Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[philosophyguy]

Lately I've seen a number of posts about engines that were successfully constructed on boards where I would never have thought an engine could succeed. (Examples: Geronimoo's DS post on an engine in the First Game setup and Fabian's sample game analysis) What I'd like to do is revisit, at a higher level, what engine builders look for when determining if an engine is viable.

Obviously, there are the components of an engine:

• +Actions (including obvious sources like Villages and odd sources like Throne Room/cantrip or Golem/cantrip/cantrip)
• +Cards (although not all are created equal; Embassy is definitely a BM drawer because it filters and only gives net +2 cards)
• +Buy

There are also engine enablers:

• Attacks
• Trashing
• Cards that smooth out your shuffle luck (e.g., Scheme, durations)
• TR/KC
• (Potentially) alt-VP that can extend the game length (as long as there's no rush strategy and your engine can sustain Garden/Duke dancing)
• Gains (Ironworks, Mint, and Possession are all different examples from this class)

So, for the advanced engine builders: are there other elements that you look for when deciding to build an engine? When can a really strong board in one area overcome a deficiency in another area?

Some of the answers to the above question, collected from the posts below:

• Whether the engine needs to get to double Province/Colony or whether it can get to single Province faster than a BM deck.
• Whether the engine builds itself faster as it gets going. An accelerating engine building phase results in a more explosive engine; conversely, an engine that just adds a Conspirator each turn is going to need to slow down the BM player much more.
• Whether the engine allows you to control the endgame. An engine with tons of +Buy, cost reduction, or megaturns (like Horn of Plenty) makes it feasible for the engine player to dictate when the game ends.
• Whether there are cards that do more than one function. Develop on a good board serves as both trashing and gaining; it's not great at either, but the combo function compensates if you can take advantage of it (as Marin does here).
• Whether the attacks are engine-friendly. Engines can handle hand-size reduction much better than BM; sans trashing/filtering, cursing is more devastating for engines than BM.

NB: Some engines work as simple stacks: Minion and Hunting Party-terminal Silver are the canonical examples of this genre. Because that engine is built around a single card, you often don't need to evaluate the engine components in the same way as you would in a fit-all-the-parts-together engine type. For instance, Hunting Party gives you +Actions and +Cards, as well as virtual trashing (by skipping repeated cards), but adding a +Buy is generally not worth it unless the +Buy is your terminal Silver (Woodcutter, Nomad Camp) or the +Buy is a cantrip (like Market), lest you break the HP chain.

[Geronimoo]

By playing with the simulator and watching the solo challenge masters it's become clear to me that engines can be extremely powerful when played well. And that's the whole problem: most random kingdoms will contain an engine with the potential to beat the big money deck that's also available, but you'll make lots of errors trying to build it unless you've played a similar setup before. The Big Money deck however pretty much plays itself.

Lately I've been favoring engine over Big Money if there's clear potential. I often fail (especially playing against the better engine builders like Marin), but the game is a lot more fun and I hope I learn something.

There are still times when you don't want to engine and often the reason is Mountebank. Strong BM enablers like Jack or Embassy which are much less vulnerable to stalling can still dominate Province games. And then there are the boards with Conspirator, Bridge or Highway which are begging to go engine but never take off because there's not enough support (often lack of +draw). Going for an engine in those cases is just FPS (fancy play syndrome) which some players are clearly suffering from (hello, Marin ).

King's Court is the only card that will let me look very hard for an engine, any engine really, because its effect is  insane (Throne Room is much much less powerful and I consider it more of a support card).

And finally if there's an engine in a Colony game, go for it! You have much more time to set it up so even if you don't get it quite right, it will still be a much better option then the Big Money deck where you have absolutely no control over the game.

[philosophyguy]

Geronimoo, how do you iterate through the simulator options to determine the best way to build an engine on a board? Like, could you actually walk through the options you tried for the first game engine (beyond just the intermediate steps that you showed)? What options did you try for numbers of Markets, when to purchase the Remodel, etc.?

I love reading your writeups of the simulator process, but I get overwhelmed by the number of options to manipulate and don't know how to strategically think about deriving a strategy and testing it.

[Lekkit]

If I'm to go for an engine there has to be one of three things going for me. Either the engine have to explode and build itself faster and faster. Something like WW+Peddler (and of course probably some other support as well), the cards I'm building my engine. Margrave is the prime example here as it gives both buy and hits my opponent. The last thing that can enable my engine building if the build up isn't too fast is alternatetive VPs. Goons is the best here, Monument and Fairgrounds comes in second (in my opinion). That way a lost Province Split won't matter as much.

