Dominion Strategy Forum

Please login or register.

Login with username, password and session length
Pages: [1]

Author Topic: End Game Example  (Read 5102 times)

0 Members and 1 Guest are viewing this topic.

DG

  • Governor
  • *****
  • Offline Offline
  • Posts: 4074
  • Respect: +2624
    • View Profile
End Game Example
« on: February 06, 2015, 05:56:51 pm »
+27

Many people have asked for articles about end game play so I'm going to use a recent league game against Stef for a discussion of some end game principles. You can only win a Dominion game through empty piles: either the province pile or three other piles. A lead at any other time may not be decisive. If you need to take more than one turn to empty piles then the opponents will always have a chance in the intervening turn to score and win. It becomes easier to for them to win if you haven't scored enough points but have lowered the piles.

In this example game, Stef and I are involved in a competitive endgame with strong decks. We are both looking for a way to close out the game in one turn or score points and deny the opponent a chance to win on the following turn. The kingdom is as follows.


Code: [Select]
Loan, Familiar, Bishop, Remake, Silk Road, Bazaar, Council Room, Ghost Ship, Mystic, Torturer
This is quite a complicated kingdom anyway and it could certainly have been played differently in the early game. Let's skip that though and jump to the endgame at turn 18, with Stef going first but 6 vp behind.

Stef's deck - 6 bazaars, 3 council rooms, ghost ship, 5 mystics, remake, bishop, 2 silver, 2 provinces, duchy, 2 curses, 1 vp chip.
DG's deck - 4 bazaars, 2 council rooms, ghost ship, 2 mystics, remake, 2 bishops, familiar, silver, province, silk road, 14 vp chips.
Relevant Supply piles - 4 provinces, 7 duchies, 8 estates, 0 bazaars (empty), 5 council rooms, 3 mystics, 8 ghost ships, 7 bishops, 3 remakes, 3 curses.

Both decks draw quite reliably despite the attack from the ghost ship. In the game Stef was able to draw the entire deck with 3 actions remaining, 4 buys, 16 coins, and a hand of bazaar, 2 mystics, remake, bishop, 2 silver, 2 provinces, duchy, 2 curses. Question number one, is it possible to win on that turn? Well, I'll give the answer later. In the game Stef played the remake on a silver and a curse gaining a bishop, drew the bishop with the bazaar, and played the two bishops on the last curse and the duchy. This left him 3 points behind with 4 buys and 19 coins. Question number two, is what would you buy there? Again, we can discuss that later.

After cleaning up his deck with the card play, Stef bought one province for a 3 point lead and two torturers for extra control. Moving onto my turn,  I was able to draw the entire deck with 2 actions remaining, 3 buys, 8 coins, and a hand of bazaar, ghost ship, remake, bishop, bishop, familiar, silver, province, silk road. Question number three, is it possible to end the game on that turn?

In fact, that is the final question as I did manage to end the game there. It was possible to  remake the silk road into a mystic and the silver into a remake, draw those cards with the bazaar and familiar, and then remake the two bishops into mystics. This empties the mystic pile so buying a province and the last two curses ends the game for a 2 vp win.

Now lets go back and look at question two. My final winning score was only two points so should Stef have bought another province for a bigger lead? It's not quite as simple as that. On my final turn it was also possible to remake the silk road into a mystic and the ghost ship into a gold, draw them with the familiar and bazaar, and play them and a bishop for 16 coins to buy two provinces. Although Stef did need to buy more vp that turn, he needed to leave those three provinces in the supply and get vp from duchies.

Now lets return to question one. Was it possible for Stef to win on his turn? Yes it was possible to remake both silvers into remakes, draw them with a bazaar and mystic, then remake the remake and bishop into two more mystics, use the final mystic in hand to draw one mystic that then can draw the other. Play the final mystic to give 20 coins. This is enough to buy the final mystic, the final remake, a province, and an estate, for a 1 vp win.
« Last Edit: February 06, 2015, 06:00:03 pm by DG »
Logged

DG

  • Governor
  • *****
  • Offline Offline
  • Posts: 4074
  • Respect: +2624
    • View Profile
Re: End Game Example
« Reply #1 on: February 06, 2015, 05:57:04 pm »
+7

What end game principles does this game show?

