The best kittens, technology, and video games blog in the world.

Thursday, January 31, 2013

Europa Universalis 3 Divine Wind second impressions

NYC - Brooklyn - Prospect Park Zoo: Discovery Trail - Red Panda by wallyg from flickr (CC-NC-ND)

Over a month ago I wrote first impressions review of EU3 Divine Wind. Since then I've played it a lot more, and did a tiny bit of modding, so here's an update.

Modding is easy!

First, modding is ridiculously easy. Pretty much everything is a just a bunch of text files.

To get started you just need to make a simple mod skeleton, copy over defines.txt and any other files you want to tweak, and just edit some data.

New values you're trying don't necessarily have to result in a "better" game, it being different from one campaign to the next can be very interesting.

Some things like giving everybody a lot more diplomats, or making basic buildings not get destroyed when province changes ownership are fairly easy to do. Unfortunately "fixing" infamy or providing faster mission cancelation option for particularly stupid missions seems like a lot of work, if it's even fully possible.

The game is full of bugs

So far I had two different save game corruption bugs - and since save games are plain text files (something very rarely seen in any game ever) I managed to fix them both manually. The first was achievement exponential growth making save games double in size each time, the second was some kind of string escaping problem that added a lot of crap in the middle of save file - and interestingly moving back to last good save the problem would still reappear in a month or two.

In addition to these serious bugs there are countless things in the game that don't work like expected, or in a sane way - for instance just a few game-years ago when my vassal was sieging a province held by rebels and I sent some spies to bribe defenders away, my spies were detected and of course that screw my relationships with my vassals.

There's even more things which might or might not be bugs, but in any case work in completely undocumented ways - I still have no idea why my relationships with other countries go down (usually) for no apparent reason at random times. It's amazing since the game openly provides such huge piles of data and algorithms, and still manages to surprise every single time.

Even when things work as designed, it's really hard to know what's the design unless you alt-tab to a browser with EU3 wiki. The tutorials and in-game explanations are so bad as if they were nonexistent, but after a few games you figure things out on your own.

There is no balance

Some games make a serious attempt at balancing things - I doubt they even bothered with EU3. There are so many broken things you can do, like become Holy Roman Emperor, establish Baltic or Black Sea tolls, seize Rome/Jerusalem/Mecca, establish land route to Africa or India and so on.

These things tend to be game-breakingly good, but then if the prize wasn't great what would be the point putting major effort into such goals?

Exploits

There are quite a few fun mild exploits I recommend trying.

Start multiple wars at once. Sadly the casus belli system makes it unprofitable to get anything out of countries that joined existing wars, but there's an easy workaround - save 5 diplomats, then stars 5 wars on the same day with each member of the alliance! A massive 5-way Crusade against Ottomans/Saruhan/Candar/Kardamon/Ramazan can net you 10+ provinces and up to 5 annexations at 1 infamy each.

War goals against leader sometimes work against alliance members. If you declare war against a country and it loses war leader status, you have no casus belli against anybody except that country. But if other countries join as minor parties your war goals often still apply! So declaring a Holy War against Muslim country, you still get 1 infamy/province even for taking provinces from their Christian allies. It's quite useful against Bosnia and Serbia early game. It's not as good as multiple war technique, since you'll rarely be able to afford taking more than 4-5 provinces in peace talks per war and you won't be able to annex anyone.

Bait ships in port with 1-ship blockade. AI is somewhat competent at land warfare when it has compact country like France or Burgundy, since it's pretty much impossible to screw that up. Naval warfare on the other hand - it's just amazingly bad at. If you want to sink some ships that happen to be in port, just move your entire fleet one province away and leave one ship blockading their port. AI will invariably sail out of the port to attack that ship, then your fleet returns to sink them.

Baiting fleets before blockades are invented. Similar technique works even before blockades - if your land army is crossing a narrow strait, AI can't resist trying to stop it, even if that crossing is fairly irrelevant to its goals. Of course your superior fleet is just one province away, waiting for them.

