Aetius

I've noticed this tendency on the part of the AI as well - once it has captured those points, it does not defend them. To be fair, defending the Sunken Road would be difficult, but ... the AI has the troops.

If an infantry brigade loses its attached skirmisher unit, it is no longer able to deploy skirmishers. Not sure when this changed, but it makes very little sense.

The introductions to the battles say a lot of things, the majority of which are wrong or misleading, but are more-or-less historically accurate and good flavor. They are, however, entirely unreliable as a source of accurate information about the scenario victory conditions. You get zero men for losing Legendary Confederate Gettysburg - +20% of zero is still zero. And +20% of $28k is a very different reward from +20% of $168k.

Except, sometimes they aren't, or are unclear. The first phase of Confederate Chickamauga is an excellent example - the victory conditions list what you need to do for the whole scenario, but they does not tell you that if you don't achieve the first few objectives, you immediately lose the entire battle. This is something that is critically important for the player to know, but can only be discovered through bad experience. Leaving aside melee dragging, I think unclear victory conditions and the battle timer are the most important remaining problems in the game (followed by player units never showing up for battle). He's pointing out that if you lose (deliberately or otherwise), Politics is a lot less useful - you get very little extra money and very few additional troops (and in some cases zero, like Confederate Gettysburg).

I think it would be better for reinforcements to enter the battlefield from their particular direction - I don't think any units should simply be placed in historical positions. The funny thing is that the game generally does a really good job of this; it only falls apart in particular phase changes in a handful of grand battles. I think having the reinforcements arriving makes the battles more realistic - Gettysburg, for example, was largely driven by when and where reinforcements arrived.

The frustration is not just on the first time through, and it gets worse as you move up in difficulty level. First, no Civil War army would ever give up a huge tactical advantage "returning to camp". Second, as a game it is very, very uncool to arbitrarily teleport the player from a good position to a bad one, especially when it involves taking away hard-won victory points and/or scattering divisions to the wind. It feels petty and mean, especially when it's obvious that the game can save the state of the battlefield between phases. And functionally, part of the real pleasure of campaign mode is having the battles develop differently from the historical battles; this is taken away when the battles are arbitrarily returned to a particular state between phases. And finally, mechanically it makes no sense - how did those troops get there? In some cases, I've seen troops teleported miles across the battlefield, with no Condition loss or other effects. It ruins the immersion.

I spread artillery out for the same reasons - I had one Shiloh game where my artillery division was stuck defending the Hornet's Nest alone. I keep cavalry together for survivability reasons, especially on Legendary where you really can't afford full-size units. It's also a lot easier to just hit the division button to gather all the cavalry together, instead of having to dig them out of the list.

My naming scheme is similar to Andre's, but I split my brigades into two classes - rifle and grenadier. Rifle units have rifles, and when they get stars they get shooting-related skills. Grenadier units have muskets, and get melee-related skills. I've also started to number divisions regardless of which Corps they are in. For example, I Corps consists of 1st, 2nd, 3rd, 4th, and 5th Divisions. II Corps consists of 6th, 7th, 8th, 9th, and 10th Divisions, etc. I found that pattern made it easier to identify divisions on the field and keep them together as a unit. Thus the 3rd rifle brigade, nominal 1st Division, II Corps, would be named 6/3 Rifle.

Only to the extent that the victory conditions might require you to keep casualties down. And yeah, it's screwy. I admit that I've been ... less than careful with allied units on Legendary. I've even pulled them out of the line to recover morale so that more of them eventually get killed. I've also let enemy units opposing them recover for the same reason. It's bloodthirsty, but on Legendary you need every weapon you can scrape together, even if its only to sell.

I haven't had any trouble with this. There have been mouse-related issues previously on high-resolution screens, so might be related?

*And* they use no Condition while routing. I've abused manual routing several times in Legendary to drain stamina from enemy units that are pursuing mine. I believe this is because Condition matters so much in melee combat. It appears to me that units lose Condition equally, even though they are different sizes. For example, if you attack a unit in melee with one unit, wait a bit, and then hit it with another unit, the defenders will run out of Condition and either flee or surrender.

I think this is one of the key things that needs to be fixed before the full release. The victory conditions and timers need to be correct, complete, and explicit. To do otherwise is arbitrarily unfair to the player, and will make people angry - not everyone has a lot of time to replay battles to figure out how to actually win them.

Having played so much on Legendary recently, my only response is to salivate at getting all those allied brigades killed so I can take their weapons.

It's not about failed tactics - the game already supplies these notifications. The problem is that they are easily missed in the chaos of battle.

I'd say the AI in this game is pretty spectacular. It is able to form a battle line, attack weak points and flanks, and react reasonably well to being flanked. The two main flaws right now that it is far too willing to accept grossly uneven exchanges of fire, and it *really* likes fortifications to the point where it will reoccupy fortifications that are already overrun. Solve those two problems and I think we'll have a vicious beast on our hands.

Please, for the love of all that is holy, can we get audio notifications of battlefield events? I just lost Shiloh because I lost track of Jenkins, and he rode halfway across the map, grabbed the Pittsburg Landing VP which had just appeared, and won the game for the CSA despite the Confederates taking 5-to-1 casualties. By the time I realized what had happened, I couldn't get units there in time to stop it.

