Not a game. A puzzle with only one right answer.
Largely unplayable in terms of a video game. This is like one of those really really old puzzle games you'd see for Windows 3.1. You know what I mean? Like one of those box-pushing games or a maze with keys and doors, and there's exactly ONE right way to clear the room? The (lone) risk/reward schedule is set up like one of those games.
In other words, it's boring. Even worse, the deceptively simple interface obfuscates the game's deliberately obtuse matching rules. The player has no agency whatsoever until they deduce which combos will cause a gem to form where, and since these rules are never explicitly stated in the tutorial, it means you'll need to lose a few games before you'll even have the foggiest idea how the game is played.
The final nail in the coffin is, when you realize you've inadvertently made a large gem in the wrong place, THE GAME WILL NOT LET YOU RESTART THE LEVEL. You need to start a New Game from the VERY FIRST LEVEL if you mess up. This is poor form and very unfair.
Labeling this title a "gem matching game" is false advertising. People hear "gem-matching game," and they think of classical puzzle games where you match small gems, the small gems disappear, and you keep doing it ad infinitum. You know, "easy to learn, difficult to master?"
Well, this game plays more like a cross between a button-hunt and a combination lock. It's a pain in the ass to learn, and if you put up with its bullshit long enough to get the hang of it, it just adds more ways you can fail so you need to start over more often, and random chance means it's more likely you'll end up in an unwinnable state than you'll solve it, even if you work the system right. Just bad game design, all around.
Candystand knows how to code and they have good art, but it's clear they don't know the first thing about game design. Gameplay should be simple and intuitive at the most basic level, and additional complexity should emerge as new risk/reward schedules are added to the game. Overall success or failure should be a long-term thing. It should NEVER hinge on the first move.
And if your game MUST have a ton of opening moves that lead to inevitable dead-ends, the player needs to be able to reset the game at any time, so they can re-do that first move over and over again until they figure out how to frigging play.
Yes, if we took the time to learn the bizarre vertical/horizontal system, we'd be able to play as well as the game's creators, but when the game's on a portal and we have hundreds of other games to choose from, we'll quit a confusing game and keep coming back to a game that's fun. This game was confusing, not fun.
I'm all for games that make you learn, but the learning can only occur in the space of experimentation. If the first wrong move screws your entire campaign, or requires you to make a dozen other moves for the game to say "NO MORE MOVES" and start you over, that game makes you waste a minute or two between learning events. Nobody's going to put up with confusion for that long.
The difference between novelty and confusion is subjective, so I can't prove I'm right. But assuming you want people to actually enjoy playing your game, you want it to be easier and more intuitive at the micro level, and complex in ways that can be addressed gradually over several (successful!) playing sessions.
It's possible this is a "thinking-man's game" and I'm just not the right audience for it. But if that's the case, neither is most of Newgrounds, I suspect. I recommend some serious introspection on the part of the authors, if their goal is to create a fun game that many people will enjoy playing.
If the goal was just to pad Candystand with as much content as possible, congratulations. Mission accomplished. Nothing me or anyone else says about the game can take that away. Maybe there's a big market for boring games with bad gameplay, I dunno. But it's surely not for the Newgrounds crowd.