So, for example....

Germany has armies in Berlin and Silesia, and a fleet in Baltic. Russia has an army in Prussia.

Germany convoys Berlin to Prussia, with Silesia supporting.

In this case, Russia _can_ retreat to Berlin. This follows the same rule that would allow a German army in Prussia to "switch places" with an army in Berlin through the use of convoys.

So, you believe that "no retreating to attackers province" rule is governed by inability to directly switch places? But these moves are not simultaneous, the retreating army already "knows" that the province is empty. I always thought that "no retreating to attackers province" is more of common sense.

The thing is, when there is a convoy, I believe that the "attacker's province" is considered to be the last sea zone in the convoy, not the originating land space.

Well, that might make sense, of course, but is it somewhere in the rules? Or is it maybe confirmed by some other situation, hopefully more known and clear?

I would vote for directly declaring that in the rules, if it is so :)

There is a section on this issue on the DATC page:

http://web.inter.nl.net/users/L.B.Kruijswijk/#4.A.5

Oh. As any discussion on such a subtle point, it's very complicated...
Anyway, here's what I understood, please correct me if I'm wrong:
1) 4.A.5 first refers to 4.A.3. In case for webDiplomacy, 4.A.3 explanation is not relevant, since we explicitly define the route - via land or via convoy.
2) Next, 4.A.5 refers to 4.A.4 to help define "from". 4.A.4 seemingly comes to determining "from" as originating army province except when convoy itself is being disrupted.
3) However, DATC commentator in 4.A.5 decides the opposite, saying that he prefers to allow that retreat. He claims that it is logical, since the border in question was not used by attacker. The thing is that to me that is NOT logical regardless of the border :/

On webdip a move ordered "via convoy" actually will take the land route if the convoy is unavailable or dislodged and the land route is possible.

Unfortunately it is not clear which interpretation of 'from' kestas took when coding this.

See: http://webdiplomacy.net/datc.php

There is no mention of this issue. 'if a fleet is being dislodged by a convoyed army then it can't retreat to the sea territory 'from' where the attack crossed in any interpretation - because there is still a fleet (the convoying fleet) there.

Thus you could code this either way to get the result, i suggest someone run a 1vs1 game on vdip, setup a situation and see what happens. Shouldn't take more than a few minutes (and it should be the same result as you will get here)

I know what happens on vdip, in Octopus setting though. That's why this thread appeared in the first place :)

So, the question was if the webdip implements rules correctly. And the problem is now what is "correctly" :/

'Correctly' is down to interpretation, is in a classic variant on vdip whatever happens there likely happens here.

Whether you interpret that as 'correct' is a seperate debate...

I'm with the DATC author's comment on letting the retreat occur. Partly because i can't ever see a reason to convoy - unless you were planning on having their army swap places...

but the retreat is geographically possible, as you don't get an army coming head on from the land province ...
and I suppose rules should be made to stick to reality as much as possible? so since it's geographically possible, it should be allowed.

It would seem strange to me if a unit could move there (in my example, the rules make it clear that Russia could move Prussia-Berlin if Berlin is convoyed), but not retreat there.

When your army is dislodged, you also don't get an army coming head on. It had already came and fought with you, that's why your army is dislodged. So why it is not geographically possible?
Speaking of reality, it is also geographically possible to swap armies and fleets, or do you think every border in reality has only one crossing point? Reality is not always a good advisor when talking about such abstract models as Diplomacy war model.

However I have to agree with orathaic on having no reason for convoy except swapping - that is, under normal rules.

Oh, that way it is a quite sensible argument, Tasnica. Thank you.
However, since it is an exception to a basic rule of no retreating to an attackers province, it should be probably clarified as such somewhere.

I'm more going with a practical 'because it is unclear it should be clarified'

Given that this is a rare situation, doesn't significantly affect game play and involves creating a special exception in the programming logic, I would guess that webdib doesn't allow retreating to the attacking territory. Just a guess though.

Could someone who is nearing a safe endgame situation try this out with a dependable ally? I'd be really interested to know. My guess is that it should work, logically.

Erm, it seems I wasn't clear enough: this *happened* in a vdip game, in Octopus setting (double movement), but there shouldn't be any difference in rule resolving here.

There IS difference that in Octopus this situation is actually normal, I mean, there may be a need for a real convoy where army can't make that move by land at all (across the sea). Fleet that was dislodged by that attack with an army across the sea *was able* to retreat back to the province where from the attack originated.

I think you can retreat there. Since units can exchange territories if one of them is convoyed, it sounds obvious you shuld be able to retreat there.

Generally you can retreat anywhere, where you could have moved successfully as well. And since you could have moved there, you should be able to retreat there.

With a convoy it can be done just as two units can swap places by one of them convoying. The key is that they are taking different paths and the rule states that if it could have moved there unsupported during the movement phase, it can retreat there. Now, the same rule should apply on VDip in Octopus (and Pirates) but it would mean they were able to move there via different paths which is quite feasible in multiple territory moves.

Oh, and that is one of the reasons you specify by land or by convoy. In the early rules, you didn't specify it and people "hijacked" armies so the dislodged unit would have a retreat.

