Thursday, July 4, 2019

Be rational

I have been working on balance for policies in Civ 6 and I have concluded that there is one particular outlier that I need to address somehow - Rationalism.


Increase the Science output of your Campus buildings by 50% if the adjacency bonus of the district is at least +3.  Increase it by 50% more if the city has population 10 or more.

There are three other cards that do similar things for the Holy Site, Theatre Square, and Commercial Hub.

I find myself virtually always running Rationalism to boost my science and mostly ignoring the other three cards.  I think this is a bad sign - you want policies to be powerful, but if there is one you auto pick every time then something is amiss. 

The core thing that goes on is that the Rationalism card boosts science more than the Theatre Square card boosts culture, or the Commercial Hub card boosts gold.  There are lots of sources of culture in the game that aren't affected by cards, such as Great Works and Wonders.  Plus if you have a ton of culture going you actually don't need to boost your culture more - you really want to boost your tourism!  If you already have a truckload of high adjacency Theatre Squares your culture is probably already plenty high.  More of it is better than nothing, but your utility drops off quick.  Commercial Hubs mostly give you gold via enabling trade routes and those aren't boosted by the card, so that card is bad.  Also getting the adjacency of +4 to qualify for the Commercial Hub bonus is rough.

The Holy Site card is good.  Nearly all your faith comes from Holy Site buildings, so it is a huge bonus.  It, like science, is hard to get other ways, so the card does something powerful.  But unless you are going for a religious victory there isn't much point in stacking faith, so this card is a really powerful one that only comes up in limited circumstances - I am fine with that.

Because science nearly all comes from Campus buildings the Rationalism card is simply too big a science multiplier.  This is also why Campuses themselves are good - you simply can't get any reasonable amount of science in any other way.  No matter what victory condition you are aiming for it makes sense to have more science - it gives you better units, more buildings, better yields, everything.  And the cap is really high, such that getting ever more science is pretty much always good, in a way that other resources don't quite match.

Unfortunately I don't have a clean way to change this.  Other cards at similar places in the tree do things like adding 2-3 science per Campus, or similarly affecting other districts.  Rationalism can easily be pumping out ~4 science per district at that point, and it can get so much higher as your cities grow and it rockets to the moon when you start pumping out Research Labs in the late game.  I could just hard nerf Rationalism to give 4 science to every university, for example, which would ruin the symmetry of the cards but would certainly fix the necessity of having it.  It would also take away the fun minigame of trying to get all your Campuses to +3 adjacency to get the big Rationalism boost.  I can't decide if I like that or not.  Unless I change all the four cards it will be ... kludgy.

It is certainly true that the AI isn't adept at maximizing Rationalism, and it needs to be to win.  It definitely tries to get big cities and high adjacency bonuses but you have to specifically hunt for +3 and 10 in particular if you want to max returns, and you have to slot Rationalism in the first place, and they will only do so randomly.

Initially Rationalism didn't require high adjacency or high population.  This led players to slot Rationalism, then just slam down Campuses in tiny cities absolutely everywhere without worrying about adjacency.  I think what the designers were trying to do when they changed to this latest version is make players care about adjacency and to grow bigger cities.  That worked.  Unfortunately it didn't fix the fundamental problem that the card is just so good you have to build your entire game around it nearly every time.

Friday, June 28, 2019

A solo mission to Mars

I just bought Terraforming Mars and have been playing the solo game some.  For those that haven't tried it, here is how it works:  A couple of random cities and greenies get planted on the board, your TR starts at 14, and you have to fully terraform by generation 14.  Assuming you win, then you try to have the highest possible score.

There are some interesting consequences to this ruleset.  You don't get milestones or awards, which reduces your access to fast points.  This means that science tags lose a lot of value because competing for the scientist award isn't a thing.  Also things that have raw points on them are a lot weaker, both because they don't get you to finishing terraforming (and that is actually a LOT of work solo) but also because the game is so long that you really want all your investment to go into income, not points.

The first time I tried I just failed.  The second time I got a score of 124, which is absurdly high.  Everything just went perfectly.  The next two times I got 90, then 82.  I think you would struggle to win the game at all with any score below 70, so 82 is pretty close to the bottom of the barrel for games where you win.

