My problems with simultaneous

[PvK]

Tsaarx, thanks for the tests. I agree your changes would be a way to make it more consistent between game types, which would be good for mods which are trying to get the same effect in both game types.

DavidG, yes there are three issues, I think, which apply equally to turn-based except I think maybe turn-based players are more likely to issue shorter move orders rather than long-range ones.

The first issue is that the blue line is just a possible course, and can even change different times you look at it. This is because the movement AI makes some arbitrary decisions, mainly which side to go around an obstacle, and it does those each time it thinks about the question.

The second issue is that the AI uses different reasoning during actual movement than it does when drawing the blue line. This seems to only apply to multi-system movement, and the decision of which warp points to use. The blue line will suggest one route, but the ships will actually choose another.

The third issue is that the AI (Last I checked it in any detail, which was a couple of patches ago) doesn't do very well with created warp points, human-made maps, and perhaps some modded quadrant types. It can also not realize it needs to take a warp point which looks like the wrong direction from the system it is in.

There is a fourth cause which generally makes sense, but can cause unexpected results, which is that temporary events during a turn can cause the movement AI to do unpredicted things. Mainly, it reconsiders each sector move, and if there are moving hostile ships blocking any choke-point, even temporarily, along the path, ships with long-range movement orders will look for alternate paths. If the obstacle is just an enemy ship zipping through a warp point, this could cause a fleet to take a detour for a few sectors, then revert back to the original course.

All of these can be worked around by giving ships more specific orders about where to go, though that is of course more work.

PvK

[PsychoTechFreak]

Thank god, I am not alone with my trouble.
But I guess it is not easy to compile all of these small issues into savegames for an enhencement report to MM, at least I have my problems with it sometimes.

PvK, if you do not need savegames for your reports, could you . Ahem, now I recall the other open issues that I still have not tested with patch 3 yet. Timeout.

[PvK]

Heh. One example: I was just looking at a game using patch 3, but using my "Rich Clusters" quadrant type from Proportions. There were several cases where the blue line would not show an intended course to the destination, as if the ship wouldn't go anywhere, but then during the turn it actually did take a course, and eventually made its way there.

PvK

[PsychoTechFreak]

Just a guess, in case of the blue line does not show up in the first attempt, is the course going through unknown systems?

[DavidG]

Originally posted by PsychoTechFreak:
Just a guess, in case of the blue line does not show up in the first attempt, is the course going through unknown systems?

Not in my case. I just tested a little more and it seems it is not a case of it just not drawing the line through the one warp point. It does for a short range course. But for a longer one it sometimes does sometimes doesn't depending on the destination sector. After a bit of clicking it seems there are two roughly equal paths to the destination sector so maybe that is what is confusing it.

[PvK]

No, the course goes through known systems. Apparently it has something to do with the number of warp points, like if the most direct-looking warp point turns out not to be the shortest, it breaks the blue line algorithm, but not the actual pathfinder.

PvK