Impressions after last update

[Tordenskjold]

First of all, this game is getting more and more awesome! Easy to play, hard to master. The last improvements has enhanced the game tremendously!

I have played through as the Allies and took Berlin in november-44. I gave the Axis one step advantage and that was not enough. Next time I will give them two step advantage in the setup meny. A part from that I divided the opitons as equally I could to get an even game. As mentioned, it turned out to be uneven.

I also observed some AI behavior that gave the Axis disadvantages. The most notably was:
1. The AI try to crush forces that are not important and even turns the opposite way of the main target (Paris in this case) to crush an enemy unit. It should rather stand passive and block the way.
2. AI forces in well defended cities and fortified positions attacks adjacent hexes instead of just being passive. It woul be much harder to destroy the unit if it just defended.
3. The AI use time an effort to take Switzerland. It seems a vaste of time and resources. AI also attacked Sweden spending huge amounts of troops but still unable to take Stockholm.
4. On several occasions it seem the AI "forget" to move fleets. I have seen Transports, Subs and Surface fleets just standing there until sunk. Also ships in port seems to be forgotten. Just to be destroyed when the city is conquerd.
5. The Italians seems to be just a marginale force. Should they not have a lot more units? They gave up late 42 without putting up a fight.
6. When the Allies reconquering a nation such as France by taking Paris, should not the map of France change so that the map is entirely an Allied nation? As of now, places that units have not went through stays in the Axis colours. When a country surrenders like Rumania to Soviet Union the entire country are marked as Allied. The same problem apply to Norway, Belgium, Holland, Tunis, Algeria and Morocco.
7. And finally a wish. Would it be possible to use production points to try to persuade countries to join ones alliance?

Thanks for a great game!

[lhughes]

How would you say the AI compares to the original vanilla game AI? So the current AI understands all the changes?

I found the original AI pretty weak..interested to hear if it has improved.

[Plaid]

As far as I know nothing has been done with AI, so it simple can't play with new rules properly.
(It also can't play properly with any rules, but its another problem).

[Bern]

This is correct. The changes are made in order to improve person to person play. I think this is a pity. I always play as the Allies and don't really have the time or patience to play email games.

Given that the Axis AI is poor (not really the devs fault - it is extremely difficult to program an offensive AI) I have been playing around with ways of enhancing the chances of the Axis AI. I want to do this without ramping up the Axis strength or production capacity etc. The only way I can see to do this is to restrict the strength of the minor powers. At present I have a situation where the Axis AI is able to achieve the fall of France within a very reasonable timeframe which does tend to open up the mid game and provide some sort of thought for the Allied human.

I'm not sure how this affects the new French Armistice rule because of a bug which is noted elsewhere. This will be corrected and will, hopefully, add interest.

The next step is Russia. I still believe that the Russian position is too strong for the purposes of play against the AI only - also noted elsewhere in the Forum. The challenge again here is to nudge the balance a bit to give the Allied human cause for a little more thought. Clearly changes to the strength of the Russian forces have a knock on effect in other areas of operation and I have not yet got the balance right. I will keep on trying because I believe this is a great game .

Bern

[rkr1958]

I will keep on trying because I believe this is a great game .

Have you thought about playing hotseat? In effect, you could be your own AI. I know the fog of war and the naval part would be a bit tricky but I think with a bit of tinkering you could come up with a set of houserules that could work. Also, you could flowchart some general strategies and counter strategies and pick them at random. Who knows ... you could even develop the strategic and tactical algorithms necessary to improve the AI.

By the way, it was this type of thinking that was the genesis of the BJR-Mod, which if you aren't familiar with was forefather of the GS expansion. The BJR-Mod was made of data and map changes and a set of house rules which we played by. These house rules were later codified into the CEaW game engine and became the GS expansion.

[Bern]

Many thanks for the suggestion - an excellent one. At present, I'm playing just the early to mid stages of the invasion of Russia as Allies with FoW turned off and trying to analyse the Axis AI play to get a better understanding of its strategic objectives and its path to achieve these. Whilst there is no fog of war, there is plenty of 'mist' in this area, the logic behind which is a bit of a struggle to make sense of.

