Pouch Deposits

The Editor and the Readership

Last time in this column, I presented a couple of my pet failed openings and asked for the readers to comment on them and to follow this topic. Indeed, I did receive a few letters, and you will find them below.

This time, though, my tack will be a bit different. This issue is noticably smaller than the first issue, and so, rather than presenting a topic for discussion in the next issue, I intend to spend some time discussing this difference. Heck, I guess that is a topic for discussion in and of itself, so if you feel like writing to me about the way the five different annual issues of the magazine are organized in my mind, feel free. Yes, I like that. Let's call this my and your topic.

As you surely know, the Retreat issues of the magazine are intended to be smaller than Movement issues. This issue proves the point nicely. Let me give you a quickie history of the last three months.

Most of my time in the last three months was taken up administering the LoebBourse "dippouch" game. Upwards of fifty investors are now active, and each, of course, needed to be supplied with a portfolio for their investments, and their questions answered. There must have been a megabyte of discussion about the rules of the game. (Check out the HISTORY of the game to see what I mean.)

I thank God that Andy Schwarz stepped forward and answered my plea for someone to take the Mastering duties away from me, even though this meant giving up his own investment portfolio. As I have pointed out in the game, Andy is fast making a name for himself in the hobby, and accepting the arduous task of Mastering "dippouch" was not only the best thing he could do for me, but the hardest and most difficult thing he could possibly choose for himself. Take it from a veteran Master and from the one who handed it to Andy: Mastering "dippouch" is the toughest, most demanding job the hobby has ever created. Andy and all who follow him in this task deserve all of our greatest appreciation. He most certainly has mine.

As it happens, my statement about not requiring articles from regular contributors for Retreat issues was heeded by most of them, and this worked out fine for me, since, as I mentioned, my time was totally consumed by the game. So I'm actually real happy with this issue. It confirmed my idea of the prototypical Retreat issue. I hope it doesn't disappoint anyone by its relative brevity compared to the first issue, because I hope it sets a tone which I plan to follow. Retreat issues will be a nice rest for all of us (especially me) who work on this magazine.

This is not to say, of course, that I will ever discourage contributions. No indeed. I'm just saying that it worked out well for me this time that the contributions didn't number as many as last time. It also fit right in with my totally baseless prediction about the size of Retreat issues.

So I guess what I'm saying is, "Isn't this Retreat issue idea great?" There. There's your question. Send me your answer, keeping in mind that the said idea was mine to begin with, so consider your audience (just joking).

Okay, enough about this issue. Let's let you see the letters received from the readers of the first issue.

Mail Concerning The Topic Presented Last Time

From Travis Ruelle (ruelle@comware.com)

I made a very similar opening as Italy to the one you described in your "Pouch Deposits" section, at least in the opening moves. Strong ties to both Austria and France, fake war with Austria, go for Munich. Only France backed out on me, so it took me a little longer to get it. And I didn't go for Warsaw, or try to stab Austria. Anyway, I thought you might like hearing that you're not the only one "foolish" (as the Gamer's Guide calls us) enough to try that.

My response: Good! It's high time all us fools came out of the closet! (Oh, wait...maybe not....)

From Bruce Hitchcock (BruceH@accessone.com)

Manus, you asked for openings to explore, so how about England? Specifically, I have heard advice that if England is going to attack France it must be done at once because France gains too much in the first year. So, what are your thoughts on the F Lon-Eng and A Lvp-Wal opening? Is there any way to camoflauge this opening and use it against the German (perhaps with a convoy into Belgium)?

My response: I'm sure the readership will chime in with all sorts of comments. Here is a quickie from me. The most memorable game I played where this opening was employed by England involved a negotiated E/F/G three-way, which England chose to break right off the bat with German knowledge. Germany, though, saw this rift between England and France as his chance to build three, and got into Belgium with support from both of the combatants. This resulted in a one-turn surprise E/F, taking Belgium from German hands in Spring of 1902. However, it remained under German rule because England and France were very easy for Germany to bust apart diplomatically given their earlier falling-out. Once there was a second disagreement, the two were irreconcilable, and Germany played each against the other for a good long time.

Mail Concerning Other Articles

From Mark Gilbert (markg@winternet.com)

Regarding Simon Szykman's article on dedication points: I liked your article in the Dip Pouch about dedication points. I am also frustrated by the slow progress and abandonments in the games that I am in. I definitely agree that the -1/+3 rule is quite silly.

The other thing I have wondered about is a possible bonus for joining a game that needs replacments. It would obviously have to be much smaller than the penalty for quitting, but this might help encourage people to help out with mercy positions, etc., and get games going again faster.