I ran a quick test on my webdip instance.
http://forum.webdiplomacy.net/download/file.php?id=650

If that doesn't work for you...
http://forum.webdiplomacy.net/viewtopic.php?f=12&t=895

It makes sense to me. If you could have moved there unaided, you should be able to retreat there. It appears that webdip uses the "from" to determine that the attacker is coming from the water territory not its original territory.

And, if you didn't want the retreating unit to move/retreat there, then don't do the convoy.

Guys, I don't think it's that complicated. You can retreat to anywhere you could have moved in the previous phase. (That is why you can't retreat to somewhere where there was a standoff.) Since it is legal to swap units via convoy (see page 16 here: http://www.stabbeurfou.org/docs/regles/en/diplomacy_rulebook.pdf), it is legal to retreat to where a convoy came from, even if the convoyed army ends up where you moved from.

The thing is "You can retreat to anywhere you could have moved in the previous phase" goes in conflict here with "You can't retreat to the province of the attacker".

But I believe it is already justified and should stay as it is.

No, It doesn't because the attacker entered your territory from a sea zone. He may have started the phase where you are retreating to but he attacked from the sea zone of the last fleet in the convoy.

First: by the link Mr A you may see that "from" is treated as the starting point, except for case of convoyed army trying to disrupt attack on convoy.
Second: the retreating rule as I know it says nothing about entering from wherever, or else. Retreating rule generally says that you can retreat to wherever you might have moved, with one of the exceptions being "the province from which the attacker came", which is obviously not the last fleet in convoy.

localghost said - "Retreating rule generally says that you can retreat to wherever you might have moved, with one of the exceptions being "the province from which the attacker came", "

That is incorrect. "The province from which the attacker came" is not an exception to "you can retreat to wherever you might have moved" because you couldn't have moved to that province.

"The province from which the attacker came" is one of the exceptions to "you may retreat to an open spot" with the other exception being "a province that had a bounce".

The bottom line is that the official rulebook has a conflict of rules and the DATC allows for it to be interpreted either way. The way webdip does it is one of the acceptable ways.

According to the DATC, DPTG says the retreat is allowed. Excerpt follows:

4.A.5. RETREAT WHEN DISLODGED BY CONVOY
In a very rare situation (see test case 6.H.11 and 6.H.12) a unit can be dislodged by a convoy from an adjacent place. Then the question arises whether the dislodged unit may retreat to the starting place of the convoying army.

The following two interpretations are possible:

The unit may not retreat to the starting place of the attacker.
The unit may retreat to the starting place of the attacker when the attack was with convoy.
This is according to the DPTG.
First note that if issue 4.A.3 is played according to choice c, then this issue is not relevant anymore. If it is played according to choice b, then this is only an issue in the extremely theoretical situation that both units attempted to move by convoy (see test case 6.H.12).

All the rulebooks are ambiguous on this issue. They say first "A dislodged unit must retreat to an adjacent province that it could ordinarily move to if unopposed by other units.". According to this sentence the unit may retreat to the starting place of the attacker (choice b). However, in further explanation all the rulebooks say: "The unit can not retreat to the province from which the attacker came.". For the interpretation of the word 'from' see also issue 4.A.4. This sentence must be interpreted that the particular retreat is not allowed.

Alderian, yes, I wasn't precise enough.

Draugnar, last paragraph you cited describes exactly the conflict I'm talking about.

And anyway, I don't see any problem with this anymore, however I believe it would be reasonable to mention this situation in retreat rules.

Well, we don't have the full rules on the site anyhow (violation of copyright to do so) but the DPTG that WorldDipCon uses for adjudication allows the retreat. So when the primary governing tournament body uses it, you can pretty much accept it as the standard method of adjudicating the turn.

they was I see it in order to retreat somewhere both conditions must be satisfied;
1. an adjacent province that it could ordinarily move to if unopposed by other units
2. The unit can not retreat to the province from which the attacker came.
in this rare situation condition 2 is not satisfied therefore I think retreating there should not be possible.

@hellalt, read the DATC entry I quoted above and it's reference to the Diplomacy Players Technical Guide (DPTG). Of course, had they never even bothered to write that "exception" it wouldn't even be an issue. After all, you coudln't have moved there during the regular phase anyhow if it wasn't a convoy, so why bother to include that? they did it to clarify but actually caused more confusion with it.

localghost: You didn't understand my point, I think. The rules give explicit exception to rule that units can't swap places when a convoy is involved.

hellalt, those two conditions don't go together, as I mentioned just a bit upthread.

You can explain the retreat rules two ways.

The first and simple way: You can retreat anywhere that you could have moved to unaided.

The more complicated way: You can retreat to any open adjacent territory that did not have a bounce and that the attacker did not come from.

Applying an exception condition of one explanation method to an alternative explanation method is not valid.

Does that make any sense? I'm not sure I'm wording it just right.

hellalt: But isn't there an inconsistency if you interpret things your way?

In the regular moves phase, there is also the rule that a unit can't swap places with another unit moving in the opposite direction. That is analogous to your condition 2..

An exception is made to the original rules when convoys come into play. Therefore I see that it is only logical to extend that exception to the analogous situation when units are retreating.