The intention here was to simply identify areas where a minor strength reduction in Russian forces would improve the position of the Axis. Perhaps I should be thinking more along the lines of your suggestions.

Bern

[pzgndr]

I have made some modest adjustments to GS 2 for Allied play versus Axis AI. Made Switzerland non-DOWable and brought Denmark and Norway in as Axis in April 1940. I also added a few units to the setup; a German tank and tac bomber in Germany and a German and Italian tank in Libya. I'm slowly working through a campaign. Poland fell in Oct 39 and France fell in Jul 40, OK. As of end of 1941 Germany still has not attacked Russia and Axis hasn't done much in Libya. U-boats have appeared in force finally.

I realize no AI programming improvements have been made and aren't likely to happen, but still there are some play balancing options available to make the computer opponent more challenging. Could someone confirm if we can save campaign mods under a different name for selection at new game start? IE, '1939_Axis' or '1939_Allies', etc.? For those of us still interested in playing a computer game versus the AI, not hotseat or pbem.

[rkr1958]

Could someone confirm if we can save campaign mods under a different name for selection at new game start? IE, '1939_Axis' or '1939_Allies', etc.? For those of us still interested in playing a computer game versus the AI, not hotseat or pbem.

I'm not 100% certain but I fairly sure you can save to those names. I do know for sure that that the first four characters of any scenario can only be 1939, 1940, 1941, 1942, 1943, 1944 or 1945.

[pzgndr]

I'm not 100% certain but I fairly sure you can save to those names.

OK I just confirmed this is possible. I renamed the 1939 SCN and DATA files to "1939_AxisAI" so should also be able to create "1939_AlliedAI" etc. Although the old CEAW Editor does not work with GS 2.0, it is useful for identifying hex coordinates, so modders can then add/delete/edit at-start units fairly easily, as well as edit the country data. If we can get the new editor, that would be helpful.

[Bern]

I have made some modest adjustments to GS 2 for Allied play versus Axis AI. Made Switzerland non-DOWable and brought Denmark and Norway in as Axis in April 1940. I also added a few units to the setup; a German tank and tac bomber in Germany and a German and Italian tank in Libya. I'm slowly working through a campaign. Poland fell in Oct 39 and France fell in Jul 40, OK. As of end of 1941 Germany still has not attacked Russia and Axis hasn't done much in Libya. U-boats have appeared in force finally.

I realize no AI programming improvements have been made and aren't likely to happen, but still there are some play balancing options available to make the computer opponent more challenging. Could someone confirm if we can save campaign mods under a different name for selection at new game start? IE, '1939_Axis' or '1939_Allies', etc.? For those of us still interested in playing a computer game versus the AI, not hotseat or pbem.

Yep - I've gone along the same lines, although I haven't added to the Axis strength - well apart from one mech unit in the initial set up. The results are the same pretty much. I've also adjusted the Yugoslav forces which enabled the Axis AI to force surrender in quick time. I believed that this would facilitate the Axis in its build up for the invasion of Russia. Sadly I was wrong.

I've now played a number of games without FoG so that I could see and, maybe, understand the Axis strategy with regard to Russia. Given the significance of this event to the game as a whole, I believed that it was likely that considerable effort would have gone into programming the Axis AI in terms of build up, strategic targets etc. I have to say that I can't see too much evidence of this - at least within an appropriate timescale.

What tends to happen is that following the fall of Yugoslavia, there is quite significant partisan action - I believe this has been criticized elsewhere as being excessive - I don't think it is. However, the Axis AI does seem to allocate a disproportionate amount of resource to eliminate this 'threat to the Reich' which in itself hampers invasion preparations. Additionally and more problematically, there seems to be little effort to build up and position forces for invasion during 1941. I cannot understand the logic here - Europe is subdued and yet there are significant Axis AI units posted all over Germany, units which really should be on the Eastern Front.

It would seem that simply modding is not going to make much difference to that situation and that really the only solution to providing a more effective Axis AI is to code in some strategies to tackle the Russian question. Either that or make Russia Axis-friendly

Bern

[pzgndr]

It would seem that simply modding is not going to make much difference to that situation

Maybe not. In the general.txt file there are partisan rules where the chances of partisans could be reduced. Be nicer if partisans were restricted until USSR is at war. There are also AI rules where Axis aggression could be increased and/or the DOW targeting rules adjusted? Not sure how these parameters work, nor much else about editing the general.txt file. Guidance??

