Southwest Front Runs Away?

[karonagames]

From the manual:

2.2. THE GAME
Gary Grigsby’s War in the East is a turn based simulation of the Eastern Front in World War
Two from June 1941 to September 1945. As the Axis or Soviet player, you take the role of
the military High Command to use the forces available to you to execute the conflict at the
strategic and operational levels of war. The game is an “Alternate History Creator” that focuses
on simulating the logistic and command and control problems that the historical commanders
on the Eastern Front had to deal with. It will allow players to explore many of the strategic and
operational “What ifs” that have been discussed by historians and armchair strategists for
many years. As such, economic and research based “what ifs” are not the focus. The game
scale is weekly turns, with a hexagonal map area representing ten miles per hex. Ground
units range from Soviet Corps and Axis Divisions down to battalion level support units and
air units at the squadron and air regiment level. Units are all controlled through a series of
Headquarters units, each with individual leaders, which represent the chain of command up
to the High Command (OKH and STAVKA) level. Combat is conducted through an automated
tactical system that models the action down to the individual aircraft, Armoured Fighting
Vehicle, and infantry squad.

When someone gets to September 1945, then we can judge whether it a simulation or not.

[Redmarkus5]

I'd love to hear what Gary has to say about that! We'll decide if it's a simulation when somebody gets to 1945? It's not designed as one or the other???

[Redmarkus5]

Is this a what-if, a design flaw, a bug or an AI trick? German AI (set to Impossible) decides that Riga just isn't worth the effort...

[Redmarkus5]

Sorry if I sound rude, but I am frustrated.

[karonagames]

We are trying to understand your frustration. You have highlighted issues with the AI, and within hours a patch has been produced in direct response to your concerns. How many other companies have been that responsive to your concerns?

I think we will continue to respond to your concerns and the obvious issues that your AARs are showing up, but I think a little bit of patience on your part would be appreciated. There will be no more patches until the New Year, and I have no doubt that you and the other players will find other issues that will need to be addressed during that time, but a fix is unlikely to be forthcoming until mid-January, so there will be very little opportunity to alleviate your current frustration.

[Flaviusx]

I made a bug report on this Riga thing. We're looking at it.

Never seen it happen in my own games because, well, I abandon Riga early on. I also don't play with the Red Army sitting on its butt for two turns. You've generated an unusual set of circumstances here and created an unusual result.

[blastpop]

The sky is falling... [:D]

With that said, in game development the game gets tested many times. When that same game gets released it quickly accumulates a lot of play, more than it was tested in terms of total turns and games played. Wargamers, tend to put things through their paces and many may not find any problems. But there are a few who being new to a game or just plain luck do something or a combination of things the testers never did/thought of.

I am sure the dev. team will take care of these issues.

[ComradeP]

I have made mention of the AI not trying to take Riga several times, although not as an issue to be resolved. My stack had a defensive value of over 100, so it was logical the AI made no attempt to attack it with a handful of divisions, but it was odd that it didn't try to attack even with larger concentrations of forces.

[KenchiSulla]

Very nice to see this kind of support guys, appreciated (even if my main focus will be PBEM)...

[notenome]

Am I the only one who thinks that this is actually kind of historical? Not for 41, but for 42. After the Second Battle of Kharkov, the Germans had a pretty clear drive until they reached Stalingrad, then got slammed on there flanks (major simplification). What the AI is doing, even if its a bug, can be seen as a feigned rout, pulling away, and then slamming back hard on the Axis player. Everyone who posted the issue has then posted that their spearheads got cut off a couple turns latter.

[Toby42]

Don't let one person (player) rush you into making changes that don't have to be made?

[Redmarkus5]

ORIGINAL: BigAnorak

We are trying to understand your frustration. You have highlighted issues with the AI, and within hours a patch has been produced in direct response to your concerns. How many other companies have been that responsive to your concerns?

I think we will continue to respond to your concerns and the obvious issues that your AARs are showing up, but I think a little bit of patience on your part would be appreciated. There will be no more patches until the New Year, and I have no doubt that you and the other players will find other issues that will need to be addressed during that time, but a fix is unlikely to be forthcoming until mid-January, so there will be very little opportunity to alleviate your current frustration.

I'm not frustrated primarily with the bugs. I'm frustrated by the insistence of a few hard core testers that these are NOT bugs. Here's a list of my observations so far:

1. AI runs away in the south during GC (for a large number of players, not only the idiotic Redmarkus). Acknowledged as a bug and reportedly fixed in record time. Great! Testing it again now...
2. Soviet GC AI won't defend many of its cities. Confirmed as a known bug "for many months" by Comrade P in another thread.
3. Axis AI bypasses Riga. Seen by Comrade P as well. Maybe a bug, maybe not?
4. Transport aircraft not flying missions. Reported by a couple of people, I believe. I'm not sure of the status.

As you say, I'm sure I will find more. I'll continue reporting them - sorry.

If what I'm reporting are design features, then keep them and don't fix them. If they are bugs, let's get a straight answer and a commitment to fix. January is fine with me and the effort IS appreciated.

