IN DEFENSE OF A SMY H

by Peter-Paul Koch


I recently drew Turkey for the third time in a row, and that did not appeal to me at all. I like a bit of variety in my Powers, but more importantly my Turkish games are very stressful. In the first game I was brought down to 2 centers and managed to crawl back to 9 centers until I was part of a three-way draw. That sounds cool, until you have to do exactly the same in your very next game (except that I got 11 centers, not 9). I had no particular wish to go through the process a third time.

Still, it was clear that the Universe ordered me to improve my Turkish play. So here we go again. But letís use our brains this time.

As usual, initial press was vague, and despite their courteous replies I trusted neither Russia nor Austria. So I went into the Spring 1901 moves without a clear alliance — or even a clear tactical goal beyond picking up BUL. What I did have was distrust of Austria and Russia. Especially of Russia, since heís closer. And bigger.

Thatís when I decided to play:

A CON - BUL
F ANK - BLA
A SMY H

Itís that last order Iíd like to discuss in this article. As far as I know it has never been treated systematically, even though I think that, in certain situations, it is preferable to the more common A SMY - CON.

I analyzed 252 games from the DP Judge ďowls_Ē series that ran from 2005 to 2012, noting Turkeyís opening with A SMY, and especially the A SMY - CON - BUL sequence. I found that in only 19 of those games (8%) the Sultan orders A SMY H in Spring 1901. Moreover, in 16 of those games it was combined with F ANK - CON, obviously as part of some Russo-Turkish deal. Only three out of 252 games used A SMY H in combination with F ANK - BLA, as I did.

So either Iím a Dip genius whoís invented a new Turkish opening, or Iím a silly newbie with no knowledge of opening literature and not enough wits to see the obvious deficiencies of his idea. This article tends toward the former point of view, but the truth will, boringly, lie somewhere in the middle.


A SMY openings

What is A SMYís best opening move? Thatís clear: A SMY - ARM. It deploys the army satisfactorily and brings it in direct contact with a pickup possibility, SEV. Combined with a succesful F ANK - BLA it is the most powerful Spring 1901 move Turkey can make. Thereís but one tiny problem: it plunges you into an irrevocable death struggle with the Tsar. If thatís what you want, go to ARM. If thatís not what you want, do something else.

So. Which orders do Sultans give A SMY?

Spring 1901 moves of A SMY
Move Absolute Percentage
A SMY - CON 133 53%
A SMY - ARM 61 24%
A SMY - ANK 38 15%
A SMY H 19 8%
A SMY - SYR 1 0%
Total 252 100%

Nothing too surprising here.

  • The most popular order is A SMY - CON. Over half of Sultans opt for it. This article will argue that that popularity is somewhat undeserved.

  • One out of four Sultans go to war with Russia immediately by moving A SMY - ARM. I laud their courage, but so far I declined to follow their example.

  • Next comes A SMY - ANK, in 31 out of 38 cases combined with a pro-Russian F ANK - CON. There is nothing the army can do in ANK what it cannot do in SMY, so this move is useless. These Sultans give A SMY something to do because all units should do something in Spring 1901.

  • 19 Sultans play A SMY H, but 16 of them combined it with the pro-Russian F ANK - CON. This is the more honest version of the moves in the previous bullet point.

  • Then thereís the lone soul that played A SMY - SYR. Letís take a moment to appreciate his capacity for confusion.

Weíre left with three eaxct copies of my move, and seven cases where A SMY - ANK is combined with F ANK - BLA, which sets up nearly the same tactical and diplomatic situation, though I will argue later that the hold order is slightly preferable.

Ten out of 252. 4%. Not a lot.


A SMY - CON — and on

Why do half of the Sultans play A SMY - CON? The commonly stated reason is that in Fall it could move on to BUL in case A BUL picks up one of the other Balkan centers.

Does the data bear this out? Letís define ďsuccessĒ as the picking up of an extra Balkan center, and see where that leads us.

Spring 1901: A SMY - CON
Fall 1901 move of A CON
Move Absolute Percentage Success Failure
A CON - BUL 96 72% 32 64
A CON S A BUL 8 6% - 8
A CON - SMY 8 6% - 8
A CON - ANK 7 5% - 7
A CON H 7 5% - 7
A CON - RUM 6 4% 3 3
A CON - SEV 1 0% - 1
Absolute 133 - 35 98
Percentages - 100% 26% 74%

A 75% failure rate is not very good. But maybe weíre being unfair here. Maybe some of the options that donít bring in a second Balkan center have other merits.

  • A CON S A BUL sounds useful in case the Balkan is wilder than expected. However, A BUL was attacked 0 out of 8 times (0%). Four times Austrian A SER and Russian F RUM were present but did not attack A BUL, and the other four times only one of them was present, so the support order was unnecessary. Supporting A BUL turns out to be useless.

