Writer: terracubist

Background

This article is partly written as a result of the lessons that have been happening recently in the #coaching channel of the Dominion Discord. Come join us on Thursdays, 7pm EST!

I only started playing Dominion competitively about a year ago, and only recently hit a thousand rated games on the client. One of the disconnects I often see between the best players in the world and those trying to improve is over intuition—what is instinctive (as the result of thousands of games) is often hard to articulate. One of my takeaways from #coaching lessons and learning the game relatively recently is that, when teaching newer players, it helps a lot to try and explain from more fundamental concepts rather than assume similarly honed instincts. For that reason, I thought it’d be useful to share what I’ve personally found success with: a procedural approach to kingdom analysis that attempts to minimize instinctive play by giving clear goals for each stage of the game (opening, midgame, endgame) before the game even starts. 

30 second elevator pitch

The overarching idea of this approach is the following: think backwards from the end, asking all of the following during turn 1:

1. Endgame: How do I score?
2. Midgame: What do I want my eventual deck to do, given how I score?
3. Opening: How do I build to that eventual deck?
4. Is there anything that could disturb my plan?

I like to think of a game like building a pyramid: scoring (endgame) relies on payload (midgame) which relies on deck control (opening). 

Figure 1: Author is bad at art

Endgame: “How do I score?”

Why start thinking from the end? I found it a lot easier to build with a goal in mind, and what better goal than nice shiny points? Also, I think it’s fairly common that how you score affects how you build (e.g. “I see Distant Lands as a way to score, which meshes nicely with a draw-to-X deck based around Cursed Village”). 

There are two parts of this stage of the process:

• Look at all the possible ways to score
  • Ex: “on this board, I see Provinces, Witch, Distant Lands, and Tomb as ways to score”
  • Is there alt VP?
    • Where alt VP = literally any other scoring outside of Prov/Duchy/Estate
    • Is that alt VP actually usable?
      • “Oh hm there’s no trashing for tomb, and terminal space for Distant Lands looks iffy”
• Is any pile liable to run?
  • Basically, “how likely is a pileout to occur, given it can shortcut the normal scoring process?”
  • Requires experience, but look for junkers, self-gainers, power cards
  • Ex: “I think with Witch curses are probably going to empty, and with Magpies that’s probably a second pile. Lackeys are cheap and incredibly powerful, so it’s possible we’ll buy a bunch of those and that pile runs.”

Note that each of the above can have a strong or weak effect: Colonies are very strong alt VP, while Pasture is relatively weak. Similarly, Groom heavily encourages piles to run, while Ironworks is usually a step slower than Groom. 

Midgame: “What do I want my eventual deck to do, given how I score?”

Once you’ve identified how you can score, it’s time to think about how to build a deck that can score in that manner. The steps for this stage are the following:

• If alt VP available, figure out how good it is to pursue it
  • Ex: “ok, Witch slows my opponent down, and completely ignoring it means -10VP, which seems like a lot given the only other way to score is provinces”
  • Ex 2: “lol triumph and horses means provinces are for scrubs”
  • If it is good to pursue, compare options against Provinces/other alt VP. 
    • Look at the support for each, and as a result try and guess how frequently you can buy/gain/play each option. As an example, spamming Monument may be viable due to Hunting Lodge and good thinning, but uncontested Horn of Plenty access will likely decimate the Province pile before monolithic Monument playing can score enough.
    • If the options aren’t mutually exclusive, consider pursuing both/all. In the example from before, maybe plan to have Monument as one of your 8 uniques, and take more Monuments as the situation presents itself. 
• If alt VP not dominant, what’s the best way to get Provinces?
  • Is there a viable +buy or way to gain Provinces?
    • If “no”, look for fast money strategies
      • Ex: “gear + trade go brrr”
    • If “yes”, what support does that deck want?
      • Ex: “I see horn of plenty and a few decent cantrips, so I want to get thin and add a little draw”
  • Basically, decide if building to single/double/megaturn
• If piles look likely to run, plan to have at least one of the following:
  • VP lead – Ex: a money deck getting to a decent lead before piles get too low due to building
  • More pile pressure than opponent – want more relative gains to dictate when the game ends

Essentially, you’re trying to figure out what payload you want to use to go about scoring/the endgame, and what the optimal deck for playing that payload is. 

Opening: “How do I build to that eventual deck?”

The opening (turns 1 through ~6) is arguably the most important stage of the game, as whatever decisions you make here have a compounding interest-esque effect: the cards you open are likely going to be your most played (and therefore most impactful) cards. As such, the goals for this stage are the following: 

• Set up deck to play payload as often as possible
  • Look for trashing
  • Look for draw/sifting
  • Look for reliability – e.g. topdecking, Gear, villagers, etc.
• Get payload cards into deck soon
  • Figure out if you want to hit certain price points, and when (ASAP? whenever?)
  • Figure out if taking early payload/gainers helps get more payload/deck control (Ex: is playing cursed gold for Sentry worth it here?)

An example of thinking through this stage might look like the following:

“Okay so I want to get lots of Distant Lands and play them via Cursed Village. That deck wants to be clean of starting junk and doesn’t like Silvers, so I’m going to open Fisherman/Cargo Ship with the goal of hitting $5 early for a Sentry, and cleaning up while maintaining some midgame-friendly econ.”

