The Game of Diplomacy

by Richard Sharp



Diplomacy was invented by Allan B. Calhamer, an American graduate student of history, political geography and law, all of which disciplines served him well in perfecting his game. Unlike so many modern boardgames, Diplomacy received careful testing and constant revision before being marketed in the form we have today. The idea which began to take shape in Allan Calhamer’s mind as early as 1945 did not in fact reach the public until 1959, by which time it had been polished and refined into a superbly balanced game. As we shall see later, the balance is as nearly perfect as one can wish for, though perhaps the natural problems inherent in the terrain cannot ever be entirely solved.

The Diplomacy board is a simplified map of Europe as it was in 1900. Some changes have been made to the real map - Bulgaria, for instance, did not have its ‘south coast’ at that time but the overall accuracy is surprisingly good when one considers that playability, not realism, was uppermost in the designer’s mind.

On this board, seven players, each representing one of the seven great powers’ of the period, manoeuvre their forces in an attempt to become masters of Europe - an event which is symbolized by control of over half the available resources, at which point the game is ended. The great powers concerned. with their distinctive colours, are as follows (where two colours are given the second is that used in the standard British set, which differs from the American one). Initial fighting strengths are also shown.

ENGLAND Dark blue/pink 2 fleets, 1 army
GERMANY Black 2 armies, 1 fleet
RUSSIA White/purple 2 fleets. 2 armies
TURKEY Yellow 2 armies, 1 fleet
AUSTRIA-HUNGARY Red 2 armies, 1 fleet
ITALY Green 2 armies, 1 fleet
FRANCE Blue 2 armies, 1 fleet

On the subject of the two different sets (there are other, obsolete ones too), it may be said that the American set is far more businesslike and attractive, the British version being designed as a ‘popular’ one, with bold, poster-like colours and ornamental drawings; the American pieces are wooden, the British plastic. I have never met anyone who did not prefer the American version, though the British one does have the virtue of being a good deal cheaper.

The playing area is divided into seventy-six named ‘provinces’: fifty-seven being land (including the ‘impassable’ Switzerland, which alone among the fifteen neutrals cannot have its status violated by a great power), and the other nineteen sea. There are also unnamed provinces, such as Ireland or the Caspian Sea, which play no part in the game and are included only for appearances.

The sea provinces all have the same characteristics, but the land ones are of several different kinds.

Home supply centres: These are the starting-points of the units which the great powers have at the start of the game - Russia contains four, the rest three each. They are of paramount importance, as they are the only provinces in which new units can be built as the game progresses. A country which loses all its home centres may still gain centres elsewhere, but cannot raise new forces until it has recovered at least one home centre. The home centres are all named cities, e.g. Edinburgh, Venice, Ankara.

Neutral supply centres: There are twelve of these, scattered about the board, and the scramble for possession of them dominates the opening of the game. They are named for minor countries, e.g. Serbia, Sweden, Spain, with the odd exception of Tunis (why not Tunisia?).

Others: The remaining twenty-two land provinces (excluding Switzerland) have no intrinsic value, and serve only to extend the playing area and allow room for manoeuvre between the supply centres. Apart from the two neutrals, North Africa and Albania, they are all nominally parts of the great powers and are named for provinces of those powers: Yorkshire, Burgundy, Armenia, and so on. However, they might just as well be neutral - the presence of a hostile unit in one is not important in itself, though it may obviously pose a threat to an adjacent supply centre.

On this varied playing area, the seven players move their armies and fleets in an attempt to occupy neutral or enemy supply centres, each of which entitles them to raise more armies and fleets, and so on. The playing time is divided into ‘game years’, and each year is subdivided into seasons: in spring, players move their units; in summer, units dislodged during the spring offensive make their retreats; autumn is the second movement season; and winter sees more retreats, plus ‘adjustments’, each player adding or removing units until his forces equal the territory he controls, one unit per supply centre held. Then back to spring again, and so on. The game begins in spring 1901; its length may vary greatly, but the average game lasts until about autumn 1910. (The record for an outright win is autumn 1904; the longest game lasted until autumn 1929. Both these were postal games, but it seems unlikely that either record would be bettered in face-to-face play.)