[DG]

Gains. Generally if you trash down a deck you need to rebuild and restock it quickly and card gaining can be the best way to do that. Minting a platinum is easier than doing something fancy with highways and bridges to buy a platinum. Gaining also allows you to avoid a conventional economy that might congest the engine.

Hand size or effective hand size is very important. With small hands an engine will often draw mismatched cards or just be unable to deliver much benefit in any turn. Choice cards work best with big hands. Duration cards like wharf, caravan, and tactician help you get a big hand to work with immediately. Laboratory type cards deliver you increased hand size but with a slightly bigger risk of having none in hand. Terminal drawing cards have more power but more risk of drawing badly unless supported by fishing villages, though they present a risk of their own since they don't have +cards. Big hand sizes also allow cellar type actions to cycle lots of cards.

Libraries and watchtowers can be an excellent substitute for normal drawing if you can provide enough actions to sustain the engine. Green cards can cut down the effectiveness of this sort of engine though since you often can't get them out of hand before you play the library. It is important to have an idea of how much of your hand size is effective and how much is occupied by green cards and treasures. This is how silver can be a bad card - it eats into your effective hand size.

The next step is to draw not just big hands but the whole deck so you have complete control and get full value from every card every turn. Once you can overdraw the deck you can gain cards during a turn and then draw and use them the same turn and this allows for really fast acceleration.

[Davio]

I think you can just put down +Actions instead of Villages.

You can KC or TR a cantrip for some faux +Actions or even use Golem. And then there's Hamlet. Is it a Village? It can be, but using the term Village may be confusing to some in thinking that only true Villages qualify.

And virtual +Buy is +Gain or Outpost/Possession I suppose?

[chwhite]

Some engines don't even need Village: if your +Card source is non-terminal (say Minion, Scrying Pool, Lab, Hunting Party, Stables, Menagerie) then Villages are not strictly necessary.  To be clear, oftentimes when using these cards as the heart of your engine, you will want villages anyway, since there's usually going to be a terminal you want to reach, or the Village is also your +Buy source, or any number of other reasons.  The Hunting Party+X stack is probably the engine type which wants a Village least often.

[philosophyguy]

Davio's point about +Actions rather than villages is helpful; I'll edit the original to add that. chwhite's point about stackable cards is well-taken, although those engines are generally easier to play and so I didn't have them in mind for this question. I'll add something to the original to clarify how those engines function.

[philosophyguy]

Also, let me bump one of the questions from the original post: when can a really strong board in one area overcome a deficiency in another? For instance, Remodel is pretty weak trashing, but Geronimoo shows how Remodel is able to accelerate the Village/Smithy/Militia/Market engine enough to make it beat BM soundly. Pro-engine builders: how can you tell when a weak card in one of the categories is good enough?

[carstimon]

That question doesn't really seem that answerable.

Some engines just don't need +buy, because you can hit $8 soon and consistently.  Then you never need to go for double provinces or alternate vp.  I think an example would be something like conspirator/village/warehouse/remake. 
Sometimes your +buy is weak.  Then you maybe have to resign yourself to getting two $16/2 buy hands.  The most common thing this happens with is council room: you usually only want to play 1 or maybe 2 a turn.
But some engines absolutely need +buy or some gaining.  If you're buying highways and there's no +buy/ironworks/workshop/(some others), then you're probably doing something wrong: you're basically spending $5 on peddlers.

I think it's always important to try to get a guess on what a reasonable outcome of a turn can be.  If your buy is only going to come from one council room, you don't need 10 golds.  A common mistake:  If you can draw your deck with room to spare and get $16, don't buy the last two grand markets.

[HiveMindEmulator]

Also, let me bump one of the questions from the original post: when can a really strong board in one area overcome a deficiency in another? For instance, Remodel is pretty weak trashing, but Geronimoo shows how Remodel is able to accelerate the Village/Smithy/Militia/Market engine enough to make it beat BM soundly. Pro-engine builders: how can you tell when a weak card in one of the categories is good enough?

