Grave mismatch between Combat Summary and the Combat report?

[Grollub]

Running 1.0.6.1108r9

Is this WAD or is there something wrong?

On the left hand side there is the combat result I saw during combat replay on a key battle in China. My troops held with grave Japanese losses. Imagine my surprise when I open up the game turn and see (right hand side) that I apparentely lost the same combat with heavy losses and also trapping a great number of units below that hex.

Is this WAD or could there be some error? Sync or other?

Save game file in next post.

[Grollub]

Save game file - Combat report.

[Grollub]

Save game file - start of turn from jap player

[Alfred]

Grollub,

Just curious. Did you watch the combat animation? If you did, what was the outcome ie did you see the Chinese defeated and forced to retreat one by one?

Alfred

[Grollub]

I saw the entire Combat replay and they did NOT retreat in that one. I actually made a little victory dance when they held.

Also, there's another strange little thing. To my belief, they shouldn't have defeated my stack with a 2-1 either as all units had at least fort level 1? I could be wrong here though.

[Chickenboy]

Grollub,

I'm thinking a sync bug of sorts. Sorry, mate. Does the problem disappear if you back up a couple turns? Rebooting between turns, particularly between turns from different players, has been helpful for me in avoiding these.

Best,

Chickenboy

[Alfred]

That raises a couple of interesting points.

1. If the enemy truly achieved a 2:1 adjusted ratio that should drop the forts level back to zero which would engineer the retreat.

2. Unlike a base, where the base fortification level applies to every unit, irrespective of how long it has been present, out in the field every unit acquires its own fortification level. It is therefore quite common to find in the same field hex, some units with zero, others with 1, others with 2 and so fortification levels. When you say they had at least a level 1 fortification level is that comment made on knowing exactly each individual unit's fortification?

What I am particularly intrigued in is the possibility that some of your units individually were forced to retreat but others were not. Normally it is one or the other with the benefit usually going to remaining on the battlefield if any unit survived the adjusted odds.

The other thing of course is the different casualties inflicted between the two reports. Which suggests to me that the answer to your query lies somewhere deep in the firepower algorithm. Certainly something for michaelm to look into, if he has the time of course.[:)]

Alfred

[Grollub]

Hmmm ... We'll of course have to double-check (again) that we're using the same build. We've had no experiences with sync bugs (if that's what it is) before. Normally, do you only have to back up one turn or do you have to go further back?

ETA; I only have this game running, so there is no risk of any multiple opponent sync bug.

[Alfred]

Yes chickenboy's answer is simpler in that it could be a sync bug and therefore not of great interest to michaelm.

Alfred

[Grollub]

ORIGINAL: Alfred

That raises a couple of interesting points.

1. If the enemy truly achieved a 2:1 adjusted ratio that should drop the forts level back to zero which would engineer the retreat.

2. Unlike a base, where the base fortification level applies to every unit, irrespective of how long it has been present, out in the field every unit acquires its own fortification level. It is therefore quite common to find in the same field hex, some units with zero, others with 1, others with 2 and so fortification levels. When you say they had at least a level 1 fortification level is that comment made on knowing exactly each individual unit's fortification?

What I am particularly intrigued in is the possibility that some of your units individually were forced to retreat but others were not. Normally it is one or the other with the benefit usually going to remaining on the battlefield if any unit survived the adjusted odds.

The other thing of course is the different casualties inflicted between the two reports. Which suggests to me that the answer to your query lies somewhere deep in the firepower algorithm. Certainly something for michaelm to look into, if he has the time of course.[:)]

Alfred

Just rechecked the fort level at the end of the last turn. I misremembered. All of them had fort level one as can be seen in the attached picture.

[RogerJNeilson]

Sounds very reminiscent of what I experienced a while back. Would watch combat replays and see units holding, then open up turn and see they'd been beaten and were retreated. Maddening.

Sync bug for us, IIRC solved by both applying a new patch in exactly the right method.

Roger

[michaelm75au]

ORIGINAL: Grollub

Save game file - Combat report.

Do you have the save you sent in that created this replay?
Would like to play the normal turn and compare to the replay file.

[Grollub]

ORIGINAL: michaelm

ORIGINAL: Grollub

Save game file - Combat report.

Do you have the save you sent in that created this replay?
Would like to play the normal turn and compare to the replay file.

Attached.

[michaelm75au]

I have several sync bug saves from 1108r9 and were fixed in this series of betas. I can't see any specific mention of sync bugs, but fixing other bugs would have squashed the cause of some sync issues.
eg running off the end of a table due to a wrong variable could change data in memory affecting the turn. The same data might not exist in same place next time the turn is replayed.
I'll play out your set and see how it goes, but will be later in week before I have time for an in-depth check of the sync.

[Grollub]

ORIGINAL: michaelm

I have several sync bug saves from 1108r9 and were fixed in this series of betas. I can't see any specific mention of sync bugs, but fixing other bugs would have squashed the cause of some sync issues.
eg running off the end of a table due to a wrong variable could change data in memory affecting the turn. The same data might not exist in same place next time the turn is replayed.
I'll play out your set and see how it goes, but will be later in week before I have time for an in-depth check of the sync.

Sounds fine - thank you for the support! [:)]

[Grollub]

Checked with my opponent - we're using the same patch.

[michaelm75au]

Comparing the normal turn and the replay under 1108r9, there is a difference between the two when one TF moved, which offset the order of the following random rolls.

Running under the current beta produced the same results for both the normal and replay.

There was one or two fixes in the area of ship movement that may have corrected a sync bug issue.

[PT boat skipper]

Thanks for your answer. I am the opponent of Grollub and he asked me to clarify with you if we could continue the game now. Ie, no significant risk of further sync issues?

ORIGINAL: michaelm

Comparing the normal turn and the replay under 1108r9, there is a difference between the two when one TF moved, which offset the order of the following random rolls.

Running under the current beta produced the same results for both the normal and replay.

There was one or two fixes in the area of ship movement that may have corrected a sync bug issue.

[Grollub]

Back in Internetland. Echoing PTBS question below.

ORIGINAL: PT boat skipper


Thanks for your answer. I am the opponent of Grollub and he asked me to clarify with you if we could continue the game now. Ie, no significant risk of further sync issues?

ORIGINAL: michaelm

Comparing the normal turn and the replay under 1108r9, there is a difference between the two when one TF moved, which offset the order of the following random rolls.

Running under the current beta produced the same results for both the normal and replay.

There was one or two fixes in the area of ship movement that may have corrected a sync bug issue.

[michaelm75au]

I can't see why not. This looked like some sort of odd timing thing. Anyway new beta specifically excludes 'hidden' units from display to non-owning player as it does in one place and not another.