One thing I have been finding about strategy is that power is weak.  Cities don't do anything to terraform, they just give you points, so collecting power in order to drop cities is only good if you already have amazing income going.  You don't need to plant cities early because there is no competition, so you really only need to slam them down on the final game turn.  The other major use of power is Steelworks and the other similar cards that turn power into oxygen - they seem good, but if you actually want to get a huge score I think they are actually weak.  You want to drop 14 greeneries to rack up points, so setting up an oxygen engine actually doesn't do much of anything - in an ideal, high scoring world you get all 14 oxygen just from massive plant production.  If your game is going badly though they can certainly be a path to winning with a low score - better than losing, no doubt.

The most clutch card I have found so far is Aquifer Pumping.  It is 18 to play, and lets you spend 8 per turn to place an ocean.  You absolutely need to get a ton of plant and heat production, but there are lots of cards that do that.  Only Aquifer Pumping really sorts a category completely on its own, and otherwise you are going to have to hack together an ocean strategy from all kinds of ocean cards and standard projects.

If I am going to beat my top score I think the cards I would want in my opening hand include Aquifer Pumping + a path to setting up huge heat production and plant production.  Massive early income cards like Strip Mine or Mohole Area seem like exactly what you want.  There are people who have worked out ways to terraform completely by turn 2 and get 500+ points assuming you can perfectly stack the deck the way you want, but in a normal game where I don't know what is coming these cards are the ones I want to see.

Anyone else that has good ideas for how to think about cards or maximize TM solo score should feel free to chime in - I want to learn to be better at it!

Tuesday, June 25, 2019

Canada the great

In the latest Civ 6 patch Canada got a buff.  All of their Mine, Camp, and Lumbermill improvements in Tundra give one more production or food than before.  This, in addition to their giant pile of other Tundra based benefits, means that Tundra is still slightly worse than other terrain types for Canada.  But only slightly!  I just can't look at a huge list of abilities obviously designed to make a civ good in one particular circumstance that flat out fails and not do anything about it. 

I went with a simple and powerful change - Canada gets +1 food on all Tundra tiles.  This turns Tundra into a place that is actually excellent for Canada to be in.  It is probably only worth about 3-4 extra stuff per city for cities sited in Tundra vs. Grassland, but this is a solid benefit.  It means that you are really quite happy to start in Tundra locations and you will preferentially expand there.  It isn't an overpowered ability though because it takes awhile to actually realize the full benefits, and you probably only get them in a couple of cities on average.

The key is that the feel is right.  The numbers actually make you want to be in Tundra and you feel rewarded for doing that, which is the way it should feel.  The feel of constant raids from people living in the snow near your cities is a little odd though... I don't remember the last time barbarians came rushing down from the frozen north to pillage Toronto, to be honest.

One thing that seems super hard to balance is Canada's special war bonus, where they cannot declare surprise wars and the AI cannot declare surprise wars on them.  I just finished a Canada game where Genghis Khan was sitting right next door and he had his usual gigantic army of cavalry and was constantly grumbling at me.  Normally this would mean a series of wars, but he just sat there making more and more cavalry and doing nothing with them.  In the midgame I became allied with him because of an emergency and discovered that he had 125 chariots on the map!  That wasn't his whole army by any means, as he definitely had over 200 units, but his chariots covered the continent and the nearby oceans wall to wall.  I can't help but think that his AI told him to declare a surprise war on me but he could not so instead of denouncing me and declaring 5 turns later he just sat there foolishly wishing he could surprise me.

If my supposition is correct it would make Canada's ability extraordinarily powerful in games vs. the AI.  A human can just denounce and wait, but an AI who gets stuck can't do anything about it, and it probably wrecks their whole game plan.  Given that the only real threat to an expert player is early AI rushes, this could even make Canada by far the most powerful civ, if only because it interacts so oddly with the AI's war decisions.

Of course the game went predictably, and when I was in the Information Era Genghis decided to attack a city state who was my friend.  I declared war on him and parked a modern army on his borders and watched in glee as wave after wave of chariots, horsemen, curassiers, and knights slammed into my units and died.  It was hilarious to see a single unit kill more than 10 enemy units in one turn as they rushed in and committed suicide, hoping their wooden carts with spikes would defeat modern infantry.  I killed around 200 units and pillaged half a dozen cities to nothing before letting Genghis have peace.