It seems like some games wait for a long time for replacements. To try to help out, I make half my game starts replacements, trying to vary the quality of the positions that I take on.

Simon's response: Thanks for the message. I'm glad you liked the article.

Your point about a bonus for joining games which are waiting for replacements is a good one that I would have been worth mentioning. But the reason I didn't mention it is not because I hadn't heard of the idea, since I have been following the recent dedication discussions on rec.games.diplomacy. It actually looks a lot like a thread that went on around a year ago, in which I was a much more vocal participant. In fact, I may have started the thread but I don't recall for sure. At the time I was in something like 18 games at once, only six of which were not stalled for one reason or another.

At any rate, I was pushing pretty hard to get the dedication system reformed and there were many ideas, including the one about giving people points for signing on as replacements. The concensus among a few of the people who actually write judge code was that things like changing late penalties or CD penalties are relatively minor changes - just a number here or there. On the other hand, giving points to replacements would take quite a large amount of time and effort for writing new code. They said they'd be happy if somebody took that on but that they had things on their to-do list that had a higher priority so the chances of that actually happening were slim (there actually is a to-do list for judge programming).

Since then, I've directed my efforts at convincing people about the things that are easy to modify, since those are the things that are actually likely to get changed. When the idea about dedication points for replacements first came up, I thought it was a fine idea and I still think it is. But, since I don't think it's likely to change, I've stopped making a big push on that point.

I think your personal decision to make half your gamestarts replacements is great. I've recently signed on to three games that needed replacements. If more people did that, we wouldn't need to suggest giving dedication points as an incentive.

From Danny Loeb (loeb@delanet.com)

Regarding Jamie Dreier's "Mastering Newbie Games": Jamie mentions enforcing the rule against negotiating past the deadline. I'm not sure if this is in the house rules. (Why not include a copy of the rules in DP!) In any case, I don't see why such a rule should apply to the people who got their moves in on time. If this really was so bad, then the Judge itself would outlaw it, since that would be easy to prevent.

Jamie Dreier's response: If I wrote something inconsistent with this rule, it was a typo. I endorse this rule entirely. I think Conrad Minshall put it well. It is not fair for some players to gain a negotiating advantage by negotiating past the deadline, when others are sticking to the deadline. Thus, anyone not marked "late" should be free to negotiate with anyone else not marked "late." But no negotiations by or with lateniks should be tolerated. Of course, it's up to the gm. But I'm glad it's in the house rules this way, and I run my games by it.

Also from Danny Loeb, regarding Dan Shoham's "Diplomacy Academy": I liked the DipAcademy articles. Are we going to see any articles in which he loses!? Or is he unbeatable!?

Also from Danny Loeb, regarding my "Missing Man Diplomacy" variant: I suggested once some other ways to cover an empty position. (By the way, these rules could also be quite easily adapted to cover NMR and CD situations as well.)
  1. Before each phase and random player is chosen to submit orders for that country.
  2. After submitting orders but before revealing them, a random player is chosen to submit orders.
  3. All players submit their orders along with orders for all unplayed countries. A random set of orders is chosen for each unplayed country. (Or a random order is chosen for each unit of an unplayed country.)
  4. All players submit their orders along with orders for all unplayed countries. All orders are revealed. Each unplayed country follows the set of orders identically submitted by the most countries. A random choice breaks any tie. (This makes negotiation quite interesting as players campaign for THEIR favorite set of orders.)
  5. The weakest player is given the right to submit orders for the strongest unplayed country. The 2nd weakest player submits orders for the 2nd strongest unplayed country, and so on.

Also from Danny Loeb, regarding Simon Szykman's article on player dedication: I like the following system of penalties. We don't assume that a player who is late or abandons is "bad" but simply busy. Thus, if a player is late, he is black-listed for a few days from taking on new committments. If a player abandons a position, he is black-listed for say a month. Such "penalties" could be sent by Judge's to an automatic server that Judge's consult after signons to a game. A human manager could remove these "penalties" on request. (Ie: if due to judge error...) The penalties would be cumulative and apply to all judges involved with this system. Players would be allowed to continue their existing ONGOING games regardless of penalties.

It would be nice to add a column to DP about work going on improving the Judge code. There is a list "judgemaint". Could someone there become a columnist? Perhaps readers could make suggestions. I have one:

Sometimes a game starts with players unsuitable for the game for one reason of another.

Or sometimes a player takes over an open position that wasn't quite that open.

I suggest that all current situations where a player is assigned a power, be replaced with a situation where the player receive a letter saying that he is being considered for the following game (and giving him the chance to back out immediately), and the GM receives a letter giving information about the player. The GM would have a special command allowing him to install the player in the position he has requested.

