy enemies with power-ups ( I wanted to boost even more the magical moment )


IDEAS
4) MAX COMBO LIMIT - to avoid perfect play ?
5) MAX COMBO REACHED BONUS - maybe something happens when you reach it?
6) GOD MODE ENTER / EXIT BONUS
7) DIFFICULTY INCREASES WHILE YOU ARE USING HIGHER COMBO

[mystran]

1) COMBO RESET - when you EXIT God mode - which means that during God mode you keep the combo high and earn even more points.
2) PROPELLER KILLS - give you x10 more score. ( not sure about this yet, but it rewards you for risky soldiers / enemies kills + using god mode )

Right, this seems like a solid potential design, although I wonder if x10 is too much, it might make any other scoring essentially irrelevant (unless this is exactly what you want), but you could try lower multipliers, even just 2x might work fine.

4) MAX COMBO LIMIT - to avoid perfect play ?
5) MAX COMBO REACHED BONUS - maybe something happens when you reach it?
6) GOD MODE ENTER / EXIT BONUS

I don't know how any of these would work in practice, but I definitely like the line of thinking here.

If you do make a combo limit, I'd imagine it probably makes sense to make it relatively low, to clearly communicate that you are not supposed to go for the maximum and then stick with forever, but rather just get there and then cash in with GOD right away (which I guess is more in line with how you want the game to play, right?)

7) DIFFICULTY INCREASES WHILE YOU ARE USING HIGHER COMBO

This could be fun as well.

[Gungrounds]

I tested it, and it is really tense and fun , but there is few problems.

System:
- combo falls 20 per second while using god mode ( when it dropped to 0, I never wanted to use god mode )
- getting hit drops combo to 0 ( this makes it super clear that using god mode is a good choice ( except if u really want super highscore ) )

Problems:
- when I started playing I was risking too much which resulted in me getting hit, it is a bit hard to say when it is too much risk
- it is not clear how much one needs to risk to reach certain high score and get the star, but this is probably fixable with better user interface
- maybe the biggest problem is that with this system game is stopping the player from using something that is unique in this game.He should be motivated to use propellers often instead of never.
- it doesn't have logical sense why combo drops while using god mode ( I know this is usually ok, but it bothers me a bit )

I think all this is fixable though, but it will be hard to know for sure until someone else than me tries to play the game.

I tried limiting max combo to smaller value, ramping up difficulty with higher combo, multiplying score with combo only while using god mode etc. but I didn't manage to make this work.

The last combo system had only one problem - "combo drop" condition.
So it is really close, but also so far away haha.

[mystran]

Problems:
- when I started playing I was risking too much which resulted in me getting hit, it is a bit hard to say when it is too much risk

Are you sure this is actually a problem? It sounds like something that would be really fun to me ... and from my point of view the whole point of "risk" is the uncertainty of whether or not you can actually deal with it in practice. Also keep in mind that when you started playing, you already had all the experience from every other iteration of the game and also perfect understanding of exactly how the whole scoring and risk&reward system is supposed to work. I'd imagine a fresh player would almost certainly be more cautious by default and gradually start increasing their risk-levels.

I think all this is fixable though, but it will be hard to know for sure until someone else than me tries to play the game.

There is a very obvious solution to this: get someone else to play it!

[Gungrounds]

I think I will make that all levels are survival. I think this is the only way for this game to make this progression star system really work.
If I was making new game, it wouldn't be a problem to make everything work around some of suggested scoring mechanics, but here it's just, not perfect enough, and it's starting to drive me crazy haha.

[Gungrounds]

This is how it works, finally.


Levels are split into waves.
With every new wave, difficulty and permanent score multiplier increases.
Temporary combo multiplier increases by collecting combo shards, decreases over time, and decrease even faster while using god mode.
Combo drop speed is proportional with the current combo. Maintaining high combo is more difficult.
Using god mode while having high combo means that combo drop will be big, while using god mode with low combo will not reduce it so much.
Propellers give x5-x10 points for killing enemies and deflecting bullets. It is especially beneficial to deflect bullets because this doesn't affects next wave ( difficulty increase ).

The result is very interesting economy: player needs to find a right balance between risking, finishing waves, using propellers, and god mode. Player that uses god mode often in right situation will benefit from high propeller multiplier. Deflecting swarm of bullets or killing multiple enemies is rewarding even though the combo drops down a bit. Player that is able to risk a lot without getting hit will get better score. Player that is capable of surviving long time will earn better score as well.

So there is a choice in how you want to play.

Of course I need to balance this properly, but this finally looks like something good.

[Gungrounds]

Yay, check this out. Almost at E3.
Tomorrow I am gonna upload new videos that will showcase the combo system and other improvements!

[tiaoferreira]

WOW! This new lay out ROX! Congratulations, man!

[Gungrounds]

Thanks @tiaoferreira

Here is the new gameplay video as promised! Please watch it in HD, without full screen switched on.
This is of course not finished, we use some art placeholders, and user interface requires a lot more work.
https://www.youtube.com/watch?v=aPwbq5_M-q0

Huge thanks to @mystran for all feedback! If you are interested to include some link in the credits section, let me know.


Description of scoring system ( not balanced yet ):

a) when player destroys an enemy, he gains score equal to = enemy_value * combo_multiplier * wave_multiplier * small_propeller_multiplier

b) combo multiplier raises by collecting combo shards that are dropped from killing enemies ( they last around 10 seconds which gives to player the freedom to not collect them right away )

c) combo falls gradually with time, and the bigger the combo is, the faster is falls... this means that the bigger combo is harder to maintain

d) combo falls gradually few times faster while the player uses OVERDRIVE ( god mode, time limited immortality ) .. this means that player that risks more and uses overdrive less will have an chance to score more points, if he doesn't get hit...

e) combo falls to 1 when player gets hit ( this means that it is always better to use overdrive to avoid the damage than it is to getting hit , but it is usually not a trivial choice because in most situations there is a way to escape )

f) all missions are infinite ( survival ) and they are separated into waves... with each wave, the score multiplier increases by 1.. for example in wave 2 player will gain 2x more points, in wave 3 he will gain 3x more points etc.

h) if player kills an enemy with small propellers ( while not using overdrive ) , he will get 2x more points

f) buduci da su leveli beskonacni te su podijeljeni na waveove, kod svakog wave se wave_multiplier povecava za 1, znaci igrac ce u drugom waveu dobiti 2x vise bodova, u trecem 3x vise bodova, u cetvrtom 4x vise bodova tj. sto duze prezivi, to ce dobivati vise bodova


EXPLANATION of PERFECT PLAY

The player that risks by not using overdrive mode will gain higher combo multiplier, a chance to earn more points.
Because the levels are infinite and over time player gains more score from wave multiplier, longer he survives, more points he will earn.
This two systems are in clash, because if player risks too much, he will not survive so long, and if he doesn't risk, he will not earn so much points from every kill.

Player that risks by killing soldiers and other enemies with small propellers, will gain 2x more points than usual. So if his combo multiplier is 80x, it will be like his combo multiplier is 160x.

Player can also earn more score by :
- avoiding finishing certain goals that will trigger the wave to increase ( for example in level where player needs to save friendly soldiers, he can try to avoid them )
- killing optional enemies like nuke tanks ( they give 10x points, but cause big explosion )

What's missing? A better user interface, lots of visual and sound effects to make it all easy to understand.


I tested many options, but for me the scoring system was more fun when I had a choice in how I want to play.
It's great that this system doesn't require that player knows all the details from the very start. It works on it's own, but if you want to do better, than some details become important.


If you have any suggestions please let me know.