Edwin P.

One change that HC should consider is for the Russian AI (and German & Italian AI) to reclaim tech chits for MPPs if it is in dire straits. What is dire straits? Perhaps if Moscow or Stalingrad falls or if Russia is reduced to fewer than X units or fewer than x cities. Or to strengthen a critical a hole in the line. Definitely something that needs playtesting to fine tune. ----------------- Naturally the opening strategy for the USSR should be influenced by Germany's strength along the Eastern Strength as should their research priorities. [ September 12, 2005, 08:09 AM: Message edited by: Edwin P. ]

HCs addition of a second production and supply center in Stalingrad will go a long way towards improving the campaign against the Russian AI as the human player will no longer be able to half Russian production by cutting Russia in two. The effect on the AI of this change is minimal as the AI would never, in SC1, seek to cut the supply lines between Moscow and Stalingrad. This change will also make it harder for the Axis Human player to win in HvH competition. [ September 10, 2005, 12:28 PM: Message edited by: Edwin P. ]

Off all the AI players, the Russian AI will probably be the easiest for HC to optimize. Why? 1. Its primarily a land battle against Axis forces supported by Axis Air Units. 2. Only two fronts to guard - Western front and the Causcaus area. 3. AI need not consider purchasing naval units or bombers or researching techs related to these. That said I would like to see the Russian player use more than one strategy to defend Russia. Perhaps at the game start the Russian AI considers selecting from among the following: 1) Standard Defense 2) Rapid Withdrawal to a Defensive Line 3) Conquest of Finland and Rapid Withdrawal 4) Conquest of Finaland and Standard Defense 5) Stalin's Raiders - Amphibous unit in the Baltic embarks and lands to seize capital of unoccupied neutral nation on German rear, gaining plunder - I did this once against a human player. It was a most surprising move to him. Then the Russian AI needs to decide where and when to concentrate his forces for a counter offensive: 1) in the North after the Allies invade Western Europe 2) in the South after the Allies invade Western Europe On the research front the Russians have an easier time than the Western Allies in deciding where to focus their limited resources as they can safely ignore naval technology investments. The question is should Russia sometimes build a bomber to disrupt German supply derived from captured Russian cities? [ September 09, 2005, 04:35 PM: Message edited by: Edwin P. ]

Thanks for the update. Much appreciated. "Seemed to help a little" - Thats true, but my guess is that the AI was also weak in tech research and also needed a boost in this area (along with improvements to its operational strategy) to play a stronger game. It will be interesting to see how HC improves the AI and what flexiblity the AI scripts offer. In SC2 I wonder if the Allied AI will, sometimes but not always, launch a Nordic Liberation Campaign if the Axis leaves an Axis controlled Norway weakly defended? [ September 07, 2005, 08:13 PM: Message edited by: Edwin P. ]

From my perspective, at the operational level the AI needs its own routines for losses from Naval combat are more risky and harder to recover from. Building on my prior post; Italian Naval AI 1. If Italy has 3:1 advantage over allied naval forces in Med then it adopts an offensive strategy in this theature and seeks to sink the Allied Naval forces in the Med. 2. If Italty has 1:2 disadvantage vs. allied naval forces then it will adopt a defensive strategy in the Med. Also, the AI needs to decide on an overall war strategy that may or may not emphasize Naval combat. [ September 08, 2005, 10:07 AM: Message edited by: Edwin P. ]