From Mark Nelson (fuemin@sun.leeds.ac.uk)

Regarding Jamie Dreier's article on the lack of Diplomacy opening theory: Whilst I agree with the thrust of Jamie's article, I disagree that the "explosion of possibilities threatens to make opening theory unmanageable".

To return to Chess. If I always open 1. P-Q4 then I must prepare myself for one of a number of defences (there are at least 10 common defences to this opening). In response to each of these choices I must prepare my variation. But, typically, Black will have 3 or 4 responses to each of my variations. So I must prepare a response to Black's subvariation...

Opening theory in Diplomacy starts from the premise that you know what you want. Suppose that I am playing France. Then I have a large number of possible openings. Certain openings are "better" if I want to be pro-England, others if I want to be pro-Germany. Can I rely on Italy not to open agressively against me? Am I going for two or three centers? Where do I want to put my fleet? There is no one great opening, but given certain assumptions about the state of diplomacy there are better and worse openings.

It is interesting to note that several authors (including myself) have written articles examining how a nation's choice of openings effects the result of the game. There are trends. We can't say that "this is a bad opening, don't use it," because there are always games where a country chose it and did well (in one postal game Germany NMR'd in Spring 1901, but still won the game!) but we can say that "France scores significantly better with this opening."

It is worth remembering that certain authors (such as Richard Sharp) have developed opening theory based around certain ideas, e.g. that Germany must keep Austria strong or that Italy and Austria must ally together against Russia and Turkey.

From The Unknown Reader

Regarding Jamie Dreier's article on the lack of Diplomacy opening theory: After reading your interesting article on "Opening Theory" I was struck by one thought. I would not consider the elements a matter of luck. Too much resides in the knowledge of your opponent and in the knowledge of the importance of strategic locations. Many people, when placed in the tactical position you described would have a specific reason to defend or attack one space over another.

The cautious player would immediately protect his home territory, while a more risky, expansionist player might try to defend the frontier. This plays a much greater role in the (addmitedly few) games I have played.

In the end, there are no completely random events, and most events that some might consider luck, often rely on trying to predict an opponents psyche. Addmitedly not a very precise thing, but not random either...

Jamie's response: I don't agree. Let me explain.

Take the main example I gave, the one where Turkey is trying to invade Italy.

Take Turkey's role for a moment. You are trying to guess whether Italy will protect Naples or Tunis. You realize that he is trying to guess, too, of course. You try to assess whether he is conservative or risky. Suppose you (correctly) decide that he is a risky-type. Does that mean he will cover Tun instead of Nap? No, it doesn't.

As for many, many situations in Diplomacy and other imperfect information games, the 'best' strategy (technically, the Nash equilibrium) is going to be a mixed strategy. A typical mixed strategy is, flip a coin, cover Nap if it comes up heads and cover Tun otherwise. There are many mixed strategies available. Actually, there are infinitely many. In fact, a large infinity! An uncountable one. Essentially, the coin Italy flips could be weighted in any number of ways.

If Italy adopts a mixed strategy, there is no way on earth that Turkey can figure out what Italy will do, not just by his smarts. He *has* to be lucky.

This sort of thing happens all the time. Practically every move in Diplomacy involves mixed strategies in the ideal case.

Now, as a matter of practical gamesplaying, you may be right that in practice many, most, players will not adopt a genuine mixed strategy at all. There is certainly room for some psychology. But I don't believe that all the psychology in the world will let you figure out for sure what the other guy is going to do.

Whether there are truly 'random' events or not is actually a red herring, I think. All that matters is whether the next move is predictable given the information available to players. So, even if coin flips are deterministic, we can regard predicting Italy's coin as out of the question!

(I would be very pleased to have Dan Shoham, or some other accomplished Game Theorist, comment on the above.)

From Michael H. Ross (mhross@sunset.backbone.olemiss.edu)

Regarding John Woolley's first mystery for "Sherlock Holmes, Consulting Diplomat": Congratulations on a fine puzzle! Trying to figure out how a second Russian SC would be lost took quite a while.

I am a chess puzzle fanatic, but Diplomacy puzzles (especially this one) are, IMO, much more challenging, since Diplomacy has greater flexibility in how pieces can move.

The Rest of the Mail

Out of modesty and lack of space and editing time, I have deliberately omitted the mass of the mail I received about the first issue of The Pouch. These, understandably, were all ebulient compliments about that first issue. Although I decided not to inflate my head all the more by reprinting them here, let me just say to all those of you who wrote with such kind and congratulatory words that I appreciated them very very much. Thanks, and keep reading!
Well, there you have it. Another Deposits column. Yep, it sure is.

Manus Hand
Your Publisher

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