  • A CON - SMY, back to where we came from. Why? In 7 out of 8 cases Turkey wanted to build F CON in Winter. Thatís an excellent reason for keeping CON empty, and kudos to these seven players for seeing it. Still, if they saw the truth why did they send A SMY - CON in the first place?

  • A CON - ANK. Sounds like trouble with Russia — and that, in turn, means that the army should never have been sent to CON. Five out of seven cases are related to Russian incursions, the sixth was to help destroy the Russian southern fleet with the Tsarís permission, and I have no idea why the seventh Sultan chose to go to ANK. Iím afraid the Sultan himself didnít, either.

  • A CON H was the favourite passtime of seven Sultans. The most likely explanation is that they ordered a nice parade to impress their latest concubine.

  • Finally, seven attempts were made to convoy the army elsewhere. Thatís the best thing you can do when not going to BUL, and three convoys were even succesful.

All orders that do not attempt to move the army to a supply center are either pointless, or show that the army should not have been sent to CON in the first place. Thus it is proven that the sole reason for A SMY - CON in Spring is to conquer a second Balkan center in Fall.


The risk of A SMY - CON

How risky is moving to CON in Spring? Worst case scenario: bounce in BUL while the Tsar invades ARM. The Spring move to CON did not yield any results and invited an incursion. How often does this happen?

Spring 1901: A SMY - CON
Russian incursions in ARM
Fall 1901 move Absolute Percentage Russian incursion in ARM
Spring 1901 Fall 1901 Spring 1902
A CON - BUL
bounce
6 out of 64 9% - 2 4
A CON - ANK 5 out of 7 71% 5 - -
A CON - BUL
succeed
3 out of 32 9% - 1 2
A CON - RUM
bounce
2 out of 3 66% - - 2
A CON S A BUL 1 out of 8 13% - 1 -
A CON - SMY 0 out of 8 0% - - -
A CON H 0 out of 7 0% - - -
A CON - RUM
succeed
0 out of 3 0% - - -
A CON - SEV 0 out of 1 0% - - -
Absolute 17 out of 133 - 5 4 8
Percentages - 13% 4% 3% 6%

In about 1 out of 8 cases, a Sultan moving A SMY - CON is subjected to a Russian invasion in ARM in Spring 1901, Fall 1901, or Spring 1902. Thatís not too huge a risk to run.

  • In five cases the Tsar invaded immediately in Spring, and the Sultan was forced to reply with A CON - ANK. These Sultans misjudged their diplomatic position.

  • 9% of Sultans playing A CON - BUL suffer an invasion of ARM — likely because ARM is indefensible and the Sultan, again, misread his diplomatic position. Still, a chance of less than 1 in 10 means the move is pretty safe.

  • It seems that attempting and failing to convoy A CON - RUM is a good indicator of an invasion of ARM in the next turn. Russia and Turkey are seriously at war, and Turkey fails to harm Russia in RUM. Hence the Tsar is in a mood and a position to extract revenge. (Still, there are only three failed attacks on RUM, so we may see a statistical artifact here. The theory sure sounds nice, though.)

While I had expected to find that many Sultans moving to CON would be invaded in ARM, the data does not bear that out. Moving A SMY - CON is relatively safe — probably because itís combined with F ANK - BLA, which almost forces the Tsar to play F SEV - BLA (bounce) and thus keeps ARM safe from incursions.

Moving A SMY - CON may be fairly safe, but that doesnít make it useful. Or succesful.


A CON - BUL

How can a Sultan know if A SMY - CON in Spring is a good idea? Which factors make a grab for a second Balkan center succesful?

Reasons for success of
Fall 1901: A CON - BUL / RUM
Reason Absolute Percentage BUL RUM
A/T vs R 8 23% 8 -
R-G or R-E war 7 20% 5 2
A-R war 5 14% 5 -
A-I war 5 14% 5 -
NMR 3 8% 3 -
Austro-Russian support 2 6% 2 -
Russian tactical errors 2 6% 1 1
A/T vs I 1 3% 1 -
I/T vs A 1 3% 1 -
R/T vs A 1 3% 1 -
Total 35 100% 32 3

  • In about half the cases Austria or Russia or both are distracted by wars elsewhere. They may fight each other, or Austria may become embroiled with Italy, or Russia may opt for a northern strategy. All these occurrences lead to a profoundly happy Sultan and Balkan pickup opportunities.

  • In about a quarter of the cases Turkey concluded an alliance with Austria against Russia. This, too, is a clear advantage for a Sultan seeking happiness and the picking up of a second Balkan center.

  • The rest is a mixed bag of other alliances, NMRs, and stupidity, though I would like to take a moment to praise the diplomatic brilliance of the two Sultans who were supported into RUM by Austrian and Russia.


Summary so far

So weíve seen the following:

  • Out of four Sultans, two send A SMY to CON, one to ARM, and one does something else — usually a pointless move to ANK.

  • Out of eight Sultans who move A SMY - CON, one is invaded in ARM over the course of the first three turns.