[Redmarkus5]

ORIGINAL: ComradeP

I have made mention of the AI not trying to take Riga several times, although not as an issue to be resolved. My stack had a defensive value of over 100, so it was logical the AI made no attempt to attack it with a handful of divisions, but it was odd that it didn't try to attack even with larger concentrations of forces.

My guess is that the Axis AI not taking Riga issue and the Soviet AI abandoning Riga, Minsk, Kiev and Odessa issues are linked. There's something weird about how the AI treats city hexes. As the game involves a lot of drives to capture specific cities (or am I mistaken?) this would seem to be a significant issue.

I know that Gary understands me, even if nobody else does, or he wouldn't have prioritised the southern front fix.

[ComradeP]

There are some issues that keep reappearing, disappearing and reappearing again. There's not too much the programmers can do about those aside from playing whack a mole with them until they discover the root cause.

The way the AI defends cities is one of those issues. Sometimes the AI heavily defends a city, sometimes it gives it up for free. The only persistent problem is that the attacking AI isn't defending cities, but it isn't clear what part of the routine is causing it do so.

The transport mission issue seems to have been some corruption introduced in one of the beta versions, it's not something that was there in a pre-release candidate version. I haven't seen it before in any case.

As to Riga: keep in mind that it's a port (and won't be isolated until Leningrad is captured), so maybe that's the primary reason the AI is less inclined to take it. However, it does seem willing to take Odessa. Why it often just runs away from some city hexes is still beyond me, but at some point the reason will probably be discovered.

The AI can be told to defend a certain area, but it doesn't always defend it efficiently. Improvements are constantly being made, though. The AI's already more impressive than in August when Flavio and me started testing.

[Redmarkus5]

ORIGINAL: ComradeP

There are some issues that keep reappearing, disappearing and reappearing again. There's not too much the programmers can do about those aside from playing whack a mole with them until they discover the root cause.

The way the AI defends cities is one of those issues. Sometimes the AI heavily defends a city, sometimes it gives it up for free. The only persistent problem is that the attacking AI isn't defending cities, but it isn't clear what part of the routine is causing it do so.

The transport mission issue seems to have been some corruption introduced in one of the beta versions, it's not something that was there in a pre-release candidate version. I haven't seen it before in any case.

As to Riga: keep in mind that it's a port (and won't be isolated until Leningrad is captured), so maybe that's the primary reason the AI is less inclined to take it. However, it does seem willing to take Odessa. Why it often just runs away from some city hexes is still beyond me, but at some point the reason will probably be discovered.

The AI can be told to defend a certain area, but it doesn't always defend it efficiently. Improvements are constantly being made, though. The AI's already more impressive than in August when Flavio and me started testing.

Many thanks for this coherent answer. If I list any bugs that don't make sense to you as bugs, please feel free to tell me!

[Joel Billings]

My guess on Riga is that it's not on an AI objective list, and it's never been a problem before. Just a guess. It should probably be on the 18th Army's objective list. If it is, and it's being bypassed, that would be surprise to me. It's on the bug list to check for. Gary just told me he thinks it's on the objective list and since it's a port it's supposed to not be bypassed (since it is not isolated). If you could send me a save at 2by3@2by3games.com at the beginning of the German turn where the 18th army is clearly going around Riga, that could help.

The fact that the AI does not defend some cities strongly is not in itself a bug. The AI is programmed to defend some key places, but it does not like to get pocketed and will pull out of most locations if threatened. There is a problem though that I see when units that are pocketed will move from a good city/urban hex even though it can't get out of the pocket, making it easier to kill. I've got that on the AI bug list.

Not sure about the Transports not flying problem. There are a few items here.

1) The current AI air transport function will put some planes in night mode, and the player must put the groups back in day mode from them to fly. We plan to disable the AI air transport function since we don't think it is doing much now anyway and is causing this confusion with little to no benefit. Down the road if we can get the AI to fly these missions for the human player in an intelligent way, we'll re-enable it. For now our suggestion is don't tell the AI to handle your air transport missions.
2) There is a problem we can repeat on turn 1 of 1942 where some JU-52's don't get listed as available for a mission when they should be. Since it is repeatable we will check into this and hopefully fix it. There were plenty of other groups listed, so it might be an indexing problem when there are lots of groups available, and if so, would not explain why no missions are flown (we think item 1 is the cause of most if not all of these reports).

That's the status.

[Redmarkus5]

thanks Joel - will email you shortly.

The AI defending cities issue goes a bit deeper. For example at Mogilev. Several screenshots from other AARs and also my own show Mogilev being left empty when the AI's front line runs right through the city and there are numerous units on either side and behind it.

Minsk or Lvow, OK. The front is in chaos, I accept.

Here's really bad ss (zoomed out) of the last time I saw this. When it happens again I'll grab a better one.

[Redmarkus5]

Sorry - my save games don't seem to go back that far, although I have autosave on... Will look out for it happening again.

[Joel Billings]

Mogilev is on the wrong side of the river.

[Redmarkus5]

OK, but it's a city - much better defensive terrain than a river if used correctly. I understand the computer logic of what you're saying, but not the military logic.

It so happens that Stalingrad is also on the wrong side of the Volga. What happens when we get there?