Agricola Strategy tips

This advice is written with 4-player games in mind for players that are quite experienced and want to become better. Maybe I at some point will write a guide for 2-player. It is also written playing the original decks EIK often with some of the expansions Wm, Fr,G,G4,G5,G6 in real life or on play-agricola.com. Much of this advice will remain true for revised version, but there should be some modifications. Also my examples are taken from the original decks.

This list should be regarded as a work in progress, written by Bernt Ivar (aka Turambar). I believe most strong players will agree with most of this, but of course there might be some disagreements. It is probably a bit biased by my own playstyle.

This is an updated version of the article. Thanks to play-agricola.com players cs2376jr,Graey,rookie and sidnv for helpful comments on the first version.

General advice

  • The most important rule: If you have a good reason to not follow any of the advice below, then do so. The better one is at the game the more often one should do this, because then you know enough to actually justify breaking the rule. Thus none of the statements below should be read as absolute, but rather as helpful guidelines that most times will make your game better.
  • Do not take bad actions. This is surprisingly good advice, since many people will way too often take bad actions. What is bad is relative to board position and round: 2w is a good action 2nd turn in round 1, but a bad action in round 6. 1s is almost always a bad action, however if you just need one stone to renovate a big house it might be a good action (although then you should ask yourself whether you could have planned ahead better). What is a bad action is also relative to what you could have gotten: Taking 2c in r1 to later play a 1c minor (say simple fireplace or flower pots) is decent enough, but probably you can take 4c in r3 or r4, and still play that minor in time, thus 2c in r1 is in this case often a bad action.
  • Early actions are worth more than late actions. Plowing a field is 2 pts and is not considered a particularly strong action in the early game.
  • A baseline good action in the late game is 2 points. Plow is 2 pts. Taking an animal/crop you do not have is 2 points. Taking 3w is worth about 2 points in fences. This is useful to compare whether doing a sequence of actions is worth it: If you do not get an average of 2 points per action you might reconsider. At the end of the game it is more common that there are no 2 point action sequences left.
  • Another part to the above advice is set yourself up so you do not have to take bad actions. If you see that there are few good actions for next time it is your turn, maybe you should grab the resource pile that lets you do a good fence/major/build action? Or play an occupation that makes a bad space better for you? Assuming that these are actually good actions. Or preemptively taking a medium food action if it prevents you from having to take a bad food action in 2 rounds. Very often when people take bad actions it is a consequence of some mistake they did a round or two earlier. Towards the end of the game it is often important to create possibilities to be able to do good actions also in the final rounds.
  • There is often a tradeoff between doing efficient actions and doing tempo actions: Actions that opens up possibilities for other actions. For example building two rooms in one go is in isolation more efficient than building one room two times. However if one gathers resources to build two rooms in one go one will often wait longer to build which means one will most likely family grow later and lose actions. Thus many times it is better to build one room two times rather than two rooms in one go.
  • Try to be efficient in scoring points. This might again seem obvious but is quite hard. For example fencing two times (with no bonuses for fencing) is 0 points and 1 action more than just fencing once towards the end of the game. This means that one has to get about 2 points of benefit from the first fence action for it to be worth it. Sometimes one gets good animal grabs in which case the first fence action was a good tempo fence move. Other times one does not get good animal grabs, in which case the first fence action was essentially wasted. One can win surprisingly many games by mostly just taking efficient actions from the board, even if one has a weak hand.
  • Starting player is in itself a bad action. Almost all new players has to learn that starting player is a worse action than they intuitively think. One takes start player primarily to do something very important as first action next round. There are exceptions to this, where you can play cards like bookshelf and animal pen that are as strong as a full action, but they are surprisingly rare. At the end of the game sp is a good action since going first the last 3 rounds is better than earlier (although also here one should think through whether one actually benefits from taking sp).
  • Think through whether the game is food short/rich in stage 1. If the game is not food rich then 3f is a good action round 3 and 4f is a very good action round 4, even if you already have enough food for the round 4 harvest.. Taking food in round 3 or 4 gives you flexibility to do other things in stage 2. Often there are a lot of important actions to be taken in stage 2, because people want to grow their family. For this reason there will also often be large resource piles. Eventually you will have to find food for the harvest in round 7, so you might as well do it in r3/4 when there are not many other very good actions.
  • What is considered a good food action varies by stage in the game: Most games 4f is really good in stage 1, good in stage 2 and a bit weak, but sometimes necessary, in later stages.
  • In a 4-player game spending an action merely to block one player is rarely correct, because the 2 other players will just benefit. The action has to either be good also for yourself or you have to be very confident that you and the player you are blocking are clearly leading the game. Doing the latter is very hard in the early/mid game, thus one should almost never do straight block moves at that point of the game(like letting the sheep run).
  • Scoring points from fields and crops are in isolation more effective than from pastures and animals, since it is easier to get to max points in crops than in animals.
  • The most important thing about playing a grain/baking game is sowing grain: A very common mistake is baking the grain away before sowing it. Usually this is wrong. Conversely if one manages to sow 2 grain fields then one can usually feed and end with max grain points without too much trouble. Sowing grain fields are in general more important than getting an oven. Many players overrate how important it is to get the oven. If one manages to sow two grain fields one can live completely fine with a cooking hearth, since it gives the possibility to bake grain with a decent conversion as well as the flexibility of taking good animal grabs.
  • The two most basic ways of feeding are eating animals and baking grain. But there is also some other ways: Buying the guilds (pottery and basketmakers workshop and joinery) can give a lot of food as well as giving points (and taking away the potential for points from others). Buying the well or playing minor improvements that give food on future round spaces (like duck pond, fruit tree, house goat) are important sources of food that can save you the trouble of building one single feeding method. Some of the best games uses trickling food from here and there for most of the game where one spends almost no actions taking only food. For example with several rsf actions, buying the well, duck pond, pottery and an occasional fishing/traveling players action one can feed a player for the 3-4 first stages.

