More About Convoys

By Rod Walker

I have always contended that the 1971 Rulebook is a clear and remarkably self-consistent document, and that there is no need to have it revised a second time. Nonetheless, it does appear that the situations involving convoys can be knotty… but not nearly so problematic in many cases as it might seem.

What I would like to do in this article is, first, trace briefly the history of the Rulebook. Second, make a couple of general observations about convoys. Third, return to some specific examples that have been considered recently and suggest how they should be adjudicated.

A Brief History of the Rules

The present (1971) Rulebook is actually the fourth. The first Rulebook was that of 1958, which differed very basically from any of its successors. The second Rulebook, embodying more or less the present rules, was put out in 1959. This was reprinted in a different format (much like the present one) in 1961 when Games Research bought the game. I do not believe any substantive changes were made in this reprint. The third Rulebook was published in 1966. It was almost exactly the same as the second, but included two examples (numbered I and II) regarding cut supports. It also changed the configuration of the board in 5 and 6-person games. However, no other changes were made.

By 1966, however, postal Diplomacy was in full cry in the USA. Already differing and conflicting interpretations of various rules had been uncovered and widely discussed. They were given names in those days: the Boardman Dilemma, Miller’s Rule, the Turner Rule, Koning’s Rule, and so on. Thick compendia of these things were being compiled in the late 60s. Loopholes abounded and GMs and players delighted in finding them; it became almost a game within the game to identify a new rule problem.

In 1970 Games Research commissioned a young man who had never played Diplomacy, Steve Marion, to draw up a draft of a revision. At about the same time, I submitted a suggested draft to John Moot, the President of GRI. John put me in touch with Steve. After a flurry of letter writing, a committee was formed which would work with Allan Calhamer in setting up a final revised Rulebook. The committee consisted of Steve, myself, John McCallum, John Boardman, and a couple of other postal GMs/players who had been prominent in the preceding orgy of rues discussions.

Two or three working drafts were generated. One of them, which I had drawn up, adopted the innovation of numbering the sections and subsections of the rules. This ultimately became the underlying basis of the new Rulebook. There were a lot of discussions over details… the victorycriterion, an expanded sample game, convoys … especially convoys. The committee did not in fact reach any final collective conclusion. Some of the debates had bogged down in fannish acrimony, and at that point Allan completed a final draft himself and sent it to GRI for printing.

We Pause Now…

for a word about the Coastal Crawl. I’ve heard some rumblings of confusion on that point. In the 1961/66 Rulebook, the term “space” was not as clearly defined as it is now. The ambiguity led John McCallum to conclude that the two coasts of a double-coasted province could be regarded as two spaces. This being so, the following orders would be legal and would succeed:

F MAO-Spa(sc). F Spa(nc)-MAO

This effects an exchange of units, which is otherwise not legal under the Rules. In the 1971 Rulebook, the term”space” is more clearly defined, and it is now not legal to exchange two fleets using both coasts of a double-coasted province (i.e., between Spa & Por, Spa & MAO, Bul & Con). It was Allan Calhamer’s specific intent to preclude such an exchange.

Convoys in General

Convoys have always presented a difficult adjudication problem. Two difficulties were well known when the new Rulebook was prepared. One was then called the “Shagrin Alternate Convoy”. It applied to two locations (such as Lon & Bel) between which an army could be convoyed by either of two different fleets (such as ENG and NTH). Dick Shagrin argued that if he had A Lon, F ENG, F NTH, and ordered A Eng-Bel convoyed by both fleets, if one of the fleets were dislodged, the other would still provide a valid convoy route and the order could still succeed. This caused a furore, with most GMs opposing the interpretation and suggesting that if either fleet were dislodged the convoy would be disrupted (although that term was not then in use). Rule XII.4 represents the usual GM response to the Shagrin Convoy in 1970, but has itself been said to raise problems.

Allan Calhamer will not agree with me here, but an important consideration in convoyed attacks is the direction from which the convoyed attack is coming. The important statement of this concept is Brannan’s Rule, named after Steve Cartier (who was and is also known as Dan Brannan). The Rule states, “The army in a convoyed attack is deemed to come from the space occupied by the last convoying fleet.” This originally had to do with whether the convoyed attack could cut the support of a unit in the space if was attacking, if that unit was supporting an attack on the last convoying fleet in the convoy chain. Thus:

FRANCE: A Spa-Nap C by F WMS & F TYS.

ITALY: F Rom-TYS S by F Nap.