It is worth pausing at this point to consider one of the most important facts about the game of Diplomacy. Each player is outnumbered six to one; none can hope to achieve the victory criterion (control of eighteen supply centres) by brute force, or even by unaided subtlety. The game, in short, is based on a paradox: I cannot win unless you help me, but you want to win too, so why should you help me? This problem is at the root of all the different strategies employed in the game: in the simplest idea, two players unite, assisting each other to become powerful, each confident that when the time comes he can decisively attack the other and win. In the more complex sort of game, several countries - ideally all seven - attempt to play off their neighbours against one another, some growing steadily fatter on the proceeds of other people’s wars while others find themselves mysteriously dwindling. I shall look at some of the strategic ideas in more detail later; for the moment it is enough to recognize that the discussion on tactics and movement which follows is merely an introduction to the tools of the game. Mastery of the moves is vital, certainly, but on its own it achieves nothing.


Movement at Diplomacy is controlled by written orders, which are exposed and executed simultaneously by the seven players. Each player writes his orders in secret, after a period allowed for negotiation with other players:

the orders are then handed to a ‘gamesmaster’ if one is available, though in face-to-face play this is a rare luxury, and orders are more usually stacked face down on the table and read aloud by one (or more) of the players when all orders have been submitted.

The writing of clear, unambiguous orders is not as simple as one would expect; in the heat of battle many mistakes are made, and strict discipline is needed: ambiguous and illegible orders fail, and units so ordered are at the mercy of their enemies. It is highly advisable to use standard notation (see Appendix); a player who writes an order which is capable of misinterpretation is unlikely to find much sympathy. As we shall see later, the deliberately miswritten order has its uses, which makes it even more important that orders be strictly adhered to and doubtful ones rigorously banned. A player who has promised to move, say, from Norway to the North Sea, and has decided not to go, will not thank a gamesmaster who, confronted with the order F(Nor) - Nor, allows himself to be persuaded that this aberration should be ignored. Errors of this type by the high command under severe pressure are a familiar enough story in real warfare, after all.


The basic move at Diplomacy is from any province to any adjacent province. This simple rule is however modified by the nature of the terrain and of the units involved.

Armies may move between any two adjacent land spaces (note that Spain and North Africa are not adjacent, as is apparent on the American board but not on the British one, though the latter is likely to be redrawn with this error corrected). They cannot move to any sea space, but may be carried across one or more by friendly fleets. Thus an army may be convoyed from Wales to Brest by a fleet in the English Channel, or by two fleets in the Mid Atlantic and the Irish Sea. In an extreme case an army may be convoyed from Smyrna to St Petersburg - one British postal game, 1972-K ended in a two-way draw between Russia and Austria in autumn 1915, and the last Russian orders included: ‘A(Smy) - StP C by Austrian F(AES), Austrian F(IOS), Austrian F(TYS), Austrian F(WMS), F(MAO), F(ENG), F(NTH), F(NWG) and F(BAR).’ The move had no tactical value, being more in the nature of a lap of honour, but one can understand the attraction!

Fleets may move freely between adjacent sea spaces, or from any sea space to any adjacent land space, or vice versa. They may also move between land spaces which are adjacent along the same coast: thus a fleet in Rome may move to the Tyrrhenian Sea, Naples or Tuscany, but it may not move, as an army could, to Venice or Apulia, which though adjacent and coastal are not adjacent along the same coast.

Some provinces present special complications for fleets. Bulgaria, Spain and St Petersburg all have ‘split coasts’, and a fleet moving to those provinces must be on one coast or the other. Thus a fleet moving from the Black Sea to Bulgaria can only arrive on the East Coast, and cannot move on next season into the Aegean area. When ambiguity is possible, the coast intended must be specified: the order F(MAO)—Spa is illegal, since the fleet could move to either coast; the order F(GOB)—StP would usually be regarded as legal, since the fleet can only move to the South Coast, but good players habitually write the coast in every case, to be on the safe side.

Provinces of apparently similar type which do not have split coasts are Kiel, Denmark, Sweden and Constantinople: a fleet may enter all these from one side and leave the next season from the other. Note however that they are all unequivocally land spaces, and fleets occupying them may not convoy, a rule which has created some havoc in the past among the wargaming fraternity.


Only one unit may occupy a province at any one time — from this simple rule springs all the mayhem and slaughter that make Diplomacy such good, clean fun. If two or more units attempt to enter or hold the same province unassisted the result is a ‘stand-off’ and no unit moves, thus:


However, any unit capable of moving to a specific province may also give support to another unit, friendly or hostile, attempting to enter or stand in that province. Each unit moves or stands with strength of itself and all its valid supports, and in case of conflicts the better supported unit prevails. If one unit succeeds in entering the disputed province, any unit already in occupation is dislodged, and must retreat or disband. In Diagram 1, Russia and Germany are involved in a typical frontier clash: Russia must hold his home centre of Warsaw against the German attack, and does so easily with

Diagram 1

A(Liv) and A(Mos) S A(War), which gives the Warsaw garrison a strength of three, equal to the best Germany can do.