Advanced general advice

  • Be aware of who gets good actions around the table: Quite often it is correct to take 3w to block someone from getting 6w (this does not violate the “do not block” advice above, since 3w is almost always a good action). But do not do this blindly: If it is mid/late game you should ask yourself whether the player who gets 6w is competing for the win. If not, then wood policing is not necessary. Also you should think whether you have the actions to spend wood policing: sometimes you do not because you have few actions or because your game is really good just doing your own thing. In both these cases it could be reasonable to not wood police.
  • In general be wary of who you are helping with your start player actions, particularly if you take it often. It does not help your game to get good actions if the player after you get as good actions without taking sp. It is quite easy to accidentally just give the game to the player sitting after you.
  • Stay flexible so you can jump on chances: if someone lets you jump in growth queue or take 6w, be ready to do so. One way to do this is to get rid of your obligations: People will often try to put off an action they will have to do sometimes soon, like building a fireplace/room or taking the sheep, as long as possible. The intuition is that then they can get tempo to do something useful done first, or get a more efficient sheep action if one waits another round. Sometimes this reasoning is correct, however sometimes the opportunity cost is higher than the gain: by waiting one more round to take the sheep you are creating an obligation for yourself to take the sheep that round, which means you might be unable to jump on a good opportunity that opens (like 6w).
  • Sometimes it is good to decide you are not doing something because your hand is not set up for it and never even try. This frees your attention to achieve things that is possible for you. Trying to do everything at once could mean that everything is done badly. For instance one should not always try to get all public bonuses. Or if one has a specific hand, say building a 6 room mansion with riding plow, playing under the assumption that one is not going to get fences might be good: This frees up your actions to do the things you have to do. At the end of the game you will sometimes be able to squeeze in some fences after all, but this is secondary like stables in an ordinary game.
  • Wood is a continuous resource in the sense that having 1 extra is almost always good. Clay is a discrete resource in the sense that the most important thing about taking it is the number of actions you spend taking clay. As an example: assume you see that you probably need 9c the entire game: 4c renovation, 4c cooking hearth and 1c to pay hoarder for a bonus point. Then you should aim at taking exactly 2 clay actions that game. This means you should try to wait until you can get 5 or 6 clay in one action and then later you can take 4 or 3 clay whenever you have a free action. It is a mistake to take 4c twice, even though 4c is in general a decent action, because that means you will need a third action. If the amount of clay you need changes (say cheap hearth is bought so you need 1c extra to get expensive hearth or both hearts are gone so you have to make do with fireplace, thus needing 1/2c less) the number of clay actions to aim for should often be the same. Stone is more continuous than clay, although one has to be careful when to stop: it is a common mistake to end with a lot of stone which has no use because all the majors are gone. Reed is generally also discrete like clay, although because of the rsf space it behaves a bit continuously, since one generally should and could optimize the combined number of reed and stone actions.
  • Think through which resources are scarce/plentiful: Some games there is a ton of wood, while others there are almost no wood. Adjust your valuation of actions after this. This is pretty advanced advice, since correctly judging the value of resources is often hard.
  • Always think vaguely about how one is feeding and notice how everyone else is feeding. One does not have to know exactly how one is feeding, but at least have some possible ideas for the next 2 harvests.
  • Do not always try to build the same farm, but rather try to do what scores the most points given your hand. For example, if one has a hand with chief,cloister dweller, manger and wooden strongbox it is completely reasonable to aim for a final farm with 0 fields and 1 pasture (covering 9 spaces).
  • Probably the most important advice for becoming a really strong player: Learn to have flexible plans and not hard commit to a single plan but have several possible plans in mind that you will commit to depending on the state of the game. For instance if one has donkey and sunrise admirer in hand there is probably no need to play both. However one also does not have to choose early which one to do: If you get a lot of early wood donkey might be better, while if you get a lot of early food sunrise admirer might be better. This advice also includes being open to ignore your best cards to exploit the state of the board or to ignore all the advice here because what you are doing is better.

