Milan Balance (or Does Italy want to go West?)

Strategy discussions for the Milan variant.
Forum rules
trategy
In addition to the general Forum Guidelines (see here: viewtopic.php?f=130&t=15441), there are additional rules for posting in this forum.
1. Members should not seek advice about an active game they are in.
2. It follows that links, images, game name and/or number should not be added to a post if the game is active.
Posts which refer to a specific situation in an active game, or which link directly to an active game, are subject to editing or removal.

Milan Balance (or Does Italy want to go West?)

Postby luckythirteen » 29 Dec 2015, 21:20

In the 2015 Fall Retreat edition of the Diplomatic Pouch, the 2015 Diplomacy World Champion Toby Harris wrote that he likes the Milan Variant so much that he "hope(s) to see this as the standard map at f-5-f Dipcon!" He goes on to say "It opens up strategy and Diplomacy for all Great Powers and starts the game in a more balanced fashion for all seven players." I've never played the Milan variant before (when researching it I became sidetracked by the 1900 variant), but with high praise like that from such an accomplished player, I figured it was worth investigating. Unfortunately there just isn't a ton of material out there for the variant. Perhaps it is because it appears to be so similar to the base game people don't feel it needs much discussion? Another possibility is that it simply isn't that popular. On this site we've only had 43 games that I'd consider to be "Classic Milan" (non-friendly games using the Milan map and no other special rules like Fog of War or Build Anywhere). I'm very interested in trying out Milan and thought I'd try to get some discussion going about the variant since there isn't really much out there right now.

The best article that I was able to find about Milan was by Charles Roburn in the Sprint 2007 Retreat edition of the Diplomatic Pouch. In this article, Charles explains how the addition of a buffer space between Italy's home centers, and Austria's makes it easier for the Pope to launch a westward offensive. Towards the end of the article he has some fantastic analysis. Of particular note, Charles points out that the "total moves" required for the fastest route to victory actually INCREASES for Italy in the Milan variant, theoretically resulting in an even more difficult path to victory. He points out that this may be balanced by an increase in Diplomatic flexibility, but that the increased difficulty in getting to 18 SCs might offset this Diplomatic advantage. At the time of his article, he had a sample size of 33 games to collect data on to see just how balanced the variant was. Ultimately, he concluded 33 games was too small a sample size to generalize from, but he did notice a few trends. I thought it would be interesting to compare his sample with the data from the 43 completed "Classic Milan" games on this website, and then combine the data to see if there is anything new that we can conclude about the balance of the variant.

AUSTRIA
DPjudge and DipWorld Results (33 games): 3 solos, 1 two-way draw, 3 three-way draws, 2 four-way draws, and 24 "losses" (a surviving player who didn't participate in the draw, or an elimination).
PlayDip Results (43 games): 4 Solos, 4 two-way draws, 6 three-way draws, 0 four-way draws, and 28 "losses".
Total Results For Austria (76 games): 7 Solos, 5 two-way draws, 9 three-way draws, 2 four-way draws, and 53 "losses".

ENGLAND
DPjudge and DipWorld Results (33 games): 5 solos, 1 two-way draw, 3 three-way draws, 5 four-way draws, and 19 "losses" (a surviving player who didn't participate in the draw, or an elimination).
PlayDip Results (43 games): 2 Solos, 5 two-way draws, 7 three-way draws, 1 four-way draw, and 28 "losses".
Total Results (76 games): 7 Solos, 6 two-way draws, 10 three-way draws, 6 four-way draws, and 47 "losses".

FRANCE
DPjudge and DipWorld Results (33 games): 1 solo, 1 two-way draw, 2 three-way draws, 3 four-way draws, and 26 "losses" (a surviving player who didn't participate in the draw, or an elimination).
PlayDip Results (43 games): 3 Solos, 3 two-way draws, 3 three-way draws, 1 four-way draw, and 33 "losses".
Total Results (76 games): 4 Solos, 4 two-way draws, 5 three-way draws, 4 four-way draws, and 59 "losses".

GERMANY
DPjudge and DipWorld Results (33 games): 2 solo, 1 two-way draw, 3 three-way draws, 3 four-way draws, and 24 "losses" (a surviving player who didn't participate in the draw, or an elimination).
PlayDip Results (43 games): 2 Solos, 5 two-way draws, 5 three-way draws, 0 four-way draws, and 31 "losses".
Total Results (76 games): 4 Solos, 6 two-way draws, 8 three-way draws, 3 four-way draws, and 55 "losses".