Thing is this file appears to be very general, as it indicates, so if you mod it then it affects any scenario and probably any saved game you reload. It would be nice to customize this file for each scenario but not likely to happen. Workaround is to maintain different versions of the file and then rename/load what you want as "general.txt."

[Bern]

It would seem that simply modding is not going to make much difference to that situation

Maybe not. In the general.txt file there are partisan rules where the chances of partisans could be reduced. Be nicer if partisans were restricted until USSR is at war. There are also AI rules where Axis aggression could be increased and/or the DOW targeting rules adjusted? Not sure how these parameters work, nor much else about editing the general.txt file. Guidance??

Thing is this file appears to be very general, as it indicates, so if you mod it then it affects any scenario and probably any saved game you reload. It would be nice to customize this file for each scenario but not likely to happen. Workaround is to maintain different versions of the file and then rename/load what you want as "general.txt."

Agreed all round. I just wish that the Axis AI was enabled to build up its forces to mount a coherent invasion in 1941 - well before the bad wx sets in. This is the key in my mind to provide a challenging game. Oddly enough the Axis does have in place in good time a quite massive air resource, sufficient to blow holes in a Russian defence. What is doesn't have is any sort of ground strength, neither armour nor infantry.

One of the things I've not yet tried is to free the Axis from invading Yugoslavia. I'll have a look at this next I think. It isn't an entirely ridiculous strategy in any event and it ought to allow the Axis an early opportunity to move its troops to the eastern front. After all, it would have little else to do. Have to say on past performance, I'm not overly hopeful.

Bern

[LOGAN5]

AI seems to be more interested in attacking Switzerland than France, its sends armor and motorized until it has that capital completely surrounded until finally attacking it, by this time you can build and send many British Armored Divisions and pretty much hold France until US and Russia join the war, basically WW2 is over before it even gets started.

[Bern]

AI seems to be more interested in attacking Switzerland than France, its sends armor and motorized until it has that capital completely surrounded until finally attacking it, by this time you can build and send many British Armored Divisions and pretty much hold France until US and Russia join the war, basically WW2 is over before it even gets started.

That is true. However, as has been reported elsewhere, you can avoid that problem by making Switzerland a pro-Axis country. Again as has been said here, just by modding it is possible to allow the Axis AI the opportunity to complete the defeat of the Low Countries and France within an historical timescale. This then allows for a reasonable submarine intervention in the Atlantic - which can be improved - and provides more of a threat to what, in the vanilla game, is the fairly comfortable build up of the UK. On the surface (or under it) things appear to be going well for the AI.

That is the good news. The bad news continues to concern the Axis AI build up for Barbarossa. To me, this is the single most important issue in the game. If the Axis AI gets this right, you have a good game. If not, then the only question is how early the Axis is defeated. Simply by modding it is possible to provide the AI with a significant number of turns in which it has little to do but to position strength on the East Front for a coherent invasion within a roughly historical timeframe.

In my test games this does not happen. As I've said before, the air build up is excellent but that is not the case with ground units. My latest 'experiments' are producing a slightly better situation for the Axis AI but not good enough IMO.

There are some more changes to be made and I'll keep trying, but my overall view is
that there is something within the AI coding which is preventing the transfer of ground forces eastwards. I have to say that it is something I do not understand.

Bern

[Tordenskjold]

I just have to add that I did NOT give the Axis any advantage. I messed up. Now I am in the middle of a new game where Axis is Major. This really gives the Axis a boost and it is great fun to play the Allies now. The Axis are strong in the east and by March-43 I have still not taken Sicily and are far from landing in western Europe. The Axis are close to conquer Sweden and the British have major forces in Norway to try to halt the Axis advance. Weather the AI is better or not than before, I can't tell, but it is surely much more of a challenge than before and for me that is the main thing. Superb game.

[pzgndr]

The bad news continues to concern the Axis AI build up for Barbarossa. To me, this is the single most important issue in the game.

Unfortunately, the solution to this problem (unless the GS developers can do something about AI production priorities) may be to forego the early campaign(s) and focus on Axis AI scenarios beginning in 1941. Likewise, Allied AI scenarios beginning in 1942 or 1943 may be needed to fix the problem of Allied build up in the Med for Husky.

