The Elephant Vanishes : obvert (J) vs Historiker_SqzMyLemon_Canoerebel (A)

[Lowpe]

ORIGINAL: CaptBeefheart

Too late now, but one way around it is to break up those long strat journeys into shorter segments.

You might accept his offer and change the strat movement destinations of your affected units to bases close to where they are now. That way you'll lose a little bit of time getting them into action but not 4+ days.

Unfortunately that doesn't always work.

I don't think most players understand the total devastation this bug causes when it hits, especially here. Japan generally has the advantage of interior lines in defense and the ability of a rapid response. This in effect has been taken away, many units sidelined (delayed) from 2 days (fastest to weeks) and most likely means that Japan has to evacuate what he can to Korea...which might be Japan's plan anyhow, but now numerous units won't be able to make it which should have.

Now, I don't really know what happened in the turn. I don't know if there was CD guns & mines on the river that sat idly by while the invasion fleets steamed upriver, like I have said for a long time this is a cheesy move (against a river properly defended with mines and cd guns), but a move I doubt CR knew was cheesy as he is a classy guy. My guess here is that Obvert didn't have any mines and cd guns on the river which kind of makes my point of the upriver invasion being cheesy moot....

It is not a well known bug, or well known game design. When I first reported on it, nobody offered any explanations, causes, solutions and no other experiences cited. After continued, better documented examples, it is a bug but, folks there is nobody left to squash bugs anymore. Anyhow, I only shared with Obvert that there was a problem, keeping almost all details to myself because, quite frankly at that point in the game I thought I could be competitive. After the first time, I don't think I ever mentioned it happening again to Obvert. I had hoped that it was unique to our strange game.

Without knowing the details of the problem, Obvert immediately offered a turn redo. Upon reflection, I decided to continue the game as is for a variety of reasons and I continued to document the problem as it occurred.

I should mention that there is an additional feature, that Obvert hasn't run into yet since he most likely hasn't issued any orders...and that is the inability to set a destination with SR travel to some bases that have a clear path.

I don't know what the solution is here,other than that the two players need to figure it out to their satisfaction. That CR thinks this might be a tough lesson or that Obvert was using an elaborate tactic...simply informs me he doesn't fully understand what has happened.

Communication is a tricky thing, and given that there are readers here that don't really understand what has happened, and that Obvert probably gave CR even less detail of the problem, it is no surprise that misunderstandings occur.

[Alfred]

This is not really a bug as it flows from a central game design concept.
 
It is not an easy task to try to code a different implementation for this central design concept for several reasons.  Firstly, it is difficult to tinker with the strategic movement code without crippling the AI's use of strategic movement.  Secondly, and this also links up with the first difficulty, any code changes in this area have to be thoroughly tested to ensure that the dormant computer control zone is not reactivated to interfere with other human control mechanics.
 
The China Theatre is the most likely area to witness this issue due to the paucity of railways, in particular the lack of multiple lines servicing most bases, and the relatively (compared to other theatres) widely spread in built trails in the countryside.
 
Alfred

[obvert]

ORIGINAL: Alfred

This is not really a bug as it flows from a central game design concept.

It is not an easy task to try to code a different implementation for this central design concept for several reasons.  Firstly, it is difficult to tinker with the strategic movement code without crippling the AI's use of strategic movement.  Secondly, and this also links up with the first difficulty, any code changes in this area have to be thoroughly tested to ensure that the dormant computer control zone is not reactivated to interfere with other human control mechanics.

The China Theatre is the most likely area to witness this issue due to the paucity of railways, in particular the lack of multiple lines servicing most bases, and the relatively (compared to other theatres) widely spread in built trails in the countryside.

Alfred

The action of the landing a unit on the rails caused multiple units (later in the turn) to move one hex and then come out of strat mode with no delays.

While I understand the principal of having a unit stop motion when it's destination is no longer reachable this "broke the rules" (and any form of plausibility) by disembarking from strat mode without delay after actually moving.

It doesn't matter what it's called or how hard it is to code, it's not working as it should according to set rules of the game, which is the only reason I would bring it up at all. It's a disadvantage to me to have to talk to my opponent about this, and to replay the turn. I wish I didn't have to do it.