Question: does the attack of A Spa cut the support of F Nap for the attack on TYS? Brannan’s Rule says “No” (and in the mid-60s, many GMs were saying “Yes”). The Rulebook (Example 13) also now says “No”, although the language of Brannan’s Rule is deliberately omitted. It was Allan’s intent that Brannan’s Rule should not be used in Diplomacy adjudications, but no language forbidding it was ever inserted. In fact, my view is that the effect of Example 13 is to support the use of Brannan’s Rule in the absence of conflicting language. Carried to logical limits, the Rule yields some results regarded by many as peculiar. But it is still a valid method of dealing with problems that arise with respect to convoyed attacks.

Other Recent Problem Cases

I am going to number these so that they more or less correspond to the order in which they have been raised. My adjudications will be indicated with each example (underlined orders fail).

1A. The Voice of Doom Poll Situation

ENGLAND: A Bel-HoI*; F Nth C GERMAN A Den-Bel.

FRANCE: A Hol Std.

GERMANY: A Den-Bel S by A Ruh

*England’s A Bel is dislodged. The Rulebook does not say (as if often alleged) that a country “cannot participate in its own dislodgement”. The Rules are actually sparkling clear on this point. Rule IX.3 is the only Rule (with Examples 1 and 2) that applies. It states that a player can’t dislodge one of his own units by attacking it, and if he supports a foreign unit in an attack on a space occupied by one of his own units, the support (but not the attack per se is invalid if that unit does not move out. (Example 2 shows that if the foreign unit has enough support of its own to succeed, it will, regardless.) Nowhere does the Rulebook state that a player can’t convoy in an attack that dislodges one of his own units. Nor am I aware of any reason why it should.

1B. Same, With Nastier Extras


FRANCE: A Hol Std.

GERMANY: A Den-Bel; A Ruh-Hol

Nothing moves. Any player who makes the orders England does here is just colossally stupid. Well, if you screw up, you should pay the consequences …and either England did in writing those orders, or Germany did in choosing England as an ally.

2. Self-Defeating Convoy




This is a simplified version of the Pandin Paradox, which had a great vogue of interest in 1971/72, and has remained a popular favourite ever since with some people…who also presumably indulge in sessions of looking at pinheads and counting angels. This situation makes clear just how unlikely it is that any GM will ever be called upon to adjudicate any such set of orders. My ruling, however, is that it is a paradox, that’s the breaks, and nothing goes. Unless the players actually prearrange the thing, you’re not likely ever to see anything of this sort.

3A. Naples Gets Zapped

FRANCE: A Spa-Nap C by F GoL & F TYS and S by A Apu; F Rom S F TYS


F Nap is dislodged and disbanded for lack of retreat. It has been asked whether F Nap’s support is cut, implying a possible contradiction between Rule X and XII.5. Actually, there is none; Rule XII does not need to repeat what Rule X already says. The support of a dislodged unit is always cut.

3B. Behold Brannan’s Rule

FRANCE: A Spa-Nap C by F GoL & F TYS and S by A Apu


This is 3A without F Rom and with the Italian orders reversed. Here the attacks come out of and go into TYS. Because under Brannan’s Rule the attack of A Spa is coming from the direction of TYS, the two attacks are stand-offs. This is a logical extension of Brannan’s Rule and of the ruling made in Example 13. However, Allan Calhamer specifically disagrees with this ruling and would allow F Nap-TYS to succeed, disrupting the convoy.

3C. Beleaguered Convoy

ENGLAND: A Lon-Bel, F Wal-ENG S by F IRI

FRANCE: F Bre-ENG S by F Bel


Nothing goes. This situation is easily resolved under Example 13, which specifically states that the support for F Bre-ENG is not cut by the convoyed attack. Brannan’s Rule applies here and gives the reason; A Lon is coming from the direction of ENG.

4. Pandin’s Paradox

FRANCE: A Pic-Lon C by F ENG; F NWG-NTH S by F Nwy; F Bel S F ENG

GERMANY: A Yor-Bel C by F NTH; F Wal-ENG S by F IRI; F Lon S F NTH

This is Pandin’s Paradox in full cry, although Tony Pandin’s original had four Great Powers going at it. Yes, it’s a paradox. I see no reason to worry about something like this, which should occur in actual play with about the same frequency as the Mediterranean fruit fly appears in Siberia. Therefore nothing moves. But it’s a neat paradox, isn’t it?

5A. More Pandin

ENGLAND: F Wal-ENG S by F Bel, F Edi-NTH S by F Lon.

FRANCE: A Bre-Lon C by F ENG

RUSSIA: A Nwy-Bel C by F NTH

Another neat paradox. As in the previous situation, I rule nothing goes because the entire cycle is self-defeating. This one is particularly neat because I see no way of resolving it with any of the methods usually used to resolve Pandin situations. This in turn highlights my point that it is futile to go to the trouble to make new rules and whatnot for paradoxes in the first place.

