A Workable Plan

[Zosimus]

My original post was inspired by a private conversation I had with KarmaOverDogma over a year ago. He promised to reveal his strategy for winning during the AAR for the last tournament he played in, but he never did. I have given up hope that he will ever do so.

For those of you who have never heard of KarmaOverDogma, his standard rating is 2391.
As you can see, mine is only 1723.
Big Gun's is 1419.

So perhaps it would behoove us to think on what he has suggested. His exact words to me were:

- First get to the corner or corners
- Don't have anyone at your back for long
- Get over the stalemates whilst allying with one nation until the end where it is a stab to set up for the win.
- Be genuine with your one chosen ally until that fatal moment.
- Back yourself to stab to get up to 12-14 depots. If you're good and over the stalemate line then you'll succeed.

[Big Gun]

For those of you who have never heard of KarmaOverDogma, his standard rating is 2391.
As you can see, mine is only 1723.
Big Gun's is 1419.

Hey Zosi,
I know you don't like me, but there's no need to get yer knickers in a twist over it. Let's not compare apples and pears here. As 75% of the games I've played on this site are on the Versailles map, comparing the All Games ratings will give you a more balanced view.

Get that chip of your shoulder and grow up, for goodness sake! If you allow yourself such lofty opinions, you'll need to learn how to accept criticism and the occasional poke.

[boldblade]

- First get to the corner or corners
- Don't have anyone at your back for long
- Get over the stalemates whilst allying with one nation until the end where it is a stab to set up for the win.
- Be genuine with your one chosen ally until that fatal moment.
- Back yourself to stab to get up to 12-14 depots. If you're good and over the stalemate line then you'll succeed.

All lot of these are principals I have learned from some others on this site and like to employ. They have been working ok for me in anon games but I feel like that only has to happen to you once before you become wary of any good relationship you make in a game going forward.

I feel they definitely don't translate as well in f2f though. In my experience you can only pull that brutal endgame stab on your partner once. Next game you play together you can pretty much guarantee a warning is made to my potential allies from 1901.

[rd45]

- First get to the corner or corners
- Don't have anyone at your back for long
- Get over the stalemates whilst allying with one nation until the end where it is a stab to set up for the win.
- Be genuine with your one chosen ally until that fatal moment.
- Back yourself to stab to get up to 12-14 depots. If you're good and over the stalemate line then you'll succeed.

All lot of these are principals I have learned from some others on this site and like to employ. They have been working ok for me in anon games but I feel like that only has to happen to you once before you become wary of any good relationship you make in a game going forward.

I feel they definitely don't translate as well in f2f though. In my experience you can only pull that brutal endgame stab on your partner once. Next game you play together you can pretty much guarantee a warning is made to my potential allies from 1901.

I guess anonymity has to be the key to getting away with this kind of thing repeatedly. Plus the ability to identify or cultivate a certain naive optimism in one's allies.

My feeling is that once you boil a complex game down to such basic easy-to-follow principles, then the risk is that you only say something banal & obvious. Imagine Roger Federer's tennis game, described in similar terms - all you have to do is keep playing the ball over the net until your opponent makes a mistake or gets out of position, then you hit a winner - repeat until you're the champion. It's true up to a point, but not really actionable.

[rd45]

having Intro-to-Intermediate level discussions of what a "workable plan" is are very useful for the hobby, IMO.

But this, I could get behind. I've often thought it would be useful to have an actionable country-by-country breakdown of high-level plans that mostly turn out well over the course of the first 3-4 years - along with any not-obvious-to-beginners pitfalls to avoid.

Something like this...

As Germany -

• usually, pick one of England & France - whichever you get on best with - to work with against the other - YOU DON'T HAVE TO DECIDE IN SPRING 1901 - string them both along for a while if you can't decide
• your standard opening is to Denmark, Kiel and Ruhr
• be aware that an aggressive and/or twitchy France is going to give you some pain around Burgundy - use whatever diplomatic means to avoid that if poss
• <something about Belgium...>
• DON'T ATTACK AUSTRIA
• know that you'll probably have to fight Russia in the mid game, but don't try & do it too soon
• Prussia, Silesia, Bohemia and Tyrolia should probably all be DMZ, but it's up to you how formally you want to agree that
• you might think it's a good idea to bounce Russia in Sweden, but it usually isn't
• if there's a really obvious R/T, counter it with a western triple
• etc

but... you know... a bit better? And repeated for each power.

[DQ]

having Intro-to-Intermediate level discussions of what a "workable plan" is are very useful for the hobby, IMO.

