Overall rating: 3/5 (borderline)
[i]Hexa Turn[/i] brings another short, minimalist puzzle game to Steam. The puzzles range from very easy to moderate difficulty, and should take 5-10 hours to play to completion. The core mechanic is a turn-based process where you attempt to defend a goal square from an enemy triangle, which moves one square per move on a hexagonal grid, by placing black hexagons to block its path. It was pleasant and relaxing, and should appeal to players who enjoyed similar games like [i]Hexcells,[/i] [i]Hitman GO,[/i] [i]Unium,[/i] or [i]Sokobond.[/i] However, I hesitate to recommend it to a wider audience, due to conceptual inconsistencies with the game's mechanics, which I will explain below.
The mechanics are introduced by saying that "the enemy always takes the shortest path." As quickly becomes apparent by Package 2, this is wildly underspecified. When there are multiple shortest paths available, the enemy must pick one, and exactly how it does so is left up to you to find out. As far as I can tell, it may be programmed independently for each level and often isn't optimal. This becomes an important part of the gameplay and is exploited in many of the puzzles' solutions.
This becomes problematic because there are levels where, if the AI were to play optimally, the puzzle would actually be impossible. The first puzzle where I am sure this occurs is level 2.6. That is, even following the shortest-path rule, it is possible for an enemy to reach the goal, no matter what squares are placed in its path. And yet, it is possible to win on this level, because the AI chooses to go a different route. This makes for a confusing puzzle experience and makes the game conceptually less compelling, as it prevents patterns that may be found on one level from generalizing to others where the AI may behave differently (e.g. on 1.7, 1.22 and 4.18).
(On other levels, the AI would win if it played optimally, but the AI can't win following the "shortest path" rule, e.g. 1.15, 1.16, and 1.21. This is interesting and a positive part of the game design.)
In fact, the AI behavior seems to be actually [i]nondeterministic.[/i] At first, I thought that the enemy triangle would always go in the direction it was facing if possible, but this does not seem to be the case. Instead, on some levels it seems to pick randomly between multiple options, and on undoing or resetting the puzzle you may get a different result. For a strategic puzzle game, this is at best an odd design choice; at worst, it can mean that you solve a puzzle by luck rather than by the merits of your own solution. Levels where I noticed this nondeterministic behavior (and not just in symmetric cases) included 2.6, 2.7, 3.1, and 3.3. Once again, this nondeterminism becomes a key element used in the solution for some puzzles, e.g., 4.12. If the designers wanted randomness to be a part of the game, why not have the AI choose randomly between all possible options, or at least limit randomness to a specific behavior (e.g., randomly choose between left and right) that is consistent across levels?
Possibly the worst case of inconsistency is when the "bait" mechanic is introduced -- which is otherwise my favorite mechanic, and is a lot of fun to play with! In most levels, if you place a bait and there are multiple shortest paths, the enemies will favor the real goal over the bait. However, in some levels I was able to observe the enemies actually choosing the bait over the real goal instead, for example in 4.11.
In summary, the mechanics are underspecified, leading to inconsistent enemy behavior. While not a dealbreaker for the game, this means that many solutions do not generalize across levels, and leaves trial-and-error as the only reasonable way to attack puzzles where the enemy behavior is unknown. As a result, while I enjoyed the game and think it has a ton of potential, my overall recommendation is mixed.
Finally, a list of some minor issues. Some levels could be solved with fewer resources than were given, e.g. 1.7 and 3.23. There was also a bug related to undo where sometimes undoing at the wrong time gets the game into an inconsistent state, and you either could not place an item that you supposedly have available or the number of items available was not listed correctly. I observed this bug for example on level 3.24. The "1000 Turns" achievement is annoying to complete as it takes far fewer hexagons to actually complete the game, and you are just left clicking hexagons for no purpose. It would also have been nice to display the current level on the screen (or at least on reset), not just when the level is first loaded.
Despite its flaws, [i]Hexa Turn[/i] is an easy way to get some achievements for players who are looking for that. Some puzzles require head-scratching, especially towards the end, but puzzle enthusiasts will likely not find anything too difficult, and will not need to use any of the hints.