For Axis AI, I'm thinking some status change for Yugoslavia and Greece could help the 1939 campaign (ie, just activate these guys as Axis in April 1941). And significantly reducing partisans to remove that distraction should also help. If Axis AI doesn't have these distractions then Barbarossa should get full priority.

If I had more time I'd mess with it. But I won't have a chance for a couple more weeks.

[Bern]

The bad news continues to concern the Axis AI build up for Barbarossa. To me, this is the single most important issue in the game.

Unfortunately, the solution to this problem (unless the GS developers can do something about AI production priorities) may be to forego the early campaign(s) and focus on Axis AI scenarios beginning in 1941. Likewise, Allied AI scenarios beginning in 1942 or 1943 may be needed to fix the problem of Allied build up in the Med for Husky.

For Axis AI, I'm thinking some status change for Yugoslavia and Greece could help the 1939 campaign (ie, just activate these guys as Axis in April 1941). And significantly reducing partisans to remove that distraction should also help. If Axis AI doesn't have these distractions then Barbarossa should get full priority.

If I had more time I'd mess with it. But I won't have a chance for a couple more weeks.

I've tried out some changes commencing with the Barbarossa scenario. Unfortunately there still seems to be something awry with the AI. Typically by June '41, the build up on the Eastern Front is about a third as good as it should be. There is good air support; AGC is fairly strong tho' positioned a bit too far north. AGN has some units but there is nothing in the way of an AGS ready to invade. There is then some movement but regrettably it is mostly in a Westerly direction !! At this point, there are in excess of 20 units - inf and armour - still within Germany doing absolutely nothing. I can't see that changing short of an overhaul of the AI coding.

I like the idea of changing the status of some of the minor countries, although at the moment I've only got Yugoslavia as pro-Axis. Anyway I've made some changes to other countries and started a new game from 1939 to assess the effect. I also put an extra couple of subs in the Atlantic as an afterthought. That has worked really well - so far, they've devastated the convoys and the UK is having real problems vis a vis production and research.

At the rate it's going, I'd estimate France to fall about one game turn before it did historically. This will give a great deal of time for an East Front build up tho' my fear is that this is where it will all go pear-shaped because of the AI failings as described above. It remains to be seen if the minor country changes will have an effect. Fingers crossed.

Bern

[Rasputitsa]

First Impressions on GS 2.0 - Having played several practice games and started the 1939 Campaign, I am impressed with the quality of this game. If you are looking for a very accurate WW2 simulation, this is not it, but as a Grand Stategy game this is a lot of fun, well presented, with lots of detail and depth.

I hesitate to comment, as others will have done extensive testing on this development, however, here goes :

Playing against the Allied AI 1939, finished Poland easily, attacked Denmark immediately, prepared to attack France early in 1940. Hit Holland first, then moved on to Belgium and straight into France, Paris falls June 7th 1940.

I rejected the French armistice offer and was surprised how resilent the French units were, when becoming Free French under UK control. Britain could not have immediately replaced the command and supply services that these units would have needed to retain effectiveness. I would have expected more drop in effectiveness, or loss of 'steps', which would then be recovered over time.

For practice, I restarted another similar game and accepted the armistice offer and intended to move on to Spain, but changed my mind and attacked Sweden, with 8 turns left till winter. I expected another easy victory and planned that this would make available an overland advance into Norway, with sea operations based in the Baltic and safe from British interference. I was so confident that I was even massing units on the Spanish border, hoping to attack, when air units could be released from Scandinavia.

Got bogged down in Sweden into the winter (self imposed limit not to use armoured units in Scandinavia, due to historical German lack of shipping), also surprised how effective the Swedish fleet unit was, although it stayed in port most of the time (where I didn't seem able to make a naval attack on it), it seemed able to inflict serious losses to my fleet. I had selected 'Averaged results' option OFF, so I suppose unexpected results would occur and this was the effect I wanted. However, as Swedish vessels were mainly coastal defence ships, perhaps the unit should start at a lower 'step' level, as with other minor country units. Alternatively, perhaps it should be represented by a destroyer unit, to reflect that the ships were not full-scale BB.