[witpqs]

ORIGINAL: obvert

ORIGINAL: Alfred

This is not really a bug as it flows from a central game design concept.

It is not an easy task to try to code a different implementation for this central design concept for several reasons.  Firstly, it is difficult to tinker with the strategic movement code without crippling the AI's use of strategic movement.  Secondly, and this also links up with the first difficulty, any code changes in this area have to be thoroughly tested to ensure that the dormant computer control zone is not reactivated to interfere with other human control mechanics.

The China Theatre is the most likely area to witness this issue due to the paucity of railways, in particular the lack of multiple lines servicing most bases, and the relatively (compared to other theatres) widely spread in built trails in the countryside.

Alfred

The action of the landing a unit on the rails caused multiple units (later in the turn) to move one hex and then come out of strat mode with no delays.

While I understand the principal of having a unit stop motion when it's destination is no longer reachable this "broke the rules" (and any form of plausibility) by disembarking from strat mode without delay after actually moving.

It doesn't matter what it's called or how hard it is to code, it's not working as it should according to set rules of the game, which is the only reason I would bring it up at all. It's a disadvantage to me to have to talk to my opponent about this, and to replay the turn. I wish I didn't have to do it.

I guess Alfred is pointing out that it wasn't practical for them to improve on it. If that makes more sense?

[Lokasenna]

ORIGINAL: Alfred

This is not really a bug as it flows from a central game design concept.

This is basically what I was trying to say.

[Lowpe]

So it is a quirk, but I think parts of the quirk are bugged, too. And I also think there is more than one quirk.[:D]

[Lokasenna]

ORIGINAL: Lowpe

So it is a quirk, but I think parts of the quirk are bugged, too. And I also think there is more than one quirk.[:D]

Since I just posted this thought in the other AAR, I'll post it here too.

The only issue here, to me, is that they dropped to Move immediately and did so after already moving 1 hex via strategic. That's very minor in the realm of "bugs". Were I in one of their positions, I would've hashed out whether re-doing the turn because of the movement mode switch or the 1-hex movement was the best recourse. The movement being cancelled itself is not a bug.

[Lowpe]

ORIGINAL: Lokasenna

The movement being cancelled itself is not a bug.

I have said previously that movement being cancelled is okay by me...it is all the other aspects of this quirk/bug.

Would you consider it a bug if it effected units travelling by SR that don't use the portion of the rail road that was cut? Would you consider it a bug if it happened to units that didn't move...i.e. were still packing for movement? Would you consider it a bug if it happened to units that moved more than 1 hex? Would you consider it a bug if units still in SR mode could not be reassigned to a new destination without unpacking and repacking first? All these things have happened to me, and I suspect will probably happen to Obvert.

Anyhow, call it what you want, but I wouldn't call it minor when you have units stranded in move mode a week or more from the closest base staring down a major invasion.

I doubt Obvert knows the full extent of the problem until he actually looks everywhere. I certainly didn't at first.

Finally, I have my doubts that the redo will solve everything.

[Alfred]

ORIGINAL: witpqs


I guess Alfred is pointing out that it wasn't practical for them to improve on it. If that makes more sense?

Exactly.

At one point the devs did a relatively small change to the strategic movement code to fix an actual bug. In doing so they completely broke the AI.

The AE team project leader was fond of classical WITP's Mr Fragg's saying that "ahistorical play leads to ahistorical outcomes". There was very little willingness to change fundamental game design concepts which worked just to accommodate ahistorical play. Squash a bug which came to light from ahistorical play, yes that would be addressed by the AE dev team. However if the ahistorical play resulted in a less than ideal outcome albeit one still consistent with the game design and when simple player workarounds exist which take the sting out of the ahistorical play, that did not rate high on the priority list for code revisions.

I'm not saying that the full extent of the outcomes from how they implemented the game design concept would be viewed by the devs as being ideal/perfect outcomes. Nor the ramifications on gameplay of some of these outcomes was foremost in their thoughts. But to say this is not consistent with the game design is simply incorrect. Abstractions always include the good with the ugly.

Alfred

[PaxMondo]

ORIGINAL: Alfred
Abstractions always include the good with the ugly.

Alfred

[&o][&o][&o]

[obvert]

ORIGINAL: Alfred

ORIGINAL: witpqs


I guess Alfred is pointing out that it wasn't practical for them to improve on it. If that makes more sense?