ITALY
DPjudge and DipWorld Results (33 games): 1 solo, 2 two-way draw, 4 three-way draws, 5 four-way draws, and 21 "losses" (a surviving player who didn't participate in the draw, or an elimination).
PlayDip Results (43 games): 1 Solos, 2 two-way draws, 5 three-way draws, 1 four-way draw, and 34 "losses".
Total Results (76 games): 2 Solos, 4 two-way draws, 9 three-way draws, 5 four-way draws, and 55 "losses".

RUSSIA
DPjudge and DipWorld Results (33 games): 1 solo, 0 two-way draws, 1 three-way draw, 2 four-way draws, and 29 "losses" (a surviving player who didn't participate in the draw, or an elimination).
PlayDip Results (43 games): 5 Solos, 2 two-way draws, 2 three-way draws, 0 four-way draws, and 34 "losses".
Total Results (76 games): 6 Solos, 2 two-way draws, 3 three-way draws, 2 four-way draws, and 63 "losses".

TURKEY
DPjudge and DipWorld Results (33 games): 5 solo, 0 two-way draws, 5 three-way draws, 4 four-way draws, and 19 "losses" (a surviving player who didn't participate in the draw, or an elimination).
PlayDip Results (43 games): 2 Solos, 3 two-way draws, 5 three-way draws, 1 four-way draws, and 32 "losses".
Total Results (76 games): 7 Solos, 3 two-way draws, 10 three-way draws, 5 four-way draws, and 51 "losses".

ANALYSIS

First of all, I still believe the sample size is too small to take anything meaningful from the data. When comparing "our" data with the data Charles had access to, there are some pretty significant differences. This tells me that there are still too many variations in the data to draw definitive conclusions from the results. Second, there are some results that lead me to question the validity of some of these games. For example, we had far more 2-way draw results than Charles did, and far fewer 4-way draws. One of the positive elements that I've seen people mention about the Milan variant is that it is more difficult to build a stalemate line because an army in Savoy can cut the defensive support in Munich, and an Army in Burgundy can launch or support and attack on Savoy. Out of the 43 "Classic Milan" games on this site, 12 of them ended in a 2-way draw. I have a hard time believing that 28% of *any* Diplomacy variant should end in a 2-way draw (I would expect more solos or 3-way draws as the norm). Does the Milan variant really make 2-way draws that much easier when compared to "Standard" Diplomacy or are people "settling" for 2-way draws on this site instead of playing for solos? That being said, there are still some trends in the results that are interesting. Additionally, I have compared the data I collected above with some statistics gathered by Josh Burton about Standard Diplomacy to use as a control to compare how balanced Milan is when compared with Standard Diplomacy.

ITALY
First, I think the biggest question is whether or not the Milan variant helps Italy. With only 2 Solos in 76 results, Italy has soloed only 3% of the time. The median solo percentage in the Milan games above is around 7%, so Italy's 3% is significantly lower than the other powers. In the Standard Diplomacy numbers above, Italy soloed around 5% of the time and was only 1% lower than average. This means that so far at least, the Milan variant results in fewer solos for Italy when compared with the Standard Diplomacy statistics. If you compare Solos and Draws, Italy is exactly at the median in Milan, participating in a solo or draw in 28% of its games. Interestingly, this is the same percentage as Standard Diplomacy as well, so the Milan variant doesn't appear to improve Italy's chances at a draw either. That being said, Italy was below the median for Solos/Draws in Standard Diplomacy and is exactly at the median in Milan, so it may be that Italy is getting eliminated less frequently in Milan when compared with Standard Diplomacy. Of course with only 76 games, it's hard to say definitively that Milan is worse for Italy. I think we *can* safely conclude that Milan isn't drastically *improving* the situation for Italy though, with the possible exception that it is now just as likely as most of the other powers to participate in a draw.

AUSTRIA
The other power that is significantly impacted by the change to the map in the Milan variant is Austria. In my opinion, Austria is the biggest winner from the Milan variant. Austria has soloed 7 times (9%) which ties it with England and Turkey for 1st place in number of Solos. In the Standard Diplomacy numbers, Austria had a 6% solo rate, so this represents a 3% increase in solos, which ties it with England and Turkey for the largest improvement in this variant. It also participated in a win or draw 30% of the time, which is 2% more frequent than average, and also a 3% improvement over its Standard Diplomacy results. Austria had good results in both Charles' data and in our PlayDip data. It is very clear to me that the Milan variant helps Austria quite a bit. I am curious as to why Austria has such a disproportional increase in success when compared with Italy though. I'm not surprised to see Austria as the winner (in classic Diplomacy Austria frequently seems to do well if he survives the early game, and this variant helps Austria to survive), but I *am* surprised to see by how much.