Cards/Drafting

  • Know the format you are playing: Decks and draft number changes card strengths quite a bit.
  • Learn the statistics about which cards are stronger than others, but do not take them too seriously but rather treat the numbers as very vague approximations.
  • Always consider what is best for you actual hand. Plenty of times taking a weaker card that combos with your hand is better than taking a strong card.
  • Try not to draft only occupations that has to be played early: Rarely will one have time to play more than 2/3 occupations in stage 1. Similarly try to not draft only minors with high requirements.
  • Using your draft picks to block cards others need is occasionally fine, but generally it is better to draft for your own game. After all hate-drafting only blocks one of your opponents.
  • Evaluate cards in the draft based on both potential and average. Picking a card because it has potential to be really good is fine, but one should always also try to think of what is the expected return of a card. For instance wood cart is a card that can easily fool you: you imagine yourself playing it r3 and then you flip grow rd5 and you end up taking wood 12 times with your absurd amount of actions. Most times this does not happen. Similarly everyone remembers that time their game got destroyed by keys and they might be tempted to draft it just to avoid passing it to someone who destroys their game. Most games keys does basically nothing. Similarly do not evaluate a card simply on the assumption that one gets occ/sp rd1, quite a few times one does not get a chance to take those actions.
  • There is such a thing as making a too elaborate combo: If you spend 4 actions setting up your day laborer action you will have to take the action enough times that the action is justified , meaning the combined setup and day laborer actions give you on average 2 points. In reality they should give you more than 2 points per move, since you are taking an uncontested action, everyone else will get better actions than usual off the board. Generally combos using only a few cards are better than those using a lot.
  • It is easy to make a too elaborate plan: If a plan requires you to get some very specific actions at very specific moments, or that certain round cards flip in the correct order, then the plan is on average not very good and one should expect it to fail quite often.
  • Having playable minors is more important than occupations, because every game one has to take start player and family growth actions at some point. If one has unplayable minors one ends wasting them. Having unplayable occupations simply means not taking the occupation space, which is a lesser cost.
  • Do not make your cards worse. Hut builder gives a free room. One generally should not draft/play cards that make rooms cheaper if one already is planning to play hut builder, because they will be weaker than usual since they do not discount the hut builder room.
  • Some cards main utility is using them in stage 1. Traveling players cards are almost unplayable if you do not get the first TP action. Field watchman is best used for sowing 2 or 3 grain fields in stage 1. A surprising consequence is that grain cart is not particularly strong with field watchman, since the 2 occ requirement and 2w cost means that one will probably have to delay sowing the grain fields to use field watchman with grain cart. Spending stage 1 setting up grain cart+field watchman is usually weaker than spending stage 1 using field watchman to sow 2 or 3 grain fields. But once you have done that the extra grain from grain cart does not do that much.
  • When drafting minors: Always consider taking the cards with (free) points: If one decides between a card with and without a point , one has to think whether the effect of the card without the point will make up for the point(s) of the other card.
  • Think about point ceiling in the draft (how much depends quite a bit on card format). It is rare to win with no points from your cards.
  • It is very easy to over commit to your cards. Always consider whether a card is strong enough in the current situation to be played. For example if someone else plays market crier and gives you 2 grain then you probably should not play your own market woman, since the 2 grain you have will many times be enough to last you the game. This advice includes point ceiling cards: A common mistake is one tries too hard to set up the point card in hand, and for that reason avoids taking good point actions on the board. For example avoiding to take the Well because one tries to set up Chief. Quite often one should just take the well and hope one can get to stone+Chief as well. If one does that the card will actually add point ceiling to your game. If you avoid playing a 4 pt major to later be able to play a 4 pt occ it is not clear how much you actually gained.
  • 4w occs should generally only be played if you have a reasonable shot at getting the bonus points. Possible exceptions to this rule are humble farmboy and grump because they reward something that is not generally winning. Taking a 4w occ with intention of not playing it is fine if there is nothing particularly strong.
  • Notice important cards that you pass: cards that give global bonuses or warps the game. For example if you passed head of the family you know that family growth will probably be easier than most games, which changes how you play stage 1 and 2.
  • Do not play a card unless you are reasonably certain you are going to use it.