Exactly.

At one point the devs did a relatively small change to the strategic movement code to fix an actual bug. In doing so they completely broke the AI.

The AE team project leader was fond of classical WITP's Mr Fragg's saying that "ahistorical play leads to ahistorical outcomes". There was very little willingness to change fundamental game design concepts which worked just to accommodate ahistorical play. Squash a bug which came to light from ahistorical play, yes that would be addressed by the AE dev team. However if the ahistorical play resulted in a less than ideal outcome albeit one still consistent with the game design and when simple player workarounds exist which take the sting out of the ahistorical play, that did not rate high on the priority list for code revisions.

I'm not saying that the full extent of the outcomes from how they implemented the game design concept would be viewed by the devs as being ideal/perfect outcomes. Nor the ramifications on gameplay of some of these outcomes was foremost in their thoughts. But to say this is not consistent with the game design is simply incorrect. Abstractions always include the good with the ugly.

Alfred

This bolded part is particularly important to the way that AE is played now.

Almost all of what we do as players is ahistorical in it's extremes. We use what is available and anything that will work to come up with ways to outsmart our opponents. It's inevitable that this would come about since the game itself is abstracted from history, and in it's own way inherently impossible to play with historical parameters consistently. Sure, you can try to hold yourself back somewhat, but the game itself will make very non-historical decisions in how units move, interact, fight and retreat. So the trend has been to simply go with the game as designed.

That worked well when michaelm ([&o][&o]) was here to assist, to look at things and bring a coding perspective to how the function of the game was working with these newly discovered parameters from hundreds of thousands of example turns.

You seem to be saying that developers saw this problem and decided it wasn't worth fixing. Maybe that is true, but listening to Lowpe's experience of it and seeing my own limited example, I can only imagine they found it actually impossible to fix this. Putting a reasonable number of units on trains and sending them a few hundred miles to reinforce a defensive stand seems entirely historical, plausible, and within game parameters as designed. Cutting rail to thwart this is also part of game design and historically plausible. Hence the need to redo a turn when both of these design features lead to a situation where rules of the game as designed do not function as designed.

[obvert]

Damn. Went on a nice country walk today. Sunny. Gorgeous. Two village pubs and a lot of castles, rivers and new lambs bouncing through the fields.

I got home thinking there would be a turn waiting from Dan, but my email, hastily sent this morning, hadn't gone through completely. [:@]

Sent a profuse apology and sent just now.

The new turn without the APD landing on the river/rail hex went smoothly with only a few perceptible changes to other areas of the game. Nothing I think is major anyway.

So I've reset all moving units in strat mode, and lets hope Lopwe is not all-seeing in his prediction of more doom for my railed units. [;)]

[obvert]

One issue I've though about around this is the possibility it could be intentionally exploited with paradrops and minor amphibious landings to thwart a retreating or invading force, even when not controlling rail hexes. Imagine a force that landed at an enemy base at a rail terminus. Take for instance Lang Son.

What if, knowing I was retreating the Bumra Army, CR decided to drop a Chindit unit on LangSon from Burma. Entirely possible, and that could send all units leaving Saigon into a spiral of spooky behaviour at a distance. h could do the same at Udon Thani. Suddenly I'd have 10-15 divisions and who knows how many other units stranded outside of a base hex, not moving in strat mode, and a 10k AV Allied army in hot pursuit.

In the early game it could go against the Allies as they retreat on Malaya, in Burma, or in China. Even India if that is in play, although many parts of India have many multiple rail lines to the same place.

This really could be a serious problem in many situations.

[brian800000]

ORIGINAL: obvert

One issue I've though about around this is the possibility it could be intentionally exploited with paradrops and minor amphibious landings to thwart a retreating or invading force, even when not controlling rail hexes. Imagine a force that landed at an enemy base at a rail terminus. Take for instance Lang Son.

What if, knowing I was retreating the Bumra Army, CR decided to drop a Chindit unit on LangSon from Burma. Entirely possible, and that could send all units leaving Saigon into a spiral of spooky behaviour at a distance. h could do the same at Udon Thani. Suddenly I'd have 10-15 divisions and who knows how many other units stranded outside of a base hex, not moving in strat mode, and a 10k AV Allied army in hot pursuit.

