First Game Mike13z50

[Mike13z50]

He finally broke due to panic just before nightfall and I was able to run down three brigades, including Wheat's Tigers. Here you can see the turn he breaks, notice how the AI manuvers to keep the engagement range he wants. If he could catch a brigade in disorder and in the rear he would charge, I lost most of three brigades that way. Plus he avoided my artillery, the second battery only got about 4 shots off.

[Mike13z50]

Once again the AI attacks with inferior forces and nearly routes me. Another pyrric victory for the Union, I'm beginning to feel like Grant.

[Mike13z50]

I really, really, really can't emphasize how much better the AI is in HW in this version. I still haven't "lost" a battle, but in the last three I have out numbered him and been on defense, and took more casulties than I dealt. I didn't post the defensive battle of Richmond, where Halleck's Corps had fortress versus on CSA divison, that was pretty easy.

Something I hadn't seen before, I guess because this was a decisive battle, foriegn support changed as a result of the battle, plus I gained national will.

In the rest of the war, I am besieging Memphis, Louisville, and Frankfort all at the same time. I've got one gunboat and three building. The advertised "easy" siege of cities isn't a cakewalk, in two turns I took about 15,000 casulties, most in a normal siege against Memphis. I don't think I can successfully "attack the fort" without artillery, and all three batteries are in the east.

In the immortal words of Chris time for some[>:][>:][>:]

[Mike13z50]

ORIGINAL: Mike13z50

He finally broke due to panic just before nightfall and I was able to run down three brigades, including Wheat's Tigers. Here you can see the turn he breaks, notice how the AI manuvers to keep the engagement range he wants. If he could catch a brigade in disorder and in the rear he would charge, I lost most of three brigades that way. Plus he avoided my artillery, the second battery only got about 4 shots off.

Now this is strange, I just captured Wheat's Tigers again. Also, when tabbing to do screen captures I lost the ability to functionally open containers...they would open, but on the desktop not in the game window. I used Rasmus tactic of "surround rinse and repeat" once the AI was broken down, and was also able to pin six brigades in a mountain hex, though somehow the Jeff Davis Battery magically teleported out.

[Gil R.]

That attitude level for Gov. Morgan looks like a problem, too.

[Mike13z50]

Gil,

Should LU reappear? I captured Wheat's Tigers in a previous battle (though looking at the screen capture it didn't say it surrendered) and then captured it here again. At least it appears to be gone for good. I haven't had much time to play this week, played a couple of turns tonight. The CSA keeps attacking Richmond, now he is down to one brigade. I've had to halt the advance (April 62) since all units are dreadfully understrength. Up to 7000 replacements a turn, but still down 70,000 men. If I disband a brigade in a division it will distribute its strength to the rest of the brigades during the turn, right?

So far: Captured Richmond and Memphis, besieging Louiville.

Jacksons divison is south of Memphis. Van Dorn's corps is stooging around in Kentucky, and the remnents of the ANV is hanging out in Petersburg. Like others have said, the AI seems to put to many brigades into garrisons vs field armies.

Did Erik get a chance to analyse the three empty armies saved game?

Mike

[christof139]

Should LU reappear? I captured Wheat's Tigers in a previous battle (though looking at the screen capture it didn't say it surrendered) and then captured it here again.

Wheat's Tigers, all one Company of them, are indestructible because they eat the breakfast of Champions!!!, Wheaties!!! [X(][8|][:)]

Chris [>:][>:][>:]

[ericbabe]

ORIGINAL: Mike13z50
I'm not sure if this is a bug, or an AI weakness. The AI moved their coverning force (Magruder) into Kentucky (which then joined the CSA) and left Richmond uncovered except by the Fort. I rolled the AOP into the James River and took Richmond in Dec1861. The quirky thing is, that he then showed up with three empty armies, and two understrength corps and moved Magruder back into Virginia. I'll save the game if you want to take a look at it. Attached in the strategic map.

I may try hard-coding AI behavior to keep the biggest army within a radius around the capital. It does seem to tend to send AoP/ANV off wandering when it doesn't have any tantalizing options in the VA theater. The empty armies/corps are simply happening because the AI rendezvous all its newly created mil-groups at its capital.