But this, I could get behind. I've often thought it would be useful to have an actionable country-by-country breakdown of high-level plans that mostly turn out well over the course of the first 3-4 years - along with any not-obvious-to-beginners pitfalls to avoid.

Something like this...

As Germany -

• usually, pick one of England & France - whichever you get on best with - to work with against the other - YOU DON'T HAVE TO DECIDE IN SPRING 1901 - string them both along for a while if you can't decide
• your standard opening is to Denmark, Kiel and Ruhr
• be aware that an aggressive and/or twitchy France is going to give you some pain around Burgundy - use whatever diplomatic means to avoid that if poss
• <something about Belgium...>
• DON'T ATTACK AUSTRIA
• know that you'll probably have to fight Russia in the mid game, but don't try & do it too soon
• Prussia, Silesia, Bohemia and Tyrolia should probably all be DMZ, but it's up to you how formally you want to agree that
• you might think it's a good idea to bounce Russia in Sweden, but it usually isn't
• if there's a really obvious R/T, counter it with a western triple
• etc

but... you know... a bit better? And repeated for each power.

About 15? years ago when they were putting together "Diplomacy, the computer game" I was asked to beta test it. It was awful - anyone who has played it can attest. But I thought hey! Lets give the 7 powers an "opening book" where they pick "a workable plan" and angle for that - they can fall apart after 1904 or even 1903 but having something that could reasonably get through 1902 would be a huge win.

That ... turns out to be much harder than anticipated. I _think_ the work has already been done - people have written scores of articles about opening theory for each power - but aggregating them and getting a game engine to tie them to diplomatic endeavors is a monumental task, certainly one that was beyond the devs at the time. Think about how bad the AI is at diplomacy in something like Civilization - they can teach it the mechanics of the game (which frankly is more than the paradox devs did for Dip) but teaching it how to negotiate is apparently not worth the cycles.

I'm not the smart guy when it comes to this but I've talked with a bunch of them, and I guess I'm hopeful we will get there in the next 15 years or so.

[rd45]

I _think_ the work has already been done - people have written scores of articles about opening theory for each power

That's pretty much exactly the problem, from a beginner's point of view. Advice to a beginner that essentially says

1. read everything that's ever been written
2. implement what you just learned

is not realistic. IMO there's scope for something more accessible, that could short-cut the process of figuring it all for yourself over multiple games. The hard work here isn't to write all those articles, or even to read them - it's to aggregate them into a digestible format.

Digestible for a human being, that is. Teaching an AI is a very different challenge, not one that I have any insight into.

[Nanook]

PlayDip runs mentor games regularly--that's the best way to get the basics, in my opinion. Not just those specifically, but playing games with good players, either in a guided, learning environment, or out in the wild. The games I've learned the most from are the mentor game I played within the first three or four months of finding this site, two games with CharlieP (who no longer graces us with his presence around these parts, sadly), a FtF house game with Doc Binder, and then the first time I played FtF with DQ last May. Now, obviously, if I had started with the FtF game with DQ, I wouldn't have learned as much as I did with having the base to work from--mostly I probably would've learned "Huh, that guy's really good." But I had a base to work from, where I could go "Huh, OK, that's guy's really good, and I can kinda see what he's doing and where I can use pieces of that to improve my own game." Mentor games or guided learning games are the best place to start in my opinion, to get a grasp of "OK, this is what I can do piece-wise, and this is what I can do diplomacy-wise." Then it's pretty much practice, and playing against good players to see what kinds of things they do. I've always preferred to ally with a good player, even (especially) when I know they're better than I am, for several reasons but the salient one here being that you're going to get a better look at how they do things. Because while I'm never going to be DQ, I can still look at how he plays and say to myself "Hey, I can do that, but I don't--why don't I?" Take bits and pieces of stuff from good players that you can do but didn't think to do, and mold it into your own personal style, basically.

Those are the best ways to improve, to my thinking. Repetition and playing with good players. And always being willing to learn and improve, obviously, but that kinda goes without saying!

[DirtyHarry]

Back yourself to stab to get up to 12-14 depots. If you're good and over the stalemate line then you'll succeed.

I don't understand the statement "Back yourself to stab ..." maybe because I'm a Yank? If anyone can clarify, I'd appreciate it.

[DQ]

Back yourself to stab to get up to 12-14 depots. If you're good and over the stalemate line then you'll succeed.

I don't understand the statement "Back yourself to stab ..." maybe because I'm a Yank? If anyone can clarify, I'd appreciate it.

I read it as "Bet on yourself" in the same way you'd "back" a horse at a race or a sports team. If your stab gets you to 12-14, and you're across the line, believe in your ability to finish the job and get to 18. Make _them_ stop you rather than not taking the chance.