In the early game it could go against the Allies as they retreat on Malaya, in Burma, or in China. Even India if that is in play, although many parts of India have many multiple rail lines to the same place.

This really could be a serious problem in many situations.

In the game's defense, that may not be unfair...it wouldn't take a division or control of a hex to really mess up rail movement. See the Bridge on the River Kwai for a hypothetical example.

[JohnDillworth]

Putting a reasonable number of units on trains and sending them a few hundred miles to reinforce a defensive stand seems entirely historical

Perhaps...but I expect that Japans rail capacity was stretched to the max and they didn't have much of if sitting around in the yards just waiting to transport military units and all their equipment on a few hours notice. I guess the switched to strategic mode time abstracts gathering all the rail transport to the correct places. Still, there had to be some guerilla activity and in 1945 Japan probably didn't have many trains on speed dial. Military units take up a lot of rail cars. The web has a suprising about of data about this but this count is interesting:

"In the book “Blitzkrieg”, author Len Deighton notes that a full-strength panzer division transported by railway needed no less than 80 trains, each train with up to 55 wagons.


Now Japan traveled a bit lighter than Germany but even if you half those numbers one would need 2,200 rail carriages per Divisision

[obvert]

ORIGINAL: brian800000

ORIGINAL: obvert

One issue I've though about around this is the possibility it could be intentionally exploited with paradrops and minor amphibious landings to thwart a retreating or invading force, even when not controlling rail hexes. Imagine a force that landed at an enemy base at a rail terminus. Take for instance Lang Son.

What if, knowing I was retreating the Bumra Army, CR decided to drop a Chindit unit on LangSon from Burma. Entirely possible, and that could send all units leaving Saigon into a spiral of spooky behaviour at a distance. h could do the same at Udon Thani. Suddenly I'd have 10-15 divisions and who knows how many other units stranded outside of a base hex, not moving in strat mode, and a 10k AV Allied army in hot pursuit.

In the early game it could go against the Allies as they retreat on Malaya, in Burma, or in China. Even India if that is in play, although many parts of India have many multiple rail lines to the same place.

This really could be a serious problem in many situations.

In the game's defense, that may not be unfair...it wouldn't take a division or control of a hex to really mess up rail movement. See the Bridge on the River Kwai for a hypothetical example.

I think you're not getting the area being considered here. We're not saying the units should not be disrupted by having their movement stopped or not continuing to their destination. It's entirely historical to interdict railways and the game doesn't model what FB and ground attack F could do in late war to static rails or rolling stock.

I'm hypothesising about dropping a unit to intentionally bring this bug into play. Putting a unit at the end of a rail line on the Chinese border to send a number of units in Saigon (many hundreds of miles away) off on a 40 mile journey from which they're dumped at the side of the tracks and abandoned in a completely different movement mode.

[obvert]

ORIGINAL: JohnDillworth

Putting a reasonable number of units on trains and sending them a few hundred miles to reinforce a defensive stand seems entirely historical

Perhaps...but I expect that Japans rail capacity was stretched to the max and they didn't have much of if sitting around in the yards just waiting to transport military units and all their equipment on a few hours notice. I guess the switched to strategic mode time abstracts gathering all the rail transport to the correct places. Still, there had to be some guerilla activity and in 1945 Japan probably didn't have many trains on speed dial. Military units take up a lot of rail cars. The web has a suprising about of data about this but this count is interesting:

"In the book “Blitzkrieg”, author Len Deighton notes that a full-strength panzer division transported by railway needed no less than 80 trains, each train with up to 55 wagons.


Now Japan traveled a bit lighter than Germany but even if you half those numbers one would need 2,200 rail carriages per Divisision

Ok. Now you're bringing game design into question, not the plausibility of moving these units in a manner that the design allows for in the rules.

Japan in this game is not in the state Japan was in during the war, nor should it be considered by those conditions. China has been completely under it's Co-Prosperity sphere since mid-42. All bases are garrisoned adequately to subvert ALL guerrilla activity as conditions of the game allow.

Conjectures about Japan's wartime capacity to move units by rail isn't relevant to the discussion except that they did indeed have some capacity to do this and that is modelled in game equally for the Allies and for Japan. The Allies can also rail any number of units across the same territory without having to ship in or manufacture rail cars or engines because those are abstracted by the game.