One other change that went into the game that is interesting is that the AI clearly puts less priority on Holy Sites than before.  It used to be that you couldn't get a religion on high difficulties without doing some pretty awful stuff to get there, and even then you would often just fail to get a religion at all.  Even AIs who had no intention of religious victory would rush Holy Sites.  This seems to have gone away, and I have been able to get a religion easily when I push for it.  Some civs rush it, most others ignore it in favour of other things.  I like this a lot because I think that it makes higher difficulties a lot more enjoyable when you can access all the feature of the game.  The best possible strategy is still exactly the same (Build an army, get 'em) but you can much more easily play a peaceful game and actually get a religion if you enjoy that part of things.

Tuesday, June 18, 2019

Small steps in the right direction

The new Civ 6 patch is out, and it turns out we mostly had all the information ahead of time.  The patch consists of a lot of good changes, mostly focusing on improving mid to late game buildings that were underused, as well as making the whole diplomacy victory and world council better.  I agree with basically all of their directions and the things they are trying to achieve.  Unfortunately I don't think they went far enough in many of their changes.

Specialists are better now that the top end buildings buff all specialists within their district.  They basically go from making 2 stuff per turn to 3 stuff per turn once you have the top end buildings constructed.  However, this is not nearly enough to make specialists good.  Throughout the game a citizen will normally be working a tile generating 4 stuff or more, and that increases as the game goes on.  Specialists in the early game are nearly unusable because you simply can't feed your people if you use them, midgame their returns are a joke compared to the 5 stuff you get on good tiles, and late game specialists get buffed all the way to 3 stuff while tiles continue upward to 6-7 stuff.

Now you might want to use specialists in a city that you have nothing left to build in and which you do not wish to grow.  So for example in the late game where you are trying for a science victory and are basically on cruise control you might use specialists.  But that is far too niche to my mind, especially when it is so easy to make them a reasonable option in other situations.

In any case this puts the structure in place for me to easily buff specialists, making them improve as you add infrastructure.  That much is welcome, certainly.

Another set of buffs that was well intentioned but a little too light handed was the improvements to Industrial Zones.  Workshops and Factories have been improved, but they still aren't good enough to make Industrial Zones worth building.  It is a step in the right direction though, and certainly makes the power imbalance between the various districts less stark. 

I think the designers are really certain that big cities with all the districts are good.  In that sort of scenario, a district that generates production in order to make more districts is a fine thing.  However, the ideal way to play is to simply spam small cities everywhere and chop all the forests down to get up your science or culture infrastructure.  After that, ignore the city as it sits there pumping out the resource you need to win the game.  Big cities need so much support in terms of housing, amenities, and food that it just isn't worth it, and if you have small cities with only 2-3 districts then one of those districts cannot be a district that just makes more production.  If you increase that production though, and make it easier to get the amenities and housing needed to grow a city big, then suddenly huge cities become a real contender.

We aren't quite there yet.  But I am happy to continue to nudge the numbers to try to make it happen.

Friday, June 7, 2019

Full remembrance

I only just noticed that WOW is finally launching a classic server so people will be able to play the game as it was at launch.  The community has been asking for this for years, and I figured it would never happen.  I was wrong, it would seem.

But I don't know if the people who ask for a classic server actually want the thing they are going to get.  They want the feeling of exploration, discovery, and newness.  You can't get that back just by redoing stuff you did years ago.  They want that sense of losing yourself for hours and hours, time slipping away.  Newflash:  You are old now, and your kids and job make playing for 12 hours at a stretch impossible.

There are some things you will get though.  The game was far more grindy back then, and when you accomplished things you knew just how long it had taken.  Want to get one more tick of Blacksmithing?  Well, get out there and mine for six hours.  Want to gain a level?  Well, there aren't any quests to do, so I guess you grind pirates in a circle for a day or so.  There are some people that want this, but it isn't a thing that I am looking forward to.

