I like Agricola. It’s fun. After playing, I don’t regret pre-ordering, I’ll play more when my copy arrives, and I may even try to upgrade my set ala David Fair (and others). I’d play it at the next game session (if I had a copy).
But I do not think it is a great game.
Agricola joins a long list of games that I enjoy despite obvious flaws. I’m thinking of Age of Renaissance, 7 Ages (although my mood on that swings around), and the like. Still, I’d play those instead of great designs I don’t enjoy (Diplomacy, abstracts, etc).
I’m not going to describe mechanisms or details. Dale Yu has written more than enough to enlighten. Just to make things worse, I’m assuming you know the basics. To recap – I like Agricola. I’d play it again at the next game session (if I could). It will likely hit the table another 5-10 times at a minimum. Having gotten that out of the way, I’m going to focus on the negatives.
I’ve played several times now. How many? Well, that depends on how you count. I’ve played four games by any definition. Another four ‘to my satisfaction’, but I suspect most readers will only count one or two of those. “To my satisfaction” means (in this case) that everyone agreed on who would win if the game was played out. Most of these were quickly adjudicated between the first and second harvest.
One was called after the opening deal.
There is a whole class of games where the opening setup determines the likely winner. Card games. They have a few other characteristics (at least for good ones): 1) they are short, 2) you play many hands to reduce the luck (or determine the better player). Good players will win more than their ‘fair’ share of games, but won’t win every hand.
Agricola is a single deal card game that takes 90+ minutes to resolve.
I’ve seen arguments that the cards are individually balanced, and I generally agree (with at least one glaring exception). Some are better, but the range isn’t bad. But cards aren’t just assessed individually. Take Bridge. The Space Ace is worth one trick. The King values usually takes a trick. With the ace, a full trick (assuming no trump). Without it, it depends on where the ace was dealt, how many suited cards each hand has, etc. The deuce of clubs may be a full trick in some hands … if they have enough clubs, but usually it’s not that important.
Each of Agricola’s 350-ish cards adds a new twist on a rule. That makes 60,000 two card combinations that are much more complicated than the relationship between the King of Spades and deuce of Clubs. When you confine things to the E deck, I suspect most two card combinations have been seen. Are there three card combinations? Undoubtedly. Assuming 100 cards (for E decks only) there are 160,000 3-card combinations. You start with seven cards and seven occupations. Five hundred (or a thousand) games is enough to smooth all single cards, but doesn’t begin to assess the combinations … (and how many of those games involved new players)?
I could argue which cards I think are problematic (and I will, but not now). A fair response is that I (or others) missed a counter to that card. I may well have. But we’re still playing a card game. Now, the question is – how many routine hands do we have? For now, call a routine hand one that “Given reasonable and competent players, the ‘better’ hand will win barring mistakes.” You can have routine good hands and routine bad hands, it may be exciting to take 13 tricks with thirteen of one suit, and it’s rare, but it doesn’t require skill. Likewise, it takes little skill to lose all 13 tricks with a flat yarborough.
Apart from routine hands you can have routine games. In card games like bridge, one side has a bad hand the other one often has a good one. These routine hands lead to routine deals. Games where you deal a subset of the cards may see multiple players have very good (or bad) hands.
Race for the Galaxy, which I love, has routine games. I estimate roughly 3%. If you picked a number from 1-10%, I’d be fine with that. I’d argue if you went much higher (or lower). From my (admittedly small) sample of Agricola, I’d put its “routine game” percentage at 50-75%. My estimate may be high since every game I’ve played had new players. I think everyone will agree that this number exists above 10%, and that’s problematic for a game thats 4-8 times as long as Race. (Even if you discount routine ‘good’ hands, you’ll occasionally see a routine ‘bad’ hand, such as one that has no useful early minor improvements, which makes several of the actions much worse).
Only once have I been unsure of the winner after the second harvest, and it that game I had picked two people (out of five), who came in 1st and 2nd. I publicly predicted the winner of my second game (which we finished) during turn 3. He botched the endgame (never building a single fence, so earning -1 for enclosures and something like -5 for unused spaces) and won handily.
[Tangentially, this is why I stopped playing Cosmic Encounter for years … everyone wanted to play with 2+ power combinations. I now prefer playing with single powers, they are reasonably well balanced].
I hope scores will tighten as players get better, but the cards have a huge impact.
Agricola has other issues, as well. These aren’t nearly as important, but exist.
- The ‘family increase’ mechanic (which moves you from two actions to three) is very important, and provides a positive feedback mechanism. Feeding does produce a negative feedback as well, but not nearly at the same effect. Assuming no occupations (and no player manages to get a 4th action before anyone gets a 3rd) then the last player to grow his family will miss out on 4 actions (in a five player game). That’s effectively giving the first person an extra full turn. This effect is so important that ignoring everything to focus on family growth seems to be a dominant strategy. Worse yet, growing your family is a strong source of victory points. If growing your family cost you resources, and earned actions but no victory points, it would be a more interesting tradeoff.
- [1a, really] – The extra action from family growth is so powerful that the ‘family game’ (without cards) is simply an exercise in getting your home ready for your first child. With two players, even one extra turn seems dominant. [The family game is still a card game, but now the deal is the ordering of the rounds.]
- Because of the card interaction and family growth issue, I suspect every 4+ player game will have at least one player ‘eliminated’ early on, with a score of roughly half (or less) of the winning score..
- Livestock seem a much superior form of food & victory points than farming. Like family growth, this makes the best “resource” path also the best “victory points” path (since there are four categories for livestock, vs three for agriculture … and stables makes a fifth, arguably). You can delay your plowing until the very late game and still get good agricultural points, but you can’t put off breeding.
- Turn order effects. When the start player is passing back and forth between you and another player, it really matters if you are sitting besides each other. Also, some cards improve certain spaces, and again order matters. Caylus’ “Inn” mechanism works much better than Agricola’s blunt “Start Player.”
The good news? Often the game conceals this from you. Most hands have something, and monstrously good hands may be hidden (until the end). There’s enough going on that even a good hand can be misplayed.
Agricola took several great ideas (a Caylus-like placement system, a Cosmic-like special power system, a complex resource management system) and then shoved them together. It’s enjoyable but, like most cross-breeds, an odd beast. I suspect that, like Age of Renaissance in particular, I'll eventually get annoyed playing around with all those fiddly pieces to decide a card game. I'd be happy to be proven wrong, though.
Update: I've also ported my review to BGG.
Related Posts (on one page):