These things should be especially irrelevant in a game without stacking limits where he Allied stack in Thailand is now about 500k troops, all in one hex about to crash over a river into Bangkok!! [:D]

Do you think the bridge will hold up if 10,000+ tanks, trucks and arty pieces try to cross it within 24 hours? Might there be a few traffic jams? [:D]

[Lowpe]

The game plays as the game plays. Trying to come up with justifications to explain away the quirks, is all well and good but an exercise in role playing and creativity rather than playing the war game.

However, it is fascinating to look at pictures of the Manchuko Railways Headquarters during the war and their substantial amount of sophistication. Ditto for Chinese railways. Also, you get insights into the different gauges present, the lines, how Japan even stripped lines for steel, the effects of bombing...railway service was established 2-3 days after the abomb for example.

Kawasaki built close to 800 of these rolling stock locomotives for Manchuko and they also had over 100 passenger locomotives of very modern design, and from the picture they look pretty nice. A couple years back I posted a picture from Nashville, where an equivalent American locomotive is on display from WW2. I encourage everyone to do some basic searches and reading here, it is really quite fascinating.[:)]

It would be super nice to see a more sophisticated railway system in the game. Interdiction for example, but the game will most likely not be beta patched ever again, let alone adding more content.

I think in my WITPAE experience I have asked once for a turn do-over and it was promptly and graciously granted. I have been offered turn do-overs on perhaps 3-5 more occasions and declined them rather living with the results. I have been asked for do-overs probably 3-4 times by my opponents and always allowed them.

In this, such a public game with both sides having an AAR by highly seasoned and honorable opponents, it is hard to me to fathom a reason why the do-over would cause any concern other than the inconvenience of having to do a turn over.

Lets get back to the war...I am curious to see how Japan can answer this upriver invasion -- which I am not really sure it really is? Could they just be light moves to cut railway lines for the major landing elsewhere?

[brian800000]

ORIGINAL: obvert


I think you're not getting the area being considered here. We're not saying the units should not be disrupted by having their movement stopped or not continuing to their destination. It's entirely historical to interdict railways and the game doesn't model what FB and ground attack F could do in late war to static rails or rolling stock.

I'm hypothesising about dropping a unit to intentionally bring this bug into play. Putting a unit at the end of a rail line on the Chinese border to send a number of units in Saigon (many hundreds of miles away) off on a 40 mile journey from which they're dumped at the side of the tracks and abandoned in a completely different movement mode.

But the counterpoint is that it isn't a bug if used locally, and the opponent has no way of knowing if units are rolling in from 2,000 miles away or will be showing up at the destination the next turn. Tactically the solution is to cut the rail lines where you can.

It doesn't make sense to dump the units out of strat mode 2,000 miles away, but the workaround is there for the player to click a few more times and have more secure destinations.

[witpqs]

ORIGINAL: brian800000

ORIGINAL: obvert


I think you're not getting the area being considered here. We're not saying the units should not be disrupted by having their movement stopped or not continuing to their destination. It's entirely historical to interdict railways and the game doesn't model what FB and ground attack F could do in late war to static rails or rolling stock.

I'm hypothesising about dropping a unit to intentionally bring this bug into play. Putting a unit at the end of a rail line on the Chinese border to send a number of units in Saigon (many hundreds of miles away) off on a 40 mile journey from which they're dumped at the side of the tracks and abandoned in a completely different movement mode.

But the counterpoint is that it isn't a bug if used locally, and the opponent has no way of knowing if units are rolling in from 2,000 miles away or will be showing up at the destination the next turn. Tactically the solution is to cut the rail lines where you can.

It doesn't make sense to dump the units out of strat mode 2,000 miles away, but the workaround is there for the player to click a few more times and have more secure destinations.

Good points.

In my last game I invaded a point where I was concerned units might be railing in, so I landed a regiment sized unit on each side of the invaded bases to cut the rail lines and hold for the time required to capture the base. Never heard any complaint from my opponent so I assume nothing dis-entrained far away (maybe not at all?). No units railed into those hexes either, so I don't even know if any were sent direct or only to nearby cities.