I completely understand some of the things they want though.  You won't be teleporting or flying about, so the world will feel real and huge.  Getting places takes time and you have to interact with all of the denizens of the world and I really like that.  There will be zones with elite enemies that you have to find a group to defeat, and this is a challenge I remember fondly.

There is also all kinds of legitimately terrible stuff in classic WOW.  Dungeons like Maraudon had a 14 level range - the dorks at the entrance were level 39, but the boss was 53.  If you go in when the first monsters are reasonable, the later half of the dungeon is unwinnable.  If you go in at a level where you can fight the boss, the first half of the dungeon gives you zero experience, despite still being dangerous and requiring tons of time to slog through.  Sound fun?

I remember levelling up in classic.  Druids could turn into a cat to do damage... but that was actually worse than just hitting the enemies with a staff.  Paladin levelling was just autoattacking until enemies died, then casting one healing spell before going to the next enemy.  When you finally get to raiding warlocks were a class built to debuff enemies, and bosses had only 8 slots for debuffs, and you were expected to bring 5 warlocks and to have 7 mandatory debuffs already taken.  Have fun!

I am sure there will be a flood of people playing classic WOW at the outset.  I will probably even play for a bit.  Nostalgia!

But when I realize that you have to have a warrior tank for 5 man dungeons, and warriors are only 1 class out of 8, I will be sad.  When I note that the best healing paladins will just spec into the retribution tree because the talents in the healing tree don't do anything, I will shake my head.  When I log on for an hour and want to do something, but realize that getting a group and travelling to the dungeon will take 30 minutes, I will grumble and log off.

There are some really good things about classic WOW that could be fun to play with again.  But you can't just bring it back like it was without facing the fact that Blizzard changed WOW greatly over the years, for good reason!

My guess is that there will be tons of interest at the start, then it will fall off rapidly.  Blizzard may well do fine financially just because of all the resubs, but I think the long term playerbase for the classic game simply isn't going to be there.  I do intend to go back to Redridge and kill those murlocs on the north shore of the lake though.  Just to show them who is boss.  /Neverforget

Wednesday, June 5, 2019

New tricks for me

A new patch is coming for Civ 6 and I am excited about many of the announcements.  Two things they are trying to address in particular are the excessive cost of endgame buildings compared to their effects, and the power level of specialists.  They are heading in the right direction with building costs for sure, lowering the costs of the top buildings, but they haven't done nearly enough.  For example, the Research Lab is getting its cost lowered by about 25%, which is definitely good, but it is still heinously inefficient compared to the lower tier science buildings.  It is still well within the 'build it if you are just messing around before victory, or if you have truly nothing decent to do' territory, and I want more than that.  Unlocking endgame buildings should feel powerful, but the Research Lab is less than 25% as efficient as the University which comes before it, and that is pathetic.

However, I will give them credit that they are trying to change, they are simply doing it tiny steps at a time instead of getting to the right spot all at once.  Better than nothing.  At any rate I will continue tinkering with my mod using their new production costs and altering the outputs of the endgame buildings to work with them.

One other thing you will note on the Research Lab is that it now buffs Citizen yields.  This is a nice improvement because specialists were almost totally worthless in the old version.  They gave 2 of their thing, so Campus specialists gave 2 science, for example.  This is instead of working a tile though, so it was terrible.  Tiles are almost always vastly better than this, usually giving 4 stuff even in the early game, so getting to slot in a specialist to get 2 stuff is garbage.

We don't have all the details yet, but it looks like they intend to have specialists remain at 2 stuff, but get a buff from these buildings so they would produce 3 stuff at endgame.  Of course at endgame tiles produce ~6 stuff, so specialists are still basically trash, only good if you deliberately don't want a city to grow, or if your city is so big you can't even work any more tiles.

The bright side is that now modders will be able to easily drop these bonuses to specialists onto other buildings!  Previously I had simply altered specialists to give 5 stuff instead of 2, which meant that they were quite good in the early game (you still can't use them that much though, because otherwise your cities won't be able to make enough food) but they were fairly mediocre in the endgame.  However, with this new tech I will be able to start off specialists at 4 stuff, and upgrade that through the tech tree, just the way that tiles get better as you advance.  It will mean that specialists can be a solid choice throughout the game, something you can do but don't have to do.

