voting leans towards how pretty the image is not whether it's actually interesting to play. It's also not possible to manually test the difficulty of a puzzle before submitting and tweak as necessary - where's the 'test' that should be next to 'validate'? As others have said the very large levels are an exercise in tedium.
Puzzles cannot be "tested" because the puzzle generation algorithms are too computationally heavy to be used in the client. A 50x50 puzzle can take up to an hour to be generated on an high-end machine.
All this bikini women shit... I'm not a prude but I'm not a 13 year old either. This puerile bullshit immediately alienates me. Good luck with the market you're going for.
If you use all crystals on 2 crystal crafting you will quickly find your weapons vastly overpower the current enemies making the game incredibly passive and boring.
Is the order of the missions random? "Get 100% on 25 missions" was something like my 4th mission given when I was still on level 5 or so. Doesn't make much sense to have that so early. Perhaps you're not gathering enough data but it would be more interesting to have missions that encourage different playstyles, e.g. 'use flame only to...' or 'win without taking damage while staying in the front 50% of the screen' etc.
Also the shop is completely useless if you spend your crystal to craft items. Total earned cash - 1500. Cheapest in store item - 10x as much! Crafting is quite broken generally.
zone 10 boss - avoid his attacks until beam is used then attack during fatigue period. For second stage avoid large yellow projectiles and hit small blue ones.
the AI just isn't good enough. With good AI a lot of changes would be necessary as currently the free stars and/or stupid AI make for very easy levels but the game would be much better.
Overall the difficulty is on the easy side but I do have a couple of points. First, it would be good if there was a clear indicator which enemies have grenades. Second, it's unclear what triggers the AI to react to you. A developed sound/stealth element, 'wake up' events with set trigger or something else would be would great. For example enemies are inactive until you enter LoS or use of guns within a radius, with small/silenced guns having a smaller radius. hurt enemies scream similarly. LoS also triggers) Third, the completion time bonus is usually double what would be challenging. Fourth, as others have said, it would be good if equipment choices were less clear cut and more situational.
Overall I do really like the game.
I think a save state/return to saved state feature would be a cool addition to this game. This would allow you to test a possibility and return to an earlier position easily if it didn't work out.
Ok finished the game this time. By the end I had a 15ish flux 500 DoT damage (2xx burn+2xx poison+2xx weakness+helplessness+slow) aimed AoE with 50+ duration. Meanwhile I was permanently buffed with 250 per turn 30+ duration regen (turns out it stacks on itself oddly, with 8x heal for 3-4 casts) with swift thrown in and had infinite flux due to the half health/half flux artifact combined with the regen. I do love the rune system but hope you can make some balance changes because currently the biggest danger to the player is the player himself.
I'm with Rixion, please add "are you sure" dialogue box to self targeting offensive spells. Gameplay as is very spammy on some builds so if someone drops out of sight for a second it's very easy to kill yourself with the autotarget.
My first guy, a pugilist, ran a greed+rage etc courage permabuff, regen+powerful buff, sublimation and I found the game too easy to need to use the other tricks up my sleeve (rage+fluxfeed instaheal, a harmful aoe, large DoT cloud, blocks, etc) until eventually the tedium led me to my death. I'll try another character type but going off my first experience I think the difficulty could use a push in the direction of roguelike norms otherwise all that gameplay depth the rune system provides goes rather to waste.
The level design is really excellent. Solutions often far from what the first impression suggests, requiring lateral thinking more than brute force approach.
Puzzles cannot be "tested" because the puzzle generation algorithms are too computationally heavy to be used in the client. A 50x50 puzzle can take up to an hour to be generated on an high-end machine.