forestrouse
Posts: 35
Joined: 5/18/2020
Status: offline
|
I had reached my 10 post limit so could not finish my observations point by point. Here is the completion of the original post
Point 1 – I marveled how well the AI determined which hexes in the game are important and organized multiple Corps/Army attacks on specific hexes to achieve a goal. It used multiple tactics for breaking a hex. When available, it attacked with air power first, but even without air power, it would ensure that multiple units attacked a single unit. It did this by swapping or by moving the original unit out of the way to an important position before bringing in a second unit. It kept its air units from becoming fragile by periodically reinforcing units in a disciplined manner.
My only critique of how the land units are handled is that oft-times the AI pushed ground units into exposed places too aggressively. I made a good living as the Soviet player by picking off exposed tank units. The Axis AI picked off US/UK tank units at the French/German border in basically the same manner. Additionally, air units were needlessly exposed in front line positions. I was able to destroy a number of air units with relatively painless ground attacks. Given that Axis MPPs are precious (even with a 20% bonus), losing relatively expensive air units is costly in both MPPs and time to rebuild them.
Point 2 – the organization of amphibious landings is generally excellent. The AI carried out its equivalent of Operation Torch by cleverly using long range amphibious transport from the US. It invaded Sicily with a sufficient self-sustaining force to capture Sicily relatively quickly. The equivalent of the Operation Overlord was a marvel to observe. The AI coordinated the use of air power to destroy units in very specific hexes, and then followed that up in the same turn with parachute and landings to create a viable beachhead.
However, the AI also tried the equivalent of the Anzio and Solarno landings in a very odd manner. The Anzio landing was done with two unsupported infantry units and the Solarno landing was done with a single headquarters unit. Both were snuffed out fairly easily by the Axis AI opponent. The loss of the HQ unit meant that the force on Sicily had no coordination of supply. The subsequent attempt to drive up the boot of Italy was a failure. Corps after Corps of UK infantry units were destroyed two hexes from the crossing by a German engineer unit in conjunction with the use of Italian bomber (not sure if it was tactical or medium). The advance from 1943 to the end of the war in February 1946 was a total of 5 hexes from Sicily.
Point 3 – The AI Axis strategy of trying to knock to Soviets out early and the AI western Allied strategy of first holding North Africa then using its air and sea power advantage of launching an invasion of France is credible (and historical).
The only reason why Moscow did not fall to the AI in the game I played was that I fortified Moscow in the 1939-1941 time frame. So even though the AI took Leningrad and for a short time Stalingrad and injected a force into the Caucasus mountains, the fortifications forced the AI to attack only one hex effectively at a time when it was trying to take Moscow. This allowed me as the Soviet player to update the tech enough so that once attrition set in for the Axis side, the USSR could overrun the higher value artillery and tank units and begin the inexorable march to Berlin.
As an aside, I tried Sealion as an experiment against Allied AI in a game and the AI did a pretty good job of bottling up a supported 11 unit invasion including an Axis headquarters unit. The US entered the war early and pretty soon 11 units was not enough to get to Manchester. That is not to say that a better organized Sealion by the Axis could not work (I still have some rough edges in my play especially when it comes to supply), but that Sealion isn’t a slam dunk Axis strategy.
There is a caveat to the AI strategy of stopping the Axis from capturing Cairo. The UK AI strategy to stop DAK was to commit the entire UK Airforce to North Africa. This decision by the AI was unrealistic and had a deleterious effect in the battle for control of the North Sea and the Channel (more on this in point 5) as probably should be expected.
Point 4 – One of the reasons why the AI Axis air power stays competitive for so long in the game is that it builds out the minor power fighter units to what must be the limit. I was always wondering why even at the end of the game the Axis was able to intercept air missions. I should have figured this out earlier, but the tradeoff of spending MPPs on minor power fighter units as opposed to buying more German ground units strikes me as a great choice by the AI.
Since all major powers have basically the same build limits for all units, the Axis can only remain competitive in the air by improving technology (and eventually the Allied powers will catch up), or buying these non-German units. Even though Italy is relatively MPP poor, the Axis AI seems to still feel that the slow purchase of Italian air units is superior to anything else the Italians can do. That is probably basically correct.
Point 5 – The operation of the naval units by the AI is generally good. It recognizes that transports require a screening escort force. Only a couple of Allied AI US transports were picked off by Axis AI submarines.
However, in the battle of Axis AI versus UK AI – the UK lost most of its capital ships. Until the US AI naval fleet made its presence felt, even the English Channel was not safe for Allied ships. Even after the equivalent of Operation Overlord, the Allied AI lost shipping in the Channel to both Axis AI submarines and battle cruisers.
The main reason, I think, is that the use of scouting ahead of capital ships is not always followed by the AI. Numerous times I observed that Allied aircraft carriers were used almost as-if they were destroyers and would run into the zone of control of an undiscovered Axis submarine or destroyer. This would prevent the carrier from moving back to safer waters after they carried out its attack (or attacks). The Axis AI in its turn would then concentrate an attack using subs, e-boats and then heavies (heavy cruisers, battle cruisers or a battle ship) and sink the carrier. Other AI UK capital ships met the same fate.
Another aspect of the problem stems from the fact there was no UK Airforce (as mentioned in point 3) to provide a semblance of control around the British Isles. The UK AI also did not invest in even a single Maritime bomber (at least I don’t think so) which also would have helped in enforcing an area of control around the British Isles.
Finally, AI allied destroyer ASW was not uniformly updated as the game progressed. While I did not keep scientific track of which destroyers were at what ASW level, many Canadian destroyers, in particular, were not periodically updated in ASW level. As a result ASW level 0 destroyers would run into German submarines with ASW level 3, 4 or 5 from the middle of the game onward. This meant that German submarines, for the most part, were not destroyed.
I checked the research levels for all belligerents at the end of the game and it turns out the AI Axis side did this without researching Naval Warfare. Bottom line, the AI Axis navy did far more damage than should have been possible.
As to Sugar's point on AI buying of naval units: I agree - this is also what I observed. It could well be that the AI purchase optimization does not like the relatively high MPPs required to buy naval units and time it takes for production of the unit. The goal of the game is to take and hold territory. The purchasing optimizer probably assesses that ground and air units have a far more of a direct effect of the overall attainment of the game goals than naval units. Naval units basically only affect taking of territory by blocking convey routes and by transport escort to move masses of units across bodies of water. Perhaps the optimizer needs to better account for the effect of naval units when it is making it's purchasing trade-offs. However, the original sin is losing high value capital ship assets in the first place.
I will add as a note, probably only the US could afford to and carry out rebuilding of capital ships - especially carriers. I don't think that Britain, Italy, Germany or the Soviet Union even tried to replace the capital ships they lost (certainly not even close to 1 to 1). The Germans went almost exclusively to light ships (subs, e-boats), Italy didn't replace the battle ships it lost in the Mediterranean, Britain certainly didn't replace the Hood. Maybe it is because service leaders saw how vulnerable ships are to air attack and thought that investing in air power was a better trade-off.
< Message edited by forestrouse -- 6/23/2020 6:45:31 PM >
|
Printable Version
Observations on the AI - 
] 
New Messages
No New Messages
Hot Topic w/ New Messages
Hot Topic w/o New Messages
Locked w/ New Messages
Locked w/o New Messages
Post New Thread