5B. Pandin Bear

ENGLAND: F Yor-NTH S by F Edi, F Pic-ENG S by F Lon, F Lpl-IRI S by F Wal, F BAR-NWG S by F Cly.

GERMANY: A Nwy-Edi C by F NWG, A Den-Lon C by F NTH, A Bre-Wal C by F ENG, A Gas-Cly C by F MAO, F IRI & F NAO.

This example isn’t adjudicated, but nothing would go. I hold it up as a perfect example of the tenuousness of all such hypothetical situations. As with most of them … but much more obviously in this case…there are many more units than there are supply centres to account for them. England and Germany are using 18 units …and what, prithee, is going on elsewhere in the game? In other words, in order even to have this sort of situation arise… disregarding the unlikelihood that the right orders will be issued … the respective players have got to neglect their other military fronts just to cram enough units into the area. Now, no doubt some simple paradoxes will occur, on rare occasion, but most of these are pure fantasy.

6A. The Unwanted Convoy

ENGLAND: F ENG C FRENCH A Pic –Bel* (dislodged)

FRANCE: A Pic-Bel S by A Bur, F Bre-ENG S by F MAO

This peculiar special case crops up a lot in discussions of convoy problems. It is nothing more than a red herring. An army does not need a convoy to move to an adjacent province; therefore (Rule VII.1) A Pic-Bel succeeds regardless of what happens to the convoy. Whether the convoy was wanted or not is also not relevant. Ruling otherwise would allow a player to take advantage of a “technicality” which is in fact not even applicable.

6B. The Really Unwanted Convoy

ENGLAND: F Wal-ENG S by F IRI, A Lon-Bel C by F NTH and S by A Hol


Both French units are dislodged. This situation is an admittedly knotty one. It is like the old Shagrin alternate convoy thing in some ways. My position is to treat any foreign convoy as unwanted unless the army could not move without it. However, as a GM, my House Rules do provide that the army’s orders should specify which fleets it expects to move by (and I would then ignore any convoy orders not mentioned in the army’s order). The situation here is not paradoxical, but rather muddied up by the French player. Although this tactic is admittedly clever, it rubs me the wrong way. GMs should probably have some HR regarding convoys which will cover this situation. But it will probably never arise in most games.

6C. Yet More Unwantedness

ENGLAND: F Wal-Eng S by F Iri, A Lon-Bel C by F Nth and S by A Hol.

FRANCE: F Eng C ENGLISH A Lon -Bel. (dislodged)

My comments above in 6C apply here as well.

Proposed Solutions

(1) It has been suggested that Rule XII.4 should be changed so that in situations where alternative convoy routes are possible, instead of the fact that any of the convoy routes has been disrupted being sufficient to stop the convoy, all of them would have to be disrupted. [This change has now been made in the current 4th edition rules © 2000.] However, this would have the effect of allowing the Shagrin Alternate Convoy, which that rule specifically prohibits now. Consider:




Unsure whether France or Turkey will attack him, Italy orders both fleets to convoy. Thus:


ITALY: A Tun-Nap C by F ION & F TYS


The adjudication, per XII.4, prevents A Tun from moving; F ION is of course dislodged anyway. But the proposed alteration of the Rule would allow Italy to hedge his bet by ordering both fleets to convoy. Admittedly a rare circumstance, of course…but aren’t they all, here?

(2) The Verheiden Rule was proposed a decade ago. It’s been adopted into the Houserules of some GMs. The rule is to change Rule XII.5 so that it reads:

“A CONVOYED ATTACK MAY NOT AFFECT THE CONVOYING FLEETS. If a convoyed army’s attack would affect the outcome of an attack on any of its convoying fleets, however directly or indirectly, then the convoyed attack may not take place and the army to be convoyed must hold in its original position.”

Whether it in turn would lead to new adjudication problems, I don’t know; but past experience indicates that it would. In any event, because it goes further than the present Rules in obviating the useful Brannan Rule, I have never been much in favour of it. I must emphasize that the problems it seeks to resolve are primarily hypothetical problems.

The search for a rule, or rules, which would resolve all such difficulties in the game is, I feel, a chimera. Accounting for all contingencies would probably pad the present Rulebook to several more pages… and one of its chief charms (and one of the inventor’s chief intentions) is its comparative brevity (as opposed to most other wargames). This search is also a chimera in another sense; that is, it is (perhaps unconsciously) based on the feeling that the GM is some sort of glorified computer who automatically produces game results. Certainly a degree of uniformity in how games are adjudicated is desirable. There was too much variation under the old Rulebook. But some individuality, at least in respect to rare and hypothetical situations, is not necessarily a bad thing.

When (and if) situations like these arise…well, GMs will have to do what they’re being paid to do: make a decision.

(Reprinted from Diplomacy World 28 (Autumn 1981)