Firaxis is also attempting to buff Canada.  Pre patch I think Canada is actually worse than a blank civ, because their tundra start bias is a serious disadvantage and they have a collection of terrible bonuses.  However, they are getting a slight buff to the point where I think they are better than nothing at all.

I decided that 'barely better than nothing at all' isn't good enough!  I am going to add +1 food to all Tundra tiles to the Canadian arsenal, which means they actually really want to be on Tundra and can build excellent cities in that environment.  Their cities elsewhere will be totally vanilla, but at least Tundra will actually be good for them.  Want to play a diplomatic game where you can build great cities in territory no one else wants?  Play Canada, they do that well!  This is my hope, at least.

So all told this patch is a bunch of good ideas in the right direction, just not taken quite far enough.  Could be worse.

Monday, June 3, 2019

Start your engines

This past week I played a couple of Terraforming Mars 1v1 games.  I have concluded that TM is best in that format, in large part because of the way the hate cards work.  In 4p games hate cards aren't usually good, but when they get played they feel terrible because it seems so utterly random.  There is an awful lot of power in convincing people that you are the player in last place just so they don't hit you, and I hate that dynamic.  I also find that it sucks to want to avoid production - getting a steel production feels weak because I often feel like it is just going to get stolen.

I also dislike the way that plant production just makes you a target for meteors in 4p.  When anyone plays a meteor they just look about for someone to punish, and you don't want to be that someone, so plant production feels actively pointless.

In 2p though these dynamics are totally different.  For one, you see much less of the deck so the hate cards are far less likely to exist - you can get a steel production without being convinced it will get stolen at some point.  Also the negative points that are on many hate cards remove a lot of the sting - getting stolen from still sucks, but at least your only competitor lost a point!  The worst part about hate cards in 4p games is that you often watch the person in last place lose a point just to punish you even though you aren't winning.  The lower number of cards also means that it is far less likely that meteors will destroy all of your early plants, and if they do get destroyed at least your only opponent had to pay a lot to accomplish that.

The hate cards in 2p are actually really good, and that is okay by me.  I don't mind that they are super efficient, because at least I don't have to feel like someone is torching their own game just to hurt me like I do in 4p.

I also like that the income cards are a lot better.  In many 4p games income cards are terribly weak if they aren't played by turn 2 because the game may end on turn 8, and it feels silly to be ignoring income when the game is only just starting up.  In 2p the game goes long (my last 2 games were 12 and 13 turns) so you can comfortably play income cards until turn 7-8, which lets you really have fun building a powerful engine.

In fact those income cards are so good they can turn around games where you lose the critical milestones.  In my last game I was the plant corporation, so I took it for granted that I would win the Gardener milestone for having 3 forests.  Unfortunately for me Naked Man started with Inventrix and quickly grabbed the 16 card milestone, then blew up 5 of my plants cheaply, then played Arctic Algae to get 11 plants.  He bought a forest to get the Gardener milestone ahead of me, then bought a city to get the city milestone.  With those points on board and 3 milestones he was *20* ahead.  I thought for certain I had lost.

But I had been doing things.  I had my heat production up to 11, plants to 4, and energy to 3.  I grabbed Physics Complex, and began pounding out temperature increases, Physics points, Herbivores, Tardigrades, and more forests.  He had poured everything into getting those milestones, so he had basically no production at all.  The 20 point lead he had just melted in the face of my 6 / turn for 5 turns, and I won by about 12 points.  In the 4p game if you get no milestones it is disastrous because it is so hard to have enough income to do anything as powerful as getting 5 points for 8 bucks.  But in 2p where the game goes long, you can easily win doing it the way I did, by having a monstrous income that just keeps paying off. 

Milestones are excellent, obviously, but you have to be careful that you don't throw away too much in pursuit of them, and I like where that balance lies in 2p a lot more than in 4p.  In the game previous to the one I just described I managed to get all 3 milestones, 2 awards, and 3 Jovian payoff cards and won by 36, so sometimes you do just get to do it all.  In any case 2p TM feels like a fantastic game, and I wish I could try playing it competitively a bunch to try to get better at it.  Sort of like Carcassone, the game feels super interesting and deep when 2p, and like you are really at the mercy of random player choices in 4p.