On Convoy Path Specification

Allan Calhamer

What This Is

As a longtime devoted e-mail player, your humble publisher approached the inventor of The Game with a suggestion that the Internet "judge" computer's convoy syntax be adopted for the Hasbro computer and board game.

For those of you who may be unaware of the judge's syntax, and with my reasons for preferring it over any other convoy syntax, I reproduce my letter to Mr. Calhamer, and then his response.

My Inquiry


As you might be aware, I was asked to proof-read the game manual for the Hasbro CD-ROM and to contribute a couple of articles to it. Like tens of thousands of others, I have played for years using Internet judge computers, which require a convoying army to specify its convoy path, thus:

A Lon - NTH - Bel

This took just a little getting used to, as I had of course played face-to-face without this little nuance. However, the convention is so extremely useful (not only to the computer but to avoid ugly points of the rules) that the multitudes of online players quickly find themselves swearing by it.

This rule avoids all the "unwanted convoy" problems. The only "tactic" it removes (assuming that the kidnap convoy is not a tactic that should be condoned -- and I have heard from Edi Birsan that you have come out against it; yay!) is the provision for multiple convoy routes. For example:
      A Lon - Bel
      F NTH C A Lon - Bel
      F ENG C A Lon - Bel
(with the convoy going through [per the 1982 rules] even if one of the fleets is dislodged -- as you know, the 1976 rule had the convoy failing).

There are two schools of thought on this:

  1. Multiple routes should not be allowed anyway.
    The provision for multiple routes, and allowing the army to follow whichever route is not disrupted is analogous to the situation of having to guess which opposing unit is supporting and which is moving. It seems to those in this position that the ability to have extra routes should not be allowed, since it effectively means you don't have to guess in a situation in which, elsewhere on the board, you would have to guess.

  2. Multiple routes should be allowed but...
    These people have no problem with the judge computer's rule to specify a route, because the fact is that the wish to specify multiple routes is so extremely rare that, in practice, the issue never comes up. However, they stand on principle that -- since multiple routes are allowed by the face-to-face rulebook, they should be allowed by the judge computers.
My question to you is: In which school of thought would you count yourself?

It seems to those of us who are familiar with the judge computer's syntax that it might should be adopted by you and Hasbro if you like it. The reason being that it completely eliminates all ambuiguity in orders. Frankly, the part in the rules which forbids ambiguous orders has always seemed to me to be contradicted by the fact that orders like:

A Bel-Hol
A Hol-Bel
F NTH C A Bel-Hol
are considered valid. That seems very ambiguous to me. It may be a kidnap but it may be Germany helping his Anglo-French partners swap centers.

I have heard that you are in favor, in this case, of requiring any army that could reach its destination overland but wishes to do so by convoy to actually specify that it wishes to be convoyed. I think this is an excellent decision.

I am writing to ask if you would consider extending that requirement, for the sake of consistency and simplicity, to adopt the judge computer's rule that the full convoy path be specified in the army's order.

I myself am currently in school of thought #1, and do not consider myself a big fan of allowing multiple convoy routes. I am curious to hear your opinion on these issues.

I did not bring the judge computer's solution up with Hasbro since it is non-Calhamerian and I knew they were taking direction from you. But now that the hobby is generally aware of the contents of the game manuals, there has been an expression of sentiment for making an attempt to get the nice, clean solution of the judge computer adopted by the powers that be. And of course, that means you.

I do not know if you have ever even been made aware of this solution. If this is the first you have heard of it (or even if it isn't), please let me know your opinion of it.

I apologize for the length of this message, but it is a matter that is considered very important by the hobby, so I am curious to know your response.

Stab you soon,

Mr. Calhamer's Response

Dear Manus,

I have read with interest your well-thought-out arguments against the rule permitting multiple convoy routes. You have noted that earlier they were not permitted. I decided to permit them in response to fan representations; now, quite a few years later, I find fans building up an attack on that position.

In a discussion of a case in which France has the position A Tun, A Rom, F TYS, and F ION; and Italy has the position F LYO, F ADR, and F Nap.

And the French attack thus:
      A Tun-Nap
      F TYR C A Tun-Nap
      A Rom S A Tun-Nap

Against this would be offered the defense:
      F LYO-TYS
      F ADR-ION
      F Nap S F LYO-TYS
; or, in the alternative, F Nap S F ADR-ION.

It appears that this attack wins 100% of the time. Except for this attack, all optimal attacks result in guessing games.

If the multiple convoy routes were not permitted, all optimal attacks in this situation would lead to guessing games.

Obviously the good player would be likely to order the multiple convoy routes.

A defense of the present rule might go as follows:

It should be mentioned that the attacker can achieve this 100% certainty only if he has four pieces to the defender's three, and a precise position. Interestingly, the defender actually outnumbers the attacker in Fleets, three to two; but the attacker's Fleets are perfectly placed for the attack, whereas the defender's Fleets are scattered and out of touch with the objective.

Direct analogy of convoys to supports suffers because supports fail if attacked with a force of one; convoys fail only if attacked with a force of two. Furthermore, supporting a supporting piece does not protect at all against the cutting of the support of that supporting piece; whereas supporting a convoying Fleet does protect it against vitiating the convoy by dislodgment.

Thus convoys are stronger than supports anyway. In return, of course, you can't use them in as many situations.

Also under the present rules it is all the same whether one convoy route succeeds or more than one; whereas it makes a difference how many supports to a given action survive uncut. This much in and of itself seems realistic.

You argue that the attacker is allowed too many options. If one convoy route fails, he has the automatic option of using the other. It is true that if a player writes a conditional order it is no good. However, the orders generally have certain well-recognized conditions implicit in them. Thus if you order an Army to support, you know that, on the condition that it is attacked, it will ignore its order to support and defend its position. Frequently you will rely on this conditional possibility when you choose the support order.

Your argument from realism, and harmony with the tenor of the rules generally, that the Army would not know in advance which convoy route would be disrupted, is a hard one to meet. However, we might reasonably suppose that maintaining control of a body of water was a six-month job for a Fleet. Each convoying Fleet gives up its move for this task. It might further be supposed that crossing the body of water, capturing a province on the far shore, setting up on a new shore, being supplied by sea, and so forth, was a six-month job for an Army. The ousting of a convoying Fleet might be seen as a deterioration of sea control, similarly over a period of six months. Now the Army, engaged in a six-month operation to gain a foothold on the far shore, might easily during that time be diverted entirely to the other convoy route.

The counter-argument, however, that, if only one route were specified, and it were disrupted, then the Army should be free to move overland in an entirely different direction (i.e., conditionally), would fail because the six-month period would have been devoted to attempting the crossing.

Thanks very much,

Allan Calhamer

If you wish to e-mail feedback on this article to the author, and clicking on the envelope above does not work for you, feel free to use the "Dear DP..." mail interface.