  • Out of four Sultans who move A SMY - CON, two subsequently order A CON - BUL but bounce, one orders the same and succeeds, and one does something that proves the army should never have been sent to CON.

  • Out of four Sultans who conquer a second Balkan center, two succeed because Austria and/or Russia fight elsewhere, one thanks to Austrian support against Russia, and one for other reasons.

So the A SMY - CON move is a fairly safe one, but itís only useful in certain specific diplomatic circumstances: Austria or Russia fighting elsewhere, or an Austrian alliance against Russia.

We can go even further: A SMY - CON only makes sense if the Sultan knows about these diplomatic circumstances in Spring. If he doesnít heís just guessing when he sends off A SMY - CON.


The useless army

Why do so many Sultans resort to guesswork? I think they donít know what to do with A SMY. Letís face it: if you donít want to go to ARM straight away, A SMY is a rather useless army. Who hasnít wished for F SMY instead?

Still, Sultans think they are required to make it do something because all units should do something in Spring 1901. Thatís mostly true, but not entirely. A VEN H is fully accepted as part of Italian openings.

I believe that the same should go for A SMY H. I believe that, if it cannot play a well-defined role in ARM or the Balkan, A SMY should be a garrison unit, aimed at defense rather than offense.

A SMY has one use that A CON does not have. It can guard your frontiers against treacherous Tsars. By threatening A SMY - ARM in Fall you have a lot more leverage in your negotiations. Even better, you havenít actually declared war on him — you just hover there menacingly.

Turkey is strong in defense. If youíre unsure of your alliances, unclear on what the rest of the board is going to do, and generally groping around, you should consider keeping A SMY right were it is in order to strengthen your defense even more.

Using a unit defensively is not ideal in the early game, but these are the sub-optimal cards Calhamer has dealt the Sultan, so we should play them as well as we can. If 75% of the attempts to deploy into the Balkan fail, itís time to look for other options.


The useful center

Thereís a second fact thatís overlooked too often: CON is Turkeyís prime building center.

Winter 1901. The Sultan has one build, and that means a fleet. But where to build it? SMY and ANK are so dreadfully final: you proclaim to the world at large that youíre either going after Italy or Austria, or after Russia. Sometimes you want that, but at other times you want to hedge your bets.

Solution: build F CON. This offers you maximum flexibility and leverage in negotiating with both sides — unless a by now 100% useless army sits in CON for the Sultanís concubines to admire.

The other use for CON is as a destination for F BLA in case you strike a deal with the Tsar between the Spring and Fall moves. If it were occupied by a 75% useless army youíd have to either play it safe and send your fleet back to ANK (a tactically ridiculous move if there ever was one), or move A CON back to SMY (proving A SMY - CON should never have been ordered in the first place).

I believe that, because they play A SMY - CON too often, Sultans are generally not making enough effort to keep CON open in Fall and Winter.


A SMYís role

Concluding, we can see that A SMY has three distinct roles it can play:

  1. Attack the Tsar. Tactically, this is the best option. Diplomatically, it may not be wise.

  2. Grab a second Balkan center. This works only if you have a good Austrian alliance, or if Austria and/or Russia are waging war elsewhere. And you have to be sure of all that in Spring.

  3. Defend against Russia while keeping CON open for a build.

Right now the second option is overrated at the expense of the third. I would argue that, in the absence of A SMY - ARM, A SMY H should be the default order, to be used unless the diplomatic circumstances I described apply in Spring.

A SMY - ANK?

There remains one more point to argue: A SMY - ANK. It offers the same benefits as A SMY H: it covers ARM and keeps CON empty.

I believe A SMY H is superior to A SMY - ANK in case Russia plays F SEV - ARM — a stupid move, but that doesnít stop some Tsars. With an army in ANK you must hold, but with an army in SMY you can try to set up a bounce that leaves ANK free for a build.

Still, that scenario is an edge case, and my real reason for preferring A SMY H is psychological. If A SMY is a defensive unit, let it behave as one. A SMY - ANK is a non-move, so be consistent and donít move A SMY at all.

By underscoring them with a Spring 1901 HOLD order thatís bound to raise eyebrows, you can make A SMYís strong defensive capabilities clear for all to see — especially the Tsar.


Conclusion

Thus I argue that, in the absence of a move to ARM, the default order for A SMY should become a hold instead of a move to CON. This offers extra security against the Tsar without declaring war on him, and it keeps CON free for a build. Conversely, three out of four times the A SMY - CON move is useless.

If necessary you should not hesitate to send A SMY to CON, but you should first evaluate your diplomatic position for any signs of a second Balkan pickup, and not order A SMY - CON just to give the unit something to do.

A SMY H allows you to negotiate with Russia from a position of strength. Thatís a lot better than ordering a 75% useless army to occupy your prime building center in order to ogle concubines.



Peter-Paul Koch
(pp.koch@gmail.com)

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.