surrender change latest beta (.02)

[DarkestHour]

Latest beta (.02) enables units surrounded, then defeated same turn to surrender rather than rout. Previously, this only occurred if such units were isolated at the beginning of the turn. Big change.

[Joel Billings]

There was no change made that we know of to change the chance of routing versus surrendering. Are you seeing all units that have no where to retreat to surrender instead of routing? I just ran a quick test of turn one and I'm still seeing units rout (other than the forts that always surrender). Can you attach a before and after save so we can try making the attacks and see what happens.

[DarkestHour]

Will attach before and after save once I replicate the scenario which is always happening now. Knowing the 'start the turn isolated to get a surrender' rule impacted my gameplay so I just found this.

[DarkestHour]

Here are 2 examples with the 9th panzer stack and a motorised unit T29. Turn still in progress though it is difficult to do this. Someone coded an excellent AI with few obvious errors(Axis T1 abandonment of Memel enabled a Sov inf xx and arm xx to escape by sea after plundering the port - easy to block that). Anyway, never added an attachment so... will switch to standard post reply.

[DarkestHour]

pre and post surrender events attached.

[Joel Billings]

Thanks for the saves. I ran your save and I see what you mean. Once surrounded, they are in a "hex isolated" but they are still listed as they should as being "in supply" (I have a way to confirm that). In both cases when I attack them surrounded and when I attack them not surrounded, their CV crashes and they lose the battle badly. I don't see any extra combat disadvantage going on due to isolation. However, when they are forced to rout by the fact that they have nowhere to retreat, and they start taking additional losses for that, there may be some factor kicking in that puts them over the edge (the units are not in great shape) that causes them to surrender. In other tests I've just run of units just surrounded (many not in great shape), they are routing out as expected. So will you be able to count on units surrendering just because you surrounded them in the current turn? I think the answer is no. Is it possible. Yes. Is it possible that the heavy snow on the ground is impacting this, I can't say for sure. I don't think so, but without a thorough audit through the code, it's impossible to say.

[DarkestHour]

I found the same with most units routing in this scenario but some surrendering. Later, two of my soviet cav divisions surrendered (separate attacks, different hexes, other defenders in hex retreated) when under direct counter attack. Looks retreat, rout, shatter are now but 3 of 4 options.

[Joel Billings]

Gary has asked if you have an earlier save at the start and/or end of the prior German turn. If you do, please attach here. Thanks.

[Skritshell]

Can confirm is a reoccurring bug. Have noticed multiple times in a server game I am playing of units bring surrounded and surrendering when they should rout out. My friend has been experiencing a similar phenomenon.
I do not yet have 10 posts even though i have been a member for over a year... so pictures are unavailable.


has happened to a german cav regiment from 1st cav and ger infantry regiments as well. Have not noticed it against soviet units but that could be human bias as germans to pocket and soviets to rout out units.

[deaniks]

Ya I noticed the same problem as Skrit said me and him have both experienced this for example this last turn I sent my opponent it's a sever game. I surrounded two mountain div's and they both surrendered the same turn they were encircled here's some picture's

[Joel Billings]

We've only got one save where we can replicate the situation, and in that save a certain piece of data is set in a way that looks incorrect. But it appears to be set incorrectly before the unit is surrounded, and we don't have a save early enough to see why it is being set that way. Until we get a series of saves that allows us to duplicate the problem, or at least lets us see when the data is being set incorrectly, we are just guessing as to what may be going wrong. That's why I've asked for the earlier save that goes with the other saves attached in this thread. At this point all we can say is that yes, something looks wrong, but we don't know what's causing it or what changed to make it start happening. If anyone has a set of saves that shows the issue and let's us backtrack through the prior turn, please attach them. Thanks.

[Skritshell]

If you can access the server saves the game between Dean and Roadwarrior should should give you all the saves you need. Although I have no idea if it just overwrites the current save or if it creates a repository. Same thing is happening in my game vs road. Cheers

[Joel Billings]

Unfortunately the server only saves the last 2 uploads, and based on what I've seen I need more saves going back from just after it happened that that to pin down when it goes bad.

[gundam1985]

It seems if the unity are weak enough can lead to surrender without isolated. The images show the Axis troops weren't isolated from the current turn.

[Joel Billings]

Gary did find a bug in the code that could cause units to surrender when they weren't isolated. However, we're not entirely sure what caused it or how often it could happen that units would be incorrectly flagged to surrender if forced to retreat. It doesn't happen every time you surround a unit and force it to retreat. The unit has to already be flagged incorrectly that it was far from the rail. He's just put in a fix to prevent the units from getting into that state and we're internally testing it. We hope to release it to testers very soon, and then to the public (late next week if all goes well). The fix won't correct units that are improperly flagged until the next logistics phase, so you will have to run through a turn before any improperly flagged units are fixed and safe from the bug.

[Hardradi]

Is this fix in the new .03 BETA?

[Joel Billings]

No, it's in a version we just released to our testers a few hours ago. 1.01.03 had been in the pipeline before we confirmed the bug and got a fix in.

[Hardradi]

Thanks Joel