Support may be cut by an attack from the side: thus, in the diagram, give Germany an additional F(BAL). This unit is not in a position to make a direct attack on Warsaw, yet its presence ensures that Warsaw must fall:

GERMANY A(Gal) & A(Sil) S A(Pru)-War, F(BAL) - Liv
RUSSIA A(Liv) & A(Mos) S A(War)*

The Germany A(Pru) enters Warsaw, and the Russian army must retreat to Ukraine (the only available space), or disband.

Note that Germany could not be certain of taking Warsaw with either of his other two armies: instead of ordering A(Liv) to support A(War) Russia can and usually should order it to attack Prussia; so that if Germany attempts to enter Warsaw from Silesia we get:

GERMANY A(Pru) & A(Gal) S A(Sil) - War, F(BAL) - Liv
RUSSIA A Liv - Pru, A(Mos) S A (War)

What has happened here is that Russia has effectively neutralized the German fleet and made the contest a three-against-three one; we can see in this one of the simple tactical truths of Diplomacy, that an attack is usually more effective than support where the choice exists.

One important point for novices that can be brought out of the above diagram has caused much confusion to those with the old (1962) edition of the rulebook. Among many defects, this edition did not make it sufficiently clear that to cut a support one must attack the supporting unit from the side. Thus in Diagram 1, with the added German fleet, Russia cannot save himself with A(War) - Sil, A(Mos) S A(Liv) - War: the attack on Silesia comes from Warsaw, the province into which support is being given, and the Silesian support is thus cut, so that the German A(Pru) is again moving with two supports, the singly supported move of the Russian A(Liv) fails, and the A(War) is once again dislodged. This rule is obviously a sound one, as it ensures that in any two-to-one confrontation with no interference the two must prevail; I was among many who were misled by the old rule-book into thinking that the defender was on a fifty-fifty guess, needing to attack the supporter, rather than the attacker, to survive. The hobby owes a great debt to the pioneering American GMs and players who found and corrected these anomalies in the rules.

Another feature of these situations that sometimes causes confusion is that one cannot cut the cutting of a support; using the same position as before:

GERMANY F(BAL) - Liv, A(Pru)& A(Gal) S A(Sil) - War
RUSSIA A Liv - Pru, A(Mos) S A(War)

The Russian attack on Prussia cuts one of the supports for Germany’s attack on Warsaw, as shown above; the German attack on Livonia does not affect the cutting of the support in fact, F(BAL) - Liv is completely ineffective in this example. In the 1960s Conrad von Metzke was advocating a change here, but found little support.

One more interesting rule concerning supports is the so-called ‘beleaguered garrison’ rule, which states that if a province is attacked from two (or more) sides by equal forces, any unit occupying the province cannot be dislodge, even if it is weaker than all the individual attackers. Thus in the diagram assume that Russia can judge that the German A(Gal) will be occupied elsewhere, perhaps supporting an attack on Rumania: he can now order A(War) - Gal (cutting the support for this secondary attack) and A(Mos) S A(Liv) - War. If Germany orders A(Pru) S A(Sil) - War, or vice versa, the solitary unit in Warsaw is attacked by two supported armies, yet survives. It would still survive even if both the attacking forces were not Russian.

Finally it is important to understand the effect on support of the split-coast provinces. To my mind the new rulebook is clear enough on the subject, yet some quite experienced players have run into a mental block here. Briefly, a fleet adjacent to such a province may support any action in that province, even on the far coast: thus F(GOB) S F(Nor) - StP(nc) is a perfectly valid order, although F(GOB) could not itself have moved to the North Coast. But the converse is not true: a fleet already in a split-coast province can support actions only in provinces it could itself have moved to: F(Spa)(sc) can support an action in Marseilles but not in Gascony. Nothing of this contravenes the normal rules, in fact, yet the position does seem to cause problems.


A unit attacked by a superior force is dislodged (unless saved by the beleaguered garrison rule), and must retreat or disband. It can retreat to any province it could have moved to, provided that the province is still vacant, and is not vacant as a result of a stand-off. If unable to retreat, the unit is annihilated; it may also be voluntarily disbanded, even if a retreat space was available. This can be desirable, for instance, if a player wants to get the so-called ‘fast retreat home’: if Russia, say, gets a fleet disbanded or annihilated in the Norwegian Sea, he may (provided he has not lost a centre during the year) rebuild it as an army in Sevastopol, or anything else he chooses.