FRANCE
France was a powerhouse in Standard Diplomacy, soloing 7% of the time and soloing or drawing a whopping 39% of the time. One of the observations among Milan proponents is that France should be weakened by the variant, and this is reflected in the statistics. In our completed Milan games, France has soloed only 5% of the time which is 2% lower than average. The biggest drop is its participation in draws. It has only Soloed or Drawn in 22% of games, 6% lower than the average and 17% lower than in Standard Diplomacy. This is the second worst result for Solos/Draws and in the bottom 3 for Solos. My speculation is that with time, we'll see the French numbers climb up a bit, but it's obvious that they have been weakened by the variant.

RUSSIA
Russia was another powerhouse in Standard Diplomacy, with more Solos than any other power at 8%. Russia had an average number of draws in Standard Diplomacy at 32%. In Milan, Russia still has a high solo rate at 8%, but his draw % drops massively to the very bottom at 17%, a full 9% less than the average of 28%. Additionally, Russia has done very well on our site (soloing in 12% of games), but did horribly in Charles' numbers. I suspect this means that Russia continues to be a "go big or go home" sort of power that is dangerous if it manages to survive, but can also be ganged up on and eliminated pretty frequently as well. It is also proof that we need to take all of these numbers with a grain of salt, we just don't have enough completed games to see things averaged out enough yet. A couple of anomalies can throw off our statistics.

GERMANY
Germany was pretty much average in the Standard Diplomacy numbers, soloing exactly at the median 6% and soloing or drawing just slightly below average at 30%. In Milan not much is changed. Germany solos 5% of the time (1% less than the average) and is soloing or drawing exactly at the average 28%. I've seen a few games where Italy is opening north vs. Germany now that they don't have to worry as much about their border with Austria. I suspect if anything, Germany is slightly weakened by this, but the variant doesn't really seem to have had a major impact on this power's success one way or the other.

TURKEY
Turkey is another winner in this variant so far. However, I'd take this with a grain of salt because I've noticed that Turkey does better on our PlayDip site than it does in the statistics I listed above. I'm not sure if this is because my numbers for Standard Diplomacy are from 2006 and people have changed strategies over time, or if there is something a little more "meta-gameish" going on where the players on this site are using different strategies than in the Judge and DipWorld communities. In the Standard Diplomacy numbers I linked to earlier, Turkey had an adverage solo rate at 6% and a slightly above average (2% increase) Solo/Draw rate at 34%. This makes sense because Turkey can be a very difficult power to eliminate but can also get bottled up easily. In Milan, our numbers show a 9% solo rate, and a 33% draw rate. This means that Turkey is soloing more frequently in Milan and drawing about as frequently. I'm honestly not sure how to explain the increase in solos for Turkey when you consider Austria *also* has an increase in solos. Perhaps with Italy moving west more often, the A/T vs. R alliance may be more common? Maybe when Italy heads west, whoever survives between Austria and Turkey can stab Italy in the back?

ENGLAND
England is the other big winner from this Variant. Where France was the dominant power in the west in Standard Diplomacy, England has clearly taken up that mantle in Milan. In Standard Diplomacy England soloed the average 6% of the time and had a solo/draw result slightly above average at 35%. In Milan, England solos 9% of the time and has a solo/draw result a whopping 38% of the time, a full 10% higher than average. England already had an easier than normal chance of forcing its way into a draw, and considering that France and Russia are both weaker in this variant, it's no suprise that England is more powerful as a result. I wonder if Italy heads west vs. France if England is able to take advantage of this scenario more easily than Italy, bottling Italy up in MAO and allying with his cross board partner in Turkey/Austria to stab Italy in the back? Or perhaps A/T are able to ally more easily now, meaning England has an easier time vs. Russia? In either scenario, it's clear that England is the new power in the west in the Milan variant.