The thing about Remodel is that it's not just trashing. It's also gaining. So, while it may be weak in both areas, the fact that it covers both can make it strong. I talk about Remodel specifically in my Remodel article. Similarly, Oracle's draw is weak and it's attack is weak, but combined it's not that bad if you can put both to good use. The thing is that you need to look at tempo as a whole. You gain tempo in several ways: trashing and gaining make your deck faster, attacking makes your opponent's deck slower, and alternate VPs make your BM opponent's deck slower at getting half the points (by adding more available points). If you are really strong in one area (e.g. Torturer attack), then you don't need any of the others. The weaker the cards are in these areas, the more of them you need to cover. And of course it also depends on the actual engine you're building and the base speed of the BM you're up against.

[philosophyguy]

That question doesn't really seem that answerable. […]

You're right in the sense that there is not going to be a general purpose answer in the sense of an algorithm or a rule that is always true (even Coppersmith shines in some situations!). But, I'm hoping that pro engine builders can chime in with examples of when they constructed an engine even though (there was only +gains, not +buy/there was only weak trashing like Remodel/the best card draw was +2/[insert weak element of the board here]). My guess is that most players can figure out how to make an engine in obvious situations like Village/Torturer; the goal in making this thread was to get advice on how to approach engine building with boards that aren't as tailor-made for an engine.

[dondon151]

I wouldn't quite say that "attacks" in general are an engine indicator; engines don't like attacks that hand out Curses (unless your opponent ignores them entirely fsr) and weak attacks aren't going to hurt much. The attacks that really fit well into engines and ones that either punish your opponent's hand size or reduce the value of your opponent's deck. Once the engine starts firing, the opponent is screwed if he doesn't have close to the VP majority.

Alt VP is also a very strong engine indicator, but only if it's not better to rush. So like, Gardens and Silk Road without good support for the rush, sometimes Fairgrounds, usually Nobles, Island, etc. But pretty much never Duke.

EDIT: I also think that in general, engine building is somewhat underrated or underestimated because we see statistics like "BM+X gets to 4 Provinces in 14 turns" and automatically assume that BM+X is dominant - but the game doesn't end when a player gets 4 Provinces; the right engine can feasibly lose a Province split 2-6 in a game with no alt VP and still come out on top.

[RisingJaguar]

Source: http://forum.dominionstrategy.com/index.php?topic=3311.msg69230#msg69230 - Game 5

http://dominion.isotropic.org/gamelog/201207/22/game-20120722-134533-57014d96.html   Marin 3 - 2 Lekkit

I opened develop, he didn't, and yes develop can be a must-buy without a cost-7 card! I really think develop is underrated, it's so nice to open with!
If you got it with estate, you got silver on top (so finally you have the same result that the one who open silver. But you got a develop and he got one more estate!)
If you got it with moneylander (in this game), you take village/witch on top and that is crazy!
And moreover, it was the only way to trash estates and curses.
I really don't understand how develop can be behind chancellor!

Finally I want say that it was funny games and that I have really appreciate to play with Lekkit!

P.S.:In fact I prefer lose because you don't need to put comment! :p

This was one of the best game logs I had ever seen in terms of engine building.  He takes mediocre trashers to a whole new level.  I write a post below the link to show you a glimpse of what he is thinking about.  The key here is that he uses the mediocre trashers to not only defend against witch, but facilitate his own engine to use witch a lot more often.  He also leaves himself options if his shuffle luck isn't the best. 

[RisingJaguar]

I've sat here trying to think what fits best here, because I do think I am pretty good at engine building.  Unfortunately, I don't really have a great answer that seems to fit all.  I do know a things here that have not been mentioned that are to be considered with an engine.

Building an engine (usually) allows you to control the end game.  Often you will have plenty of +buys that will give you a leg up in three piling and duchy-dancing if neccessary.  You have to know if your engine can sustain this. 

Not to classify all cards in the same grouping (ie. +cards, +action, +buy).  Embassy is one that comes to mind as a card that's not designed for engines.  It is amazing if there isn't much trashing on the board.  However, it only nets 2 cards, so trashed decks are not going to be the best place for Embassy.  Another example is walled village, a kingdom with useful cantrips are going to make walled village a lot worse. 

Gainers/+buy are really important. 

However the best advice I can give you is to just... experiment with cards.  Learn how cards interact with each other, and pay attention of the order of getting cards. 

[Young Nick]

This is up there with the best articles I've read here in recent memory (~6 months).

Minor nit-pick: The link to Geronimoo's front page article is no good. Otherwise, it is perfect.

[philosophyguy]

Fixed the links in the original and added bullets summarizing some of the recent ideas from posts.

If anyone has additional games that they think illustrate some of these ideas, I'd be happy to include those links in the OP. Marin's Develop game is just sick.