The end times

Knowing the duration of a game is critical.  So often strategy games come down to various players building plans that will peak at different points in the game and the winner ends up being the person who manages to get the game to end just when their strategy peaks.  Puerto Rico is a great example, where the builders want the game to end as fast as possible because their strategy peaks right when their second big building finishes, while the shippers peak right when the last shipping point leaves the pile.  If the builders manage to finish the game quickly they will win because their strategy peaks much sooner.  Generally that means that in a game with three people shipping and one building the builder loses because all three shippers refuse to take actions that end the game and the shippers will help each other lengthen the game to generate more points from the Harbour and Wharf.  Similarly a single shipper will lose to three builders, with the winner almost always being the person who builds the Guild Hall.

During my birthday party on Saturday this was on display quite clearly.  I didn't get my timing perfectly right but on my last turn I completed my second Basement room and also finished off my big purple room to rescore it for 9 points.  I had thought the game would go one more turn and if it had I would have done better but I can't complain because I did manage to complete all of my major plans, albeit a bit awkwardly.  I won the game by a big margin, which wasn't a surprise because I was teaching all three of the other players the game.

Some of the other players asked what they had done wrong and why they lost, and as usual the obvious errors were mostly the placement of the best tiles on the 15k space.  In an auction game you don't want to let people get good stuff cheaply, but you can't keep the good stuff out of their hands indefinitely so you really want to sell it to them at a high price rather than having them buy something else for cheap.  The best tiles shouldn't be put out of reach - the optimal placement for them is *just barely* within reach.  That is when you get paid, and being rich gives you options, including the option to pay a ton for a tile if you really want to.

However, I don't think their tile placements were the reason I won by a lot.  The main factor was that I knew when the game was going to end.  I bought one big purple room and focused my game around bulking it up and completing my other stuff opportunistically.  The other players bought multiple big purple rooms and aimed to complete everything and they all ended up with a bunch of rooms incomplete.

There is nothing wrong with incomplete rooms in theory but you have to pick which ones you will leave out.  Buying a cheap 4 point room for 1 or 2 coins and sticking it on to close a door is a great play even if you make no attempt at all to finish it.  Paying 6 coins for a 2 point room that has a huge completion bonus and then never completing it is a disaster.  It isn't always true that buying a second big purple room to build around is a bad choice but you have to be really careful that you don't bite off more than you can chew.

One of the key tricks is knowing when to take the money and run.  I bought the big purple room that gives a 4 point bonus for each attached yellow room.  I quickly slammed another tile onto it that wasn't yellow because hoping to get a full 4 yellow rooms attached to it is just too optimistic.  You need to know when to start closing doors and accepting that you won't get the maximum points possible.  You can spend the game desperately trying to score 34 points from that room, but the likely result is that the game ends and you get 9 instead.  Being willing to accept a lower payout that is much more likely to come home is how you make 18 points like I did, and that is the more likely path to victory.

Knowing when to start finishing up is a skill that takes time to hone.  If you give up too quickly you won't score many points because you attach the wrong stuff, but if you wait too long you don't finish your room.  You want to end the game with every major completion bonus done, but only just barely.  Too soon, and you miss opportunities to max out.  Too late, and you get nothing.

It should not be taken that the people I played against were bad.  They played well, for first time players.  Castles is a game that is extremely sensitive to the timing of game end and I think that skill is actually one of the hardest ones to master when playing a new game.  You need a full understanding of the way the game flows before you can really make any decent decisions in that regard anyway, so until you have played a number of times figuring out how to design your plan to peak right at game end isn't really plausible.

For anyone looking for a best guess on how fast to finish up your rooms, I would suggest assuming that you need twice the number of plays as you have doors available.  If your big room has three open doors and there are about five turns left, start closing them off with anything you can.  If you have eight turns left, chill and wait it out.  With brick rooms you can generally assume about the same since there are more tiles that work and you don't get nearly as much from finishing it so you really don't want to slam crappy things on there.  Leaving a brick room incomplete isn't so bad because you got most of your points when you dropped it, so you can afford to be picky.