The best I've done on Everettsville is roughly 1.5 to 1 (13k Union, 7.5k Confederates), on patch 0.75, using an attack on the far Union left: https://www.youtube.com/watch?v=Va5MSttuRLE I've tried the frontal assault twice in tests, with one win and one loss - however, casualties were north of 10k both times. For a battle with such a small victory reward, it's just painful. When I play it again, I'm definitely going to push harder on that flank - the supporting attack on the fortifications was unnecessary, I think.

Everettsville is the worst offender - particularly on Legendary. Your choices are an uphill grind through rough terrain into layered defenses, or slipping your units around one of the two flanks through narrow gaps between entrenched Union units and the map edge. The frontal attack on Legendary is suicidal, so the flanks are the only sane option. The best I've ever done is about 8,000 casualties. It's a boring, tedious, slow-moving slugging match with little tactical challenge. Blackwater Heights isn't much better - your only avenue of attack is across a bridge into a pocket of open space surrounded by dug-in Union troops. This requires liberal use of the Hold command to make sure your units don't stray into the line of fire while you're assembling them once across the bridge. 1st Franklin is significantly easier, but it's again a choice between a fairly suicidal frontal assault or a long march around the flank. Fortunately, however, the Union units have a lot less cover at 1st Franklin so you're able to clean them up quite cheaply once in their rear. And honestly, it starts before that - 1st Winchester is a very similar scenario where flanking is the only sane choice. It does offer a more open flank on the Union left, though, if you're willing to march your troops all the way around. I'd say it's almost like most of the Confederate scenarios are designed to make you lose manpower even when you win.

There's a known issue with Stones River that's very similar - I've captured both VPs before, but was then kicked out and forced to retake them again because the battle moved to the next phase. F11 it, definitely. Chancellorsville and Fredericksburg both have these problems as well. Confederate Shiloh, on the other hand, is a good example of things working the way they should - no units are teleported between phases, the battle just continues. These problems are intensely annoying, especially on Legendary where taking the objectives once is hard enough. They take control away from the player and put the player in an arbitrarily bad position. Speaking of Chancellorsville, if you're playing defense as the Confederates you get arbitrarily kicked out of your defensive positions not once but twice - and the second time your divisions are scattered to hell and gone. This is teeth-grindingly annoying, as well as lethal to the units that get stuck by themselves. Also, my II Corps commander was placed on the far west side of the map at least a mile from any of his units.

Confederate Stones River comes to mind as an outstanding example. The only reasonable way to attack the two Nashville Pike VPs is from the north, which requires a lengthy trip into the NW corner of the map and then across into attack position, or a lengthy trip into the northeast corner of the map and then west across the river (which is risky). You can't even get into attack position until the 3rd phase of the battle - you're forced to wait repeatedly at the map's edge and fend off Union units that catch up to you while you're waiting. During the 3rd phase, the displayed timer runs out quickly, but the real timer is very long. This is helpful, because teasing apart the blob on Nashville Pike North can be a challenge. Once in position, you have to do a slow careful crawl into the rear of the fortifications - anything else invites the remaining blob units on the VPs to wipe you out. At the end of the 3rd phase, if you haven't completely taken both VPs, you're forced out of your position automatically by the scenario and repositioned to the south of the fortifications again (grrrr!!). Just getting back into attack position in the north requires almost all of the displayed timer, but the battle continues afterwards ... but not for as long as the 3rd phase real timer. Another example of timer problems is the Union battle Crossroads. If you try to hold the VP until the displayed timer runs out, you'll be demolished as you get overrun by attacks from every direction. However, if you defend the woods instead, ignore the displayed timer, and focus on thinning out the Confederates, you have plenty of time to beat them up, push them back, and retake the VP. The player seeing this battle for the first time has no idea that's the case, and will potentially sacrifice units fighting a hopeless battle. These problems are especially acute on Legendary, due to the need to whittle down the massive forces opposing you without taking too many casualties. The current displayed timers are ruthlessly short, and when enforced promote brute-force frontal-attack tactics that leave no room for maneuver and subtlety. I would also like to highlight the fact that the player has no idea if the displayed timer matters or not, nor any idea what the real timer is. Many battles permit going beyond the displayed timer, and indeed require it. Cedar Mountain as the Confederates is nearly impossible to win inside the timer, and you'll take incredible casualties if you try. In my most recent Legendary Ironman playthrough, the displayed timer expires at 28:34, but the battle continues for another 20 minutes and ended in victory for me.

Just had a cavalry unit surrender to the artillery unit it had just charged, approximately 500 men remaining to 340-ish for the artillery unit. Seems ... unreasonable.

This. In my 0.75 Legendary Ironman I bypassed Hornet's Nest entirely - drove a wedge right into Pittsburg Landing, stepped on the reinforcements, and held it for the win. You can suicide every unit on the right against Hornet's Nest and not even make a dent on Legendary.

In addition to choosing our general avatar, please let us choose our avatar's abilities, and the initial unit abilities.

Won Confederate Shiloh eariler today on my current Legendary Ironman run. The changes to the skirmisher AI apparently affect cavalry as well, and that plus the carbine changes make a big difference. The Union cavalry is now both aggressive and dangerous. I held them off, but only just, and almost lost one of my cavalry units to a well-timed charge. There is one issue though - cavalry now seems to have the ability to charge through a unit, is this intended? I had it happen a couple of times.

My cavalry units will ... not ... stand ... still. I order them into position. They move away. I order them back. They go back. They move away again. I order them back AGAIN. They go back, and then move away AGAIN. They appear to be responding to enemy units that are too far away to affect them, and not even headed in their direction.