One important rule is the effect of a successful attack on a convoying fleet:

this not only prevents the convoy succeeding, but effectively cancels it, so that the convoyed army has no effect on its destination space. For instance:


FRANCE F(MAO) S F(Bre)-ENG, A (Bur)-Bel

Because the English fleet is dislodged, the A(Lon) - Bel becomes, in effect, an illegal order, and the unsupported French order A(Bur) - Bel must succeed. This rule has one unfortunate effect: it leads to a genuinely unresolvable conflict, where two rules meet head on and it is impossible to say which should prevail. Such paradoxical situations are very rare in practical play; if one does arise, the ruling should be that no units move. I shall return to these oddities later.


Another clarification provided by the new (1971) rulebook concerns units changing places. In the ordinary way, two units cannot exchange places (‘Changing the Guard’), even in cases where logically they would easily be able to do so: e.g. F(Spa)(nc) - Por, F(Por) - Spa(sc), the so-called ‘coastal crawl’, is not permitted to succeed, though in reality the fleets would not come within a hundred miles of each other. This is a rare case where the revised rulebook went against majority opinion - most GMs before 1971 allowed this move. However, if one (or both) units are convoyed, both moves succeed: A(Ank) - Con, F(BLA) C A(Con) - Ank. This latter ruling is often useful in the Denmark area (see ‘The Loop’).

Diagram 2

Three or more units may always change places on a rotating basis, however (the ‘Merry-go-round’). This is a very important rule, easily forgotten. In the complicated four-against-four situation shown in Diagram 2 Austria, under attack from an Italo-German coalition, needs to hold Trieste and Vienna for one season, to gain time to bring reinforcements in. It looks as though Austria can be sure of saving the situation with A(Tri)-Tyr, A(Ser) S A(Vie) - Tri, A(Gal) - Boh : Vienna is safe, since both possible supports are cut; Trieste is safe, since it is being attacked by an Austrian force equal to anything the enemy can field. Unfortunately the attackers have only to order the simple A(Tyr) - Vie! and the whole position collapses, as the three units involved - A(Tyr), A(Tri) and A(Vie) - all rotate one space clockwise. Vienna falls and so does Galicia, as the two German armies are free to attack that.

I remember once in a postal game (1973-AK) claiming a stalemate based on a very similar position: neither the gamesmaster nor my opponents noticed the fallacy. (I hasten to add that I had not noticed it either.) The position arises very frequently in the Austrian sector, and also around Warsaw. In the diagram position Austria is on a guess, and has the odds in his favour, since the opponents have to guess exactly right with several alternatives to choose from.


Although units of two or more countries may collaborate, the effect is not always quite the same as when units of a single country are involved. The joker in the pack is the rule prohibiting self-dislodgement: if Austria, say, has an army in Vienna he may use another army to support a Russian attack on Vienna, or receive support from Russia for an attack of his own, but if the Austrian unit already in Vienna does not succeed in moving out the attack cannot succeed. This would be equally true, of course, if both the units attacking Vienna were Austrian. However, if both are Russian, even though Russia is attacking for Austria’s benefit and at his request, the attack may succeed. This means that international operations often have less flexibility than one-country ones (realism again!). As we shall see later, though. two countries have one advantage that a single country has not - the ‘planned retreat’.

One rather absurd situation that can arise where two allies get their wires crossed does, in effect, permit self-dislodgement: if you attack a foreign unit which is supporting one of your own units you do cut the support, and if your unit is attacked from elsewhere with sufficient support it is indeed dislodged. This does occasionally happen, amid scenes of great hilarity and embarrassment. But you cannot cut a support being given by one of your own units.

There will be more to say about international supports when we come to look at the common tactical manoeuvre known as the self-stand-off.


This chapter has dealt with the bread-and-butter of Diplomacy skills, the things a player must get right to stay alive. Without complete familiarity with the rules, any view of strategic concepts or tactical possibilities is necessarily opaque. Bread and butter, though useful, is dull; let us proceed to the jam.

The Game of Diplomacy is (c) 1978 by Richard Sharp. It is reproduced here with Richard's permission.

Back to Game of Diplomacy Main Page


Home - About this Site  - Diplomacy Rules and Maps - Dip Strategy - Variants - Dip Software - Play Diplomacy - FtF Diplomacy - Postal Diplomacy - Diplomacy Humour - Tournament Scoring - Dip Hobby History - Zines - Con Reports - UK Zine Archive Miscellaneous - FAQ - Links - Recommended Reading