In SC1 I would often play against the AI and use the editor to give the AI advantages - research chits, units or MPPs. It would be an improvement if some of these advantages could be randomly set from the AI setup screen, and it would be best if their was an option for randomly determining the AI settings so the HUMAN player did not know the level of the bonses granted to the AI. Example: In many games I would give the AI a bonus number of research chits as the SC1 AI ignored this aspect of game play, never investing in research as much as the human players. I would like to see AI bonus settings for: Beginner (0) research chits, Intermediate (1) research chits, Expert (2) research chits, and Random (0 to 3 research chits). Example: In SC2 diplomacy will impact game play. I would like to see AI bonus settings for Beginner (0) diplomancy chits, Intemediate (1) diplomacy chits, Expert (2) diplomacy chits, and Random (0 to 3 diplomacy chits). The random setting would make for a more unpredicatable game when playing against the AI. PS: I would also like to see an Intel Tech Level AI setting with options for: Level 0 - No Intel bonus Level 1 - Intel Tech Level 1 Level 2 - Intel Tech Level 2 Level 3 - Intel Tech Level 3 Level 4 - RANDOM (0 to 3rd level Intel Tech) PSS: And if it could be squeezed in a setting for AI bonus units. This would reflect the prewar construction/mobilization of units unknown to allied/axis intelligence. Level 0 - No Bonus Units Level 1 - 1 Bonus Unit (Semi-Randomly selected) Level 2 - 2 Bonus Units (Semi-Randomly selected) Level 3 - RANDOM (0 to 2 bonus units for the AI) For the Allies - imagine starting the war thinking that the Germans have only 3 sub units when they really have 5, an interesting strategic shift in the balance of naval forces that can change the game (if the AI knows how to best utilize its navay). Axis Semi-Random Units: Subs, Engineer, Bombers, Rockets, HQ Units For the Axis - imagine attacking Russia and being surprised that the Russian Army has unexpectedly fortified their positions. USSR Semi-Random Units: HQ, Rockets, Armor, Engineer Thus you have AI settings for: 1. Plunder 2. Experience (0-1-2-Random) 3. Tech Chits (0-1-2-Random) 4. Diplomacy (0-1-2-Random) 5. Intel Tech (0-1-2-Random) 6. Bonus Units (0-1-2-Random) Just a suggestion. [ September 07, 2005, 06:19 PM: Message edited by: Edwin P. ]

I was wondering if current plans are for the SC2 AI to select from a range of strategies to win, or is AI development focused on perfecting one strategy for the AI (as was done in SC1)?

You can move HQs faster by spending MPPS for operational movement. PS: HC I like the movement rate of HQs in SC1, it means that you really have to plan your offensives and stop them sometimes to allow supplies to catch up.

HQ units in SC1 represent logistical supply centers. These centers are not highly mobile.

Yep, getting...near. Near as Halley's Comet due to arrive in 2061.

Agreed, it will be interesting to see how diplomacy plays out in SC2. Can it/should it be ignored? Is it a viable option for both alliances (Axis and Allied)?

Nice and simple. I like it.

So say goodby to the Axis cookie cutter solution? or will the Allies now have the cookie cutter solution. I presume that its too early to tell. Question: Why torch, why not just a simple invasion of France - or do you feel that the Allies need to make the Axis spread his forces out before the D-Day invasion of Europe?

Thanks for the update. It gets more interesting all the time. Has HC implemented Intelligence yet? and if so is it a nice to have or must have tech?

I think that he agreed with you, and entered the extra nonsensical comments for what he would like to see as he was bored at the lab and wanted to liven things up.

Good map of the convoy routes and the larger ocean, now I have a reason to send my Axis subs to the South Atlantic. Question - can the allies allocate MPPs going to Russia between the Northern Convoy route and the Route via Iraq, or is the total split 50%/50%.

Many thanks for the update. Most interesting. It was also interesting to read about the Italian fleet was sent into the Atlantic. It will be interesting to see how the AI handles these options, as the game's complexity (from an AI perspective) has increased by a factor of 10.

Its expensive to move factories underground. In SC2 terms that would be like spending 500MPP (about the cost of a battleship) to give an industrial resource target the ability to ignore the first 4 (?) points of damage from a strategic air attack. Now, the question is - if this option were available would players use it? I think not if the cost to do so is 500MPP+.

I agree, all great ideas. The resign option would be great, so that either the Human or the AI could surrender to the other side. I would like to see the Computer should maintain a Win/Loss Record for each AI level or AI Personality to aid in evaluting player developed AI scripts.

They are generic, but I would like to see mountain tiles that are impassable to armor/mechanized units added to the mix.

Good idea, but you would need to enlarge the scale of the map. Such a change would especially apply to a mod of the Western Front of the American civil war.

Suggestion #01: AI has chance to know number; but not location, of combat enemy naval units (cruisers, battleships, subs and carriers) in North Atlantic, South Atlantic, Baltic, Blacksea, Mediterrean and Red Sea. Novice AI - 0% Beginner AI - 25% Intermediate AI - 50% Expert AI - 75% Genius AI - 100% AI will use this information to guide its selection of an operational strategy for its naval units. Creation of AI naval regions would use the same routine used for defining weather regions, and can be altered as needed for other map layouts. [ August 07, 2005, 09:04 AM: Message edited by: Edwin P. ]

June or July? </font>

I wonder if the AI will be just as varied in its strategies?

RE: Patton vs Rommel 1. Rommel was probably a better tactical commander than Patton 2. Patton was more aggresive than most other American generals and knew how to take advantage of the superior American logistics resources. 3. Both tended to lead their troops from the front, while many other commanders lead from the rear.