Coming to this from 13 Sentinels it is almost precisely the opposite of what I expected. Where 13S has a phenomenal plot full of complexity, mystery, and deep characterization, UO's written like generic pulp fantasy—no, not even that, like a monster-of-the-week TV show apisiring to generic pulp fantasy. But where 13S's combat feels sloggy and bolted-on, UO's is consistently fun and reasonably engaging.

The core failing of the game is the fact that word, "reasonably". Although battles in the game are never boring, the systems promise far more than they're actually able to deliver. The game is stuck between two worlds: a stolid traditional Attack Magic Item Flee system and an FFXII-style fully automated luxury battlesim. It's straining towards the latter and it almost manages to grasp it, but the small gap remaining makes all the difference.

The goal of a system like this is to elicit in the player strategic thinking. Consider which characters complement one another's skills, which skills are useful in which circumstances, how to make a unit that's flexible, powerful, and effective against the present challenge. Find clever combinations of skills and equipment that exploit holes in your enemies' defenses. This is what Unicorn Overlord wants, and it's what it narrowly misses.

This is how a battle goes: your unit (composed of several characters) bumps into an enemy unit. Each unit is arranged on its own 3x2 grid, and each character has a set of active and passive actions they use under certain user-defined conditions in initiative order as long as they have resources. You can customize these before the battle, swap equipment with anything in storage, and the game will tell you how much net damage you will deal and how much you will take. Once you're satisfied with your setup, you hit go and the battle plays out automatically. Since resources are very limited, most battles only last a few "turn cycles" as they would exist in a Dragon Quest style system.

In practice, this system has a few critical issues. The first and most noticeable is the degree of variance in each battle. Whether hits are critical or miss entirely is up to the random number generator, which is of course quite sensitive to initial conditions—as indeed are the deterministic behaviors of a battle. This means that you can change a prospective battle from losing to winning often enough by simply rearranging your characters or toggling skills on and off arbitrarily, a practice which quickly overwhelms the amount of time you spend on actual tactical decisions.

The ability to change a unit's programming for each battle also erodes the gap between this system and something more traditional. With some finagling, you can usually choose fairly specifically which enemy will be hit by which attack, allowing you to essentially route around the automation and pre-plan a more standard RPG battle. That level of customizability undermines the conceit of programming.

At the furthest extreme, because you can swap any equipment with anything in storage before a battle or between battles, in principle the best way to play is to keep all your top-tier gear unequipped and just swap it on each unit before they fight and off once they're done. I couldn't bring myself to go to such noxious lengths most of the time, but I did keep a few "just in case" initiative boosters around.

I hypothesize that this game would be substantially more fun if you voluntarily chose not to change your unit's loadout after deploying them. I just wish this were something that was built into the game, rather than something players have to turn to upon discovering that the game doesn't live up to its own goals.