This is obviously dependent on working through the endgame and midgame questions, so make sure you’re confident in your work there before jumping straight into clicking on cards. 

“Is there anything that could disturb my plan?”

After working through the stages of the game, I like to ask the following questions:

• “Will any attacks/opponent interactions affect my plan?”
  • Ex: Scoring via Duchy/Duke might struggle against a Bureaucrat-spamming deck playing a lot of Haunted Woods
• “Do any of the power combos/cards break/augment my plan?”
  • Check for things like CH+TF, Lurker + HG, Groom rushes, etc. which break a lot of the assumptions this approach relies on (i.e. scoring and the “endgame” starts almost immediately)
  • Check for strong synergies or tricks
    • Ex: Villa + Cursed Village, Way of the Mole + Village Green
  • Look at the canonically “good” cards and see if they fit into or warp your plan. Silk Merchant, for example, offers draw, +buy, pricepoint flexibility, temporary terminal space, and trash for benefit food. Some part of that will usually shortcut the usual concept of “get control, then build payload, then score”, so make sure not to ignore it. 

Answering these questions does rely a fair amount on game experience or study; even plenty of good players don’t know about Storeroom + Stampede, and memorizing a list of the best cards in Dominion isn’t always fun. As much as I’ve tried to minimize this approach’s reliance on external knowledge, I haven’t really found a way around this yet. 

Putting it all into practice

Kingdom 1 – 67897727

Using the approach I’ve described above, my analysis for the board above goes something like the following:

• Endgame/Scoring: On this board, there is scoring via provinces, tomb, monument, and curses from desperation and cursed gold. Monument looks hard to use given limited terminal space, but tomb looks very good with trashing options in watchtower and pooka. Watchtower looks even better given it can trash the curses from desperation and cursed gold for “free” econ, and you can score a lot of points with Masterpiece overpays and trashing all the incoming silvers. Scoring with Provinces mostly has the problem of being a dead card in deck, and only scoring 6VP at a time, whereas Masterpiece overpay into Watchtower trash converts $8 into 6VP, $9 into 7VP, etc. Provinces will probably come into play later once Silvers run, but until then Masterpiece overpays are more efficient. Without much in the way of gains or terminal space, it looks like only silvers and masterpiece could run, while curses almost definitely will. 
• Midgame/payload: Our “payload” includes anything that scores us tomb points, so we want watchtower and pooka, and scepter looks nice to play pooka again. Most of the other cards compete with pooka for terminal space, or have a better option at their price point.
• Opening/deck control: Watchtower asap to get access to cursed gold and desperation (and free VP) looks like a must. Then Pooka to start killing coppers for vp, draw, and thinning looks good. Hitting $5 for Pooka seems pretty easy once we have desperation and cursed gold available. Given that we’re aiming to consistently find watchtower and lining up pooka + scepter is nice, we don’t want to take a curse without Watchtower in hand. In terms of reliability, Haven might or might not be useful to set aside excess Watchtowers, and Watchtower itself gives topdecking which might be situationally useful.
• Disturbances: None of the boards really jump out at me as monolithic power cards, and I don’t see much in the way of attacks or player interaction here. Winning the split/pile denial doesn’t seem like much of a factor here either. 

Homework

If you’ve found this approach useful and would like to try it out, the following are the other kingdoms we used in our #coaching lessons. You can check the speaker notes here for some hints (https://docs.google.com/presentation/d/1Avd02y7_BTq6gt-mB0eUiiN8HcIBwJhvrfNSMDCeiSk/edit?usp=sharing).

• Kingdom 2 – 67586246
  
• Kingdom 3 – 63616668
  
• Kingdom 4 – 62608723
  
• Kingdom 5 – 67402462
  
• Kingdom 6 – 62066481
  
• Kingdom 7 – 57956965
  

Disclaimers

• This is what’s worked for me. Take what works for you, and adapt it as needed. This isn’t a strict checklist, merely a template.
• Freaky’s ideas of next shuffle optimization are still applicable while you’re playing the game. Everything I’m talking about is ideally done during turn 1, before you’ve bought anything. 
• This approach is more concerned with “how much do I build here”, not “how do I build”. I started doing it after I already had some idea of how to build an engine, so if you don’t think you’re at that stage go check out JNails’ excellent video series: https://www.youtube.com/playlist?list=PLsrQ73qPvJ0sIvvD0gERXQt1nOS0iIK_4 
• “All models are wrong, but some are useful” 
  • As with all general advice in Dominion, there’ll be edge cases that this approach won’t work for. The premise this approach relies on (that there are different stages of the game that are broadly deck control→payload→score) falls apart on boards with rushes, Cavalry/Villa silliness, and some of the more monolithic combos. 
• This approach attempts to reduce the amount of experience and raw intuition needed, but is much faster as you get more.
  • It’s a lot easier to evaluate the relative strength of things (e.g. double goatherd or chapel? Is village smithy with no thinning reliable here?) when you have direct experience. 
  • Also knowing power combos (Gear + Trade, CH + TF, etc.) and synergies can help shortcut a lot of the thinking.
• My ideas were loosely inspired by Dan Brooks’ forum post here: http://forum.dominionstrategy.com/index.php?topic=16194.0
• This approach works best in formats that are not time-controlled (e.g. the Dominion League) and is sometimes a little slow for the “click buttons win prizes” mood that seems to predominate the automatch queue.