 - Controlling the ending controls the game: After one poor draw for Stef, I had a small lead and needed to close out the game before his stronger deck could recover. Emptying the piles over two turns would lose me the game as Stef would score strongly and win on the intervening turn (by emptying the piles himself). This is why a clean finish in one turn was important.

- Buys and Gains control the game end: In this game we could use a combination of buys and gains to empty piles. The gains were important as we could draw the gained cards from the discard pile and play them in the same turn. The buys were important as they could be used flexibly on any kingdom card (You can't remake any card into a curse).

- Actions support the play: Many +actions were needed to support the council rooms, ghost ship, bishops, and remakes. This made the bazaars into key cards for this kingdom. Stef correctly bought this pile out first.

- Draw it to play it: Strong drawing was needed to ensure not only the reliable drawing of the deck but the drawing of the extra cards gained during the turn. Without this extra drawing the end game finish would not have been possible.

It should be no surprise that three main engine components, gains, drawing, and actions, are important for end game control. Strong engines normally provide the best end game control. Let's have a look at some other topics though.

- Attacks can sometimes control piles: In this game I could play the familiar to give a curse and remove it from the supply pile, or choose not to play the familiar if it was not to my advantage. It was a good asset. Stef's torturers would also have been influential too if the game had continued.

- Leads are strong but not as strong as control: My win here was narrow and I would not have been able to overtake a bigger lead for Stef. However, if Stef had built a bigger lead with more provinces he would still have lost on province pile.

- Alternative scoring : My deck was much stronger for having bishops. In previous turns I had scored victory chips with the bishops and bought sparingly from the province pile. If I had needed to score all my victory points with provinces then the game would have rushed to an ending on the province and duchy piles, an ending I probably would have lost. Bishops also allow the scoring of points without increasing the deck size. With bigger decks we would have been short of the drawing needed to draw any cards gained the same turn.

- Options: Both decks had many ways of scoring and emptying piles. This not only aids play but also makes it difficult for an opponent to defend. Gaining new cards for play the same turn clearly adds many, many, options.

- Preparation and Timing: It might be impossible to control exactly what cards you have in your deck as you enter the end game, but it possible to assess the usefulness of the cards you do have. In this end game the 4 cost cards were a slightly better asset than the 5 cost cards since they could be remade into duchies or mystics when needed. However there was no need to convert these cards immediately. They needed to be held back until the time was right, in this case on the final turn.

- Find a way to win: On my previous turn I could have bought an estate instead of a silk road for an extra point to prevent Stef's possible win. Should I have done that? In balance I think no. I knew there was some likely finish for my next turn using a four cost card, however I didn't know the precise capability of Stef's deck. The estate was too safe and would have denied me chances to win. I was unlikely to win an extended game.

If end games seem complex then be assured that they actually are complex. At the time I had a different plan for achieving 16 coins (if needed) that would actually have failed. Even though you can't be perfect with endgame finishes, practice will make you better.
« Last Edit: February 06, 2015, 05:58:37 pm by DG »
Logged

DG

  • Governor
  • *****
  • Offline Offline
  • Posts: 4074
  • Respect: +2624
    • View Profile
Re: End Game Example
« Reply #2 on: February 06, 2015, 05:57:14 pm »
+1

« Last Edit: February 06, 2015, 10:16:50 pm by DG »
Logged

Honkeyfresh

  • Minion
  • *****
  • Offline Offline
  • Posts: 569
  • DSF lowest ratio upvotes-posts 14 yrs & counting
  • Respect: +192
    • View Profile
Re: End Game Example
« Reply #3 on: July 08, 2015, 09:17:40 pm »
+3

Very nice work in here.  This is a topic that I think gets short shrifted as people are often more focus on the building rather than the engame. 
« Last Edit: July 09, 2015, 04:34:27 pm by Honkeyfresh »
Logged
"Rap game Julio Franco, Chuck Norris, Texas Ranger/ Ice on my fingers look like I slap-boxed a penguin." -- Riff Raff Proverbs 