[Erik Rutins]

ORIGINAL: ericbabe
I may try hard-coding AI behavior to keep the biggest army within a radius around the capital. It does seem to tend to send AoP/ANV off wandering when it doesn't have any tantalizing options in the VA theater. The empty armies/corps are simply happening because the AI rendezvous all its newly created mil-groups at its capital.

Yeah, I think adding some kind of weight to the National Capital beyond what any other city gets will go a long way towards resolving this.

[Gil R.]

Brilliant idea, and one I would never have thought of. That's why you're the programmer and I'm the guy who tries to figure out which special ability to give the Cow Cavalry...

[Mike13z50]

I'm still slogging along on my first game, Union, July61 scenerio (options on first post).

For the first time I called reinforcements to a battle last night, and none of the generals of the reinforcing corps showed up, I thought that was fixed. The brigades did have their attributes and abilities intact which was nice, but the generals need to go with their units.

In another battle it became clear that the HW deployment algorithm needs to be tweeked.

I had brigades from the 6th division scattered all along the front. Divisions should deploy contigious.

In the battle I had the Army of the Potomac with 1 Corps of three divisions, and two additional divisions. There was also an independent Corps of three divisions in the province.

If we are going to have higher headquarters doesn't it make sense for the subunits to deploy contigious to each other?

If you have multiple armies each army should deploy in defined space, all the army's subunits should deploy next to each other, and all a corps divisons should deploy next to each other. Generally a division deploys in a clump of 5-7 (CSA+cassion), a Corps should deploy in a clump of 18-21 and army etc.

One of the things I enjoyed about TalonSoft's ACW series was the moral boost from keeping brigades of a division in proximity, something to think about in an upcoming patch.

To recap my current game, I took Richmond in summer 61. Have been fighting multiple battles against CSA in and around James river. In the spirit of the beta test I'm going to keep playing, and give the south time to rebuild. I will limit offensive in 1862 to taking the rest of Virginia, Kentucky, and Tennessee and establishing the blockade.

Further posts when I notice anything interesting. If as has been reported elsewhere the south AI puts most new brigades in towns/forts should be a quick game. One thing I do like, is the AI seems to spend a lot less on new fort construction and garrisions the forts it does build.

[Mike13z50]

I was trying to stay quiet and play with the production system for a while, but the AI decided it was time to attack. He suddenly appeared with new divisions in both the east and west.

In the east he attempted to head up the valley, I don't know where he was going because on the next turn I took him out.

[Mike13z50]

In the west he tried to push me out of western Tennessee, but didn't bring enough horses. Literally, he had four good cavalry brigades which made for an interesting fight. This was the last gasp of the Confederacy, as I took Lexington on the next turn and he surrendered. Even though I was on "Total Victory" the 32 point difference forced the end.

That should be doubled to a 64 point difference, like the VP total is doubled for Total Victory option.

[Mike13z50]

The CSA surrenders in Oct 1862. There were a total of 21 battles, 11 of which were in Fredricksburg or James River. The AI was very aggressive both strategically and tactically. The major flaw in the CSA strategic AI (as has been posted elsewhere) is he over garrisons. The major flaw in the CSA tactially AI is predictable left flank attack, and his tendency to allow you to surround and destroy individual brigades as he continues to shift left.

I consistently was able to achieve numerical superiority at the point of contact, the main exceptions where in the 2nd and 3rd Battles of the James River, in which the AI was forced to attack my (captured) fortress line. In between the two battles I upgraded the cannon in the fort, and it seemed to do more damage. Next time I have a fort defense I'm turning the damage report (I think it is called that), the new feature that tells you what happens on each volley, and see what kind of guns the forts have.

[Mike13z50]

This was the location of known armies at the end of the game, wait unitl you see the next screen.

[Mike13z50]

There is no obvious scale to the strength graph, but you can see that only recently did I outnumber the CSA. As you can see in the battle report screen I almost always had more strength in battle, and the armies screen shows the AI with only 4 understrength containers. Yet the strength report show he had almost equal mobilized manpower.

The difference, only my frontline cities are garrisoned (I mobilize those qual2 garrison bde's and use them for sieges).
The AI has heavy garrisons throughout the country. Of course the south does need costal garrison, but as has been said elsewhere the build routines and strength distribution algorithms need to be tweeked.