Learn the dynamics of the family growth queue:

  • Growing your family is very important. Growing early is better than growing late.
  • Building the first room is almost always good. Generally you would like to build it as late as possible unless grow already flipped: If you see that you will certainly be able to build first room in r5 you should probably not build in r4. On the other hand, be wary of cards that let people jump ahead and build first in r5.
  • If you have the first room and grow does not flip in round 5 you should think through whether you wish to take start player to get first grow. This depends on several factors, the most important ones being having good minors to be comfortable taking sp first move r6 and grow r7 as well as having acceptable food at the same time. If the person sitting after you also can grow, you should consider start playering more than if it is the player before you.
  • If one is double building it is, if possible, best to do so in the same round one wishes to grow since then one can grow also the round after without taking start player.
  • Start playering in round 5 to build the third room in round 6 is only worth it if grow flipped in r5. If not you might as well just wait for your turn. That way you also do not commit the following mistake:
  • Generally do not start player in stage 2 if you are not currently in the growth queue: It forces the players who are growing first (and thus leading) to at least take bad actions (sp) for grow.

Early game

  • Last actions r1 and r2 are quite special: generally the board will only have actions like 2w,2c plow. If you have played a card that makes some available action better you should often take it last action r1 and/or r2. A very common mistake people make is playing hill farmer, dancer or seasonal worker in r1 and then not take the corresponding action in r1 (instead they take 2w/2c). Or take 2c r1 and then not play fireplace r2 even if major is available.This is almost always a mistake: If you take fireplace r2 over 2w you freed an action for some later round where you can take more than 2 wood instead of major.
  • Do not make many different ways of feeding, particularly in stage 1 and 2, because you spend to many of your few actions on taking food. This means if you are playing a traveling players card or fish trap you should rarely go for the first sheep grab. Later in the game if you need much food it is more reasonable to get it from many different sources.
  • Do not start player twice in stage 1. You waste too many quality actions. There are surprisingly few exceptions to this rule: Even if one has something like axe+ladder it is probably best to play one of them in stage 1 and the other in stage 2 for growth and use it for rooms 4 and 5.
  • Do not create too many good last actions for yourself: If you play several cards that says that for you fishing or traveling or plow or similar is good, then you can easily have too many things to do with your last action and end up not using some of them.
  • Getting the first fireplace/hearth is often good if you have no other food plan. It is even good if sheep comes late, because that means you have easy feeding in stage 2 in a low food game.

Mid-game

  • Mid-game is probably the hardest part of the game. Here you need to decide whether to build another room to grow again or to set up fields and crops or pastures. Experience gives one intuition for these decisions.
  • Taking solid actions like 4w,3w, plow is rarely terrible.
  • If grow flips early then building a 4th room is often worth it. If grow flips late it is often not worth it.

Late-game

  • Getting rid of ones commitments is important: Having to spend first move in r14 fencing is not good, and often one can avoid it by planning ahead.
  • Learn the dynamics of end game renovate and fence queues: A player in stone rooms cannot take reno+fence, so the player will need to use the actual fence action. Who takes start player r13 will often determine who ends up renovating.
  • Think a bit about who your opponents realistically are: If your best opponent also have a constable farm, then playing constable is a 0 point move for you, so you should do something else.