CONCLUSION

We don't really have enough data to make any definitive conclusions yet, but based on the data above, I don't think anyone can argue that the Milan variant is more balanced than Standard Diplomacy. However, one also cannot argue that Milan is definitively *less* balanced than Standard Diplomacy either. In both maps Italy seems to get the "short end of the stick". Besides this you basically have just exchanged an above average France and Russia with an above average England, Turkey, and possibly Austria. I haven't finished my analysis of Italy's opening strategies yet, but it is very clear to me that Italy has a lot more options. I like options. I've seen many anti-French openings, (weaker) anti-Austrian openings, and even a few anti-German openings. I would assume the Lepanto "anti-Turkish/delay" opening is still legitimate as well. I question whether or not part of the reason that we have seen such large success for England and poor results for Italy is that Italian players are just excited to have different options and are trying out new ideas. Following along this same mindset, are the French suffering more than they should because the Italians are so excited that they can move west now and still trying things out? Once the Italians get bored of opening "west" and truly explore what their "best" options are based on the Diplomatic situation, will we see I/F ally more frequently in the opening phases of the game? Perhaps I should do more analysis on Gunboat games to see how people are opening when Diplomacy isn't a factor?

I think the real question I'd like to discuss is whether or not Italy's "western" opening is actually a good idea or not. Does an early game attack on France just result in England and/or Turkey getting the upper hand in the mid-game? Is it still better for Italy to focus its early game on Austria in the Milan variant? *If* the ability for Italy to open against France is just as valid as the Austrian attack, regardless of whether or not the variant improves Italy's chances at a Solo, I would argue the Milan variant is "better than classic" simply because it gives Italy more opportunities. However, the data listed above makes me question if attacking France early is any better for Italy in the mid-game than it is in classic Diplomacy. Clearly it's better in the early game, but I'm wondering if this is a short sighted strategy helping the "witches" (England/Turkey) more than Italy once you reach the mid-game. If this is the case, all the Milan variant is really doing is helping out Austria, and the statistics we've seen so far would support that theory.

So, finally the heart of this post. Does Italy really want to head west in the Milan variant? What Diplomatic situation should Italy be looking for to make this happen? I would normally assume you'd want Russia heading north to slow England, but this might leave Russia vulnerable to the now more powerful A/T. Perhaps Italy actually wants the Juggernaut (R/T) so that Russia can open North and Turkey can distract Austria, and then Italy can join with R or T in the mid-game? Or is it simply not possible for Italy to get 18 without the Balkans, so it's still better to attack Austria first. In this case, does the Milan variant make things safer for Italy in the western border with France? If not, does the addition of the added opening game strategy make things better for Italy than classic Diplomacy, or is it just making things more difficult?

GUNBOAT

Edit: I ran the numbers for the 22 completed "Gunboat" (no press) games. This is a really insignificant amount of data so it's hard to draw conclusions. The good news is that each power has at least one solo. The bad news is that England has soloed 4 times (18%) and Turkey 5 times (23%) and England wins or draws 36% of the time, and Turkey 45% of the time! everyone else is almost exactly the same having a 5% solo rate (1 game each) and an 18% solo or draw rate. Austria is the lone exception with one additional solo (9%) but also only a 14% solo or draw rate. Thus, the (admittedly small) Gunboat data supports the theory that the "Witches" (England/Turkey) gain more from the Milan variant than Italy does.
We are what we repeatedly do. Excellence, then, is not an act, but a habit.

--Aristotle
User avatar
luckythirteen
 
Posts: 137
Joined: 25 Oct 2010, 21:30
Class: Ambassador
Standard rating: (956)
All-game rating: (1007)
Timezone: GMT-6

Re: Milan Balance (or Does Italy want to go West?)

Postby Jegpeg » 30 Dec 2015, 03:03

The best game I have played on site was Milan, not sure whether it helps Italy and Austria but it made an opening possible that you don't see in standard (at least I never have).

Game 48020, I was Austrian and died in 1927 , 1 year before Germany soloed.

The opening was France got 3 builds in 01 then turned on Italy while England and Germany allied against Russia leaving me to fight Turkey.
A mistimed stab by England then resulted in a gR alliance against England leading him to die in 1908 along with Italy who could survive against the extra units of France.
The remaining 5 all survived to 1921 when France's long slow demise came to an end at which point I was 2nd on 9 SCs to Germany on 12.
At this point all three ganged up on me and I failed to persuade Russia and Turkey that to continue would mean they could not prevent a German solo.

Through the game at times I was allied with all my opponents except England and it had a huge number of twists and stabs. So much better than the usual where anyone you fight against can never be an allay later in the game.
Jegpeg
 
Posts: 1265
Joined: 08 Dec 2009, 20:56
Location: Scotland
Class: Star Ambassador
Standard rating: (1125)
All-game rating: (1401)
Timezone: GMT


Return to Milan

Who is online

Users browsing this forum: No registered users and 1 guest