Naval patrols. That's not as exploitative, but useful when you have massive naval superiority, but don't want to micromanage. Just set your whole navy to go in circles (set route with some shift-clicks, select start patrol) - for example around eastern Mediterranean when you're fighting assorted Muslim minors. AI will see that your ships sailed away and will leave port even when it has no reason to, then it won't get back in time when your patrol reappears.

Land route to Africa or India. This is a surprising game mechanic, but can be extremely powerful. Provinces are counted as "overseas" when they're on another continent (somehow "continent" boundaries are not quite geographic), or when there is unbroken land route of actual control (not ownership) from your capital to the province. This makes is extremely powerful for European power to control land route to Africa or India, since then they count for far more money and manpower. You lose some special abilities of "overseas" provinces like colonialism casus belli, ability to culture shift with one colonist etc., but since what matters is actual control not ownership you could temporarily make these provinces "overseas" again by letting rebels take one of key provinces on your land route.

Breaking alliances. Do you want to attack your ally's ally? First, attack a third country and get your ally into your war on your side. Then attack your target. Since they're in existing war on your side, they can't take the victim's side in the new war, so their alliance will break. It's pretty damn useful if you get alliance with the Holy Roman Emperor, get them into a silly war on your side, then you're free to attack anybody in the empire at flimsiest CB and they won't intervene.

In other words - it's an awesome game to play, and an awesome game to mod!

Wednesday, January 30, 2013

Medieval 2 Total War Concentrated Vanilla 0.61

Jumping after each other on mom! by Tambako the Jaguar from flickr (CC-ND)

Since 0.60 release I fixed a lot of minor bugs, and added a bunch of minor enhancements to the mod builder (big thanks to everyone who tested and reported the bugs, especially Guimbix).

There aren't any huge features, but enough minor fixes accumulated that I feel it's good time for an official release.

One feature that got added is File > Generate Map TGA command you can run after building the mod to see how campaign map looks without launching the game, and trying a different random seed if you don't like this one.  The map file is saved to output/political_map.tga

And one more thing - if you're using old versions of retail Medieval 2, make sure it's updated to 1.3 or the mod won't work. (you're also saving yourself a ton of bugs this way)

Fixed bugs and minor improvement

  • start.bat pauses with "Press any key to continue" after running the script, so you can read error message if it fails, usually due to JRuby not being installed properly.
  • Campaign map heigh map level 0 changed to 1, fixing many campaign map glitches
  • Belgrade's resources (timber vs wood and missing comma) fixed
  • Floods disabled completely due to being buggy in vanilla
  • Simplified building tree option works now
  • Clustering% fixed to mean that (it accidentally meant random%, reverse of what it was supposed to)
  • Mod builder now extracts map information from TGAs at run time, instead of using preprocessed data. (so it's easier to port it to alternative maps, even if it's not yet there)
  • When allocating the last few settlements, clustering is often impossible. Before the update algorithm tried to look at neighbours-of-neighbours, then just picked a random settlements. Now it goes quite deep searching for not too distant settlement.
  • There's a switch for "allocating rebels last" if you need even more clustering than you get by default. This often causes rebels to be pushed to borders of the map, so only use this if really necessary.
  • Various help texts and error messages improved.

Credits and links

Concentrated Vanilla 0.61 includes some files from Medieval 2 Total War game by Creative Assembly, and the following minimods:
  • Sinuhet's Battle Mechanics
  • Lusted's Better BAI/CAI
  • Agart's Cities/Castles Strat
Download link.

If you're interested in how this mod evolved, posts about previous versions (with links inside):

Tuesday, January 22, 2013

Gatecrash Sealed Simulator

Rude? by Tjflex2 from flickr (CC-NC-ND)

Continuing the grand tradition I made Sealed Simulator for Gatecrash, which you can use to practice before the weekend Prerelease.

Here's Gatecrash Sealed Simulator.

Or if you want to play with older sets:
The simulator works in a way identical to RTR simulator. Strangely we had 4 single-set Limited formats in a row - I don't think that ever happened before, or is all that likely to happen anytime soon.

The one for Dragon's Maze will need to change since it will be a three-set format. And then presumably two more single-set formats.

Any feedback welcome.

Good luck on the Prerelease and may the best Crocodile Frog win!