I experienced no crashes or lockups during the game. The outcome of the game was distorted by my early victory in Fredricksburg and then capture of Richmond when the CSA left it uncovered. As long as the south hold Fredricksburg in strength the Union is forced to garrsion both Washington and Annapolis in almost equivilent strength, which limits the norths offensive options in the east.

The one major feature change I would like is adding the sortable leader screen, with abilities. I tried keeping a spreadsheet but it was to much trouble.

Nextup: Southern Steel

[Gil R.]

This is some of the most valuable feedback we have received -- thank you!

I believe that you have clearly demonstrated that the AI needs to be a bit less garrison-happy. I think this is something Eric should be able to tweak, perhaps even for the next version of the beta patch.

Incidentally, regarding the quick taking of Richmond, we're probably going to bulk up James River by adding some starter fort attributes and artillery, which should allow the province to hold out a bit longer. I take it that you would approve of such a change?

[Walloc]

I can only agree with Mike's findings. I'd say less garrisoning in both cities and forts. Some times it seems 60 70% of its forces are in garrison.

I've had to 2 types of games.

1st ANV decides to leave N. Virginia and go to Kenctucky or Tennesee.
Somes times with no retreat results some times not. In that case there is lots a fighting west. Still over garrision but u dont feel it a much. But as i siege his cities and forts East. I find them filled with good units. Even LUs.
It becomes a mob up party.
He has containers so if it just toke out all his units easily 70 80k there would be a fight and a more "fun" game IMHO. Its essential IMHO that it also forms army sized units from all these garrisons.

2nd ANV stays in NW. I usually dont from fighting ANV until 63 and i have upgraded my army to fight it on more even terms offensivly. I do elimimate the single divisions it some try and send to West Virginia.
In the west the AI has some spred division its never really try assemple a real army.
I go around smacking those division and since i heavily outnumber his single division i usually capture almost all of his units.
Do that too the 2 3 seperat divisions that is out west and the city mobbing up starts.
Again i find lots of units. Yes taking 3 forts slow me down time wise, but i win eventually and the time "bought" seemingly isnt used for any thing. So its less "fun" than doing some real fighting. Again it has many many units in garrisons.

What i would like to see if the AI that do build containers take the garrisions out and just as importandly strategicly is tweeked to make army sized units. All those single division is too simple to destroy in detail, IMHO.

Another solution in time could be that in HW, unlike QC. The AI doesnt run until beaten.
Problem is if im heavily outnumbering the AI that essentially means its all but destroyed.
When i fight army vs army i might capture some unit but far from all the army. So it lives to fight another day. Captured brigades doesnt come back. [:)]

I know its not going to make it in the gold patch, but for later. I hope Eric could program the HW AI to retreat if heavily outnumbered. At leased in some circumstances.
If in a no retreat situasion strategicly it should if programmingly possibly stay and fight.
Btw this would only work if strategic retreat rules are keept as it is now. Meaning u can surround him and have no retreat situasion. Else he just alwasy retreat and u never get any where.

Those 2 thing would greatly add to the game IMO and are glad to hear the first is being worked on.


Kind regards,

Rasmus

[Erik Rutins]

ORIGINAL: Gil R.
Incidentally, regarding the quick taking of Richmond, we're probably going to bulk up James River by adding some starter fort attributes and artillery, which should allow the province to hold out a bit longer. I take it that you would approve of such a change?

That might help, as long as we also encourage the strategic AI to focus more on defending its National Capitol or at least staying close to it. I also think that reducing the over-garrisoning that started late in the internal beta will really, really help.

[Mike13z50]

Incidentally, regarding the quick taking of Richmond, we're probably going to bulk up James River by adding some starter fort attributes and artillery, which should allow the province to hold out a bit longer. I take it that you would approve of such a change?

That should help, but what would really help, though unhistorical, is starting with Fredricksburg fortified. You really have to have a good army to take out a fortified province that is defended, and the Union does not have that until 63.

Freddy is the decisive point strategically. Once Freddy falls, the South is toast, instead of the Union having to garrison Maryland, DC, and PA the South has to garrison Richmond, Petersburg, and Lynchburg. If the Southern AI has its own routines set Fredricksburg higher even than Richmond.