Now while I was looking at comments on steam, this guy's comment was first and I disagree with every single thing he said, but I mean I guess everyone has their own opinion....:
Ultimate General: Civil War is a fun game when you first play it. I've even written a few guides for the game.
It bills itself as an authentic historical game with good AI. At first glance, the game does all of this wonderfully.
Deeper analysis of the game reveals fundamental design problems that only become visible with time.
Many of these problems are known to the developers. However due to the mechanism the devs use to solicit feedback these issues remain unaddressed since early access. Other problems are more fundamental, such as a lack of true player agency (a la Mass Effect 3)Problems with the AI
Once you get past being beaten by the AI in the beginning and get used to the quirks of the game then the advertised 'strong' AI vanishes. Each map basically devolves into one of two strategies depending on if you are on offense or defense.
The challenge the AI provides against you is not in the AI being smart. Rather the challenge is in the sheer numbers the AI has versus you.
The problem gets worse if you play on half speed. With careful micromanagement, it's possible to defeat armies with 5x the number of men you have as well as drive back the relentless AI charges (if you aren't playing offensively that is). You can even cancel AI charges if you do things right. The AI pretty much always reacts the same way to you and it is easy to learn how the AI reacts and do the same thing every game. This is a regression from UG:GB where the AI had different personalities and acted differently.Problems with Game Mechanics
Some of these problems with AI involve various bugs. These include AI breaking charges if you reposition units, AI favoring charges even if the unit is exhausted (which you can halt the advance by micromanaging artillery). Others are more fundamental. For example, early on in early access the AI had problems defending their artillery. So the game devs, instead of making the AI better at defending their artillery, made artillery invincible to melee.
Similar examples of game mechanic problems exist elsewhere. Melee mechanics were broken since early access. 2 brigades of 200 men in melee will perform 4x casualties as 1 brigade with 400 men in melee. Thus, some players (in a single player game mind you) liked to stack lots of tiny cavalry brigades and abuse the melee mechanic bug to beat battles. This bug was reported and these players complained the game was 'too easy'. Instead of fixing the broken melee system (or *gasp* asking the players to not use such strategies in a single player game), it was decided that 'cavalry was too powerful' (since this was the most commonly abused unit). Thus, cavalry got nerfed to the ground so much that the only way to use them is to abuse the melee bug. A 400 men cavalry brigade can't even chase down 90 skirmishers without getting routed.Problems with Realism
These issues of game design, where bugs are not fixed but instead hidden, permeates to the entire game. It results in unrealistic situations that hurts realism of the game. Such as 750 cavalrymen who manage to surprise 200 artillerymen from the rear failing to beat the 200 artilleryman and instead being routed and losing half your men. BUT if those very same 750 cavalrymen were dismounted before engaging the artillery, then the results won't be as disastrous (AI will still not lose men though thanks to invincible AI artillery). It results in the best way of dealing with artillery being shooting them (and eating canisters to the face) rather than charging the cannons. It results in on harder difficulties having to play the game a certain way rather than having true agency.Problems with Player Agency
On the topic of player agency, UG:Civil War does a poor job of this. On easier difficulties player agency is somewhat there as the AI is not gifted tons of men to produce artificial difficulty. Thus as the game goes on if you manage to defeat the AI (where the AI loses 3-5x your men) the AI won't be able to reinforce. This creates other problems such as the 2nd half of the campaign being too easy.
Due to this complaint (and also due to complaints from players exploiting game bugs and AI weakness on higher difficulties) combined with the lack of difficulty levels, hard and legendary difficulties are not only incredibly broken but also unfun, a tedious chore, and lack player agency. The AI patently ignores the 'army intelligence' screen where you see how many men it should have and deploys much more men than the number shown. This means you can't play with a small, elite army that's lowered the AI army strength, no, you have to bring as much men as possible.
Since the game ignores the army strength on higher difficulties there's no point to capturing or killing AI at all. Instead it's a matter of force preservation (since you only get limited recruits). As the AI is gifted a new army each time you destroy an army, the facade of player agency collapses completely. It simply does not matter what you do or how many men the AI loses.
Fundamentally these problems are a result of a lack of dynamic campaign. This is compounded by game bugs exploited by players on hard/legendary that were never fixed which makes the game 'too easy' for them. Or the lack of a true dynamic campaign where you can choose to end the war early if you manage to defeat the enemy army.
No, the solution chosen by the devs was to brush the bugs under the rug and just give the AI more men (or men invulnerable to their weakness such as invincible artillery). This doesn't result in a fun game. Rather it results in a game where battles become a CHORE to micromanage and abuse AI weaknesses (not to mention all the small bugs which still have not been fixed). It's not any more challenging or hard in a tactical level.Problems with Feedback
One aggravating factor which lead to these decisions has to do with feedback. Due to the many different ways one can play a game, issues which may crop up for some players may not crop up for other players. Similarly, due to the vast difference between normal and hard in difficulty (different AI bouses), playstyles/bugs which may work on normal may not work on hard.
Instead of soliciting feedback from the playerbase as a whole, the general impression I have from watching the game progress since early access is that a small privileged group of testers (who tend to be experts and or play using certain strategies) have their suggestions favored more by the devs. This isn't necessarily the fault of the devs, as these players are simply more active in providing feedback.
This generally wouldn't be a matter if it wasn't for the fact that some of these more active players have a habit of discounting other player's problems (since it doesn't occur for them with how they play). This leads to issues where unless a problem is faced by a majority of players, problems which only affect some players are routinely ignored or outright dismissed. It also leads to poor balance changes and honestly a terrible experience for new players who are thrown into the game (eg the patches making the first intro/tutorial level harder and harder and harder to the point where new players have to go on the forum and ask for how to beat the FIRST LEVEL).Hostility to Modding
Many of these fundamental issues would be solvable if there were modding support by the devs. One would think the modder DarthMod would have supported modding in his/her commercial games, having come from the modding community.
Instead, impediments to modding were added on purpose to reduce the basic modding possible in UG:CW.Conclusion
I would honestly prefer to play the first early game access version of the game rather than the current version of the game as it is now. I lack faith the devs will be able to fix fundamental problems based on experiences since early access.