Spring 1941, decided to proceed with attack on Russia, whist still waiting for Copenhagen to fall, which it eventually did. Southern front of Russia attack was weak because of units tied down in Sweden and never got to attack Norway. However, managed to get just beyond Smolensk when winter arrived and still learning how to eliminate encircled units. Seems like encircled units can draw supply from cities within the encircled area and continue happily with no need to break out. One Russian Mech unit attacked into Poland and was able to stay in supply well behind my lines, by progressively moving from city to city. So I should have garrisoned the Polish cities, but how much supply should you be able to take from an enemy city (with none of your own factories giving supply like Leningrad and Kharkov could) so much to think about and choices to make.

Now up to Feb 1942, US units are flooding across the Atlantic, the English Channel is filling with ships. I have six upgraded U-boat groups and am sinking a lot of transports, but how many are getting through ? I am moving air and other units into France and have beaten off the first landing, but there must be more to come.

This is all against the AI, which I know the mod is not optimised for, but I am having a great time, with lots to think about.

I changed the air unit 'sentry' graphic from Zzz to an X, my heroic pilots could not be asleep. I notice that the 's' key shows on any units (although only acts on air units), this means that I can mark my X on any unit, as a reminder of where future actions are required.

If I had played better, then the results would have been even more realistic, but still quite good.

The documentation is very very good, a 444 page manual for a 'mod' is exceptional, with huge amounts of gameplay tips and hints.

Grand Strategy 2.0 remains true to the original CEaW, with good quality graphics, depth and detail, with a good GUI. The extended choice of game options (22 of them), e.g. with, or without, oil restrictions, random convoy routes, elite units, etc., makes for a great variation in play. Many thanks for sharing this.

[stoner]

I'm not going to complain about the AI. Contrary to some people's thoughts expressed here, I think desiging an AI for a game this complex is an incredibly difficult thing to do. I can consistently beat the AI both as Allies and Axis in original CEAW with advantage set at Maximum for the AI. In GS2.0, I've found that it is possible to beat the AI with advantage set at Moderate, but it is by no means a sure thing. And if advantage is set at Maximum for the AI, I have no chance whatsoever. In this respect, GS2.0 is an improvement in the AI department as it makes it more challenging. But not much more challenging since it still has certain foilables that, once learned, allow you to exploit AI stupidity.

But I do have an issue with something I've experienced new in GS2.0 and wonder if any others have encountered it and/or if any designers who may be on this list have an answer for it.

I am used to seeing bulletproof units when there is a big spread in research upgrades between the opposing units. A unit with a 8 or more upgrade advantage can usually inflict serious losses with little impact on itself. When the upgrades are close, on the other hand, the losses are usually close.

In my AI games in GW2.0, however, I've experienced as the Axis a new development that has me baffled. With Allied advantage set at Moderate, I encounter Russian tanks in 1943 where German panzers have 12-13 upgrades and the Russkies have 13-14 upgrades...and the Germans get annihilated in tank-to-tank battle. And I mean *annihilated*. As in 10:0 total elimination in a single attack. And this is not an anomaly - it happens every time, even with Manstein and an SS panzer unit. I actually have to hide my panzers from these T-2000s and attack with *infantry* which, inexplicably, do about even agianst these Russian super tanks.

Has anyone else experienced this phenomenon? And does anyone have an explanation for it?

[Peter Stauffenberg]

The AI cheats with tech and even more when you select an advantage for the AI. Against human opponents you won't see such a tech gap if both sides put some effort into labs.

When you see tanks receiving lots of damage then the opponent has put a lot of tech into antitank. Every point in antitank will reduce the survivability of the tank. So I suspect that the Russia AI has very high antitank tech.

The way to counter this is to put focus in armor so you gain more survivability. The downside is that armor is the armor tech that will increase oil consumption as well.

Have you put any effort into armor for your armor tanks? If you haven't then your tanks have low survivability. Late in the game when the Tiger II tanks show up then the Germans should have 10 or 11 survivability. It increases further with SS and XP levels. So my experience is that Axis tanks late game are very hard to kill. Tac bombers inflict maybe 1 hit per attack and land units 1-2. So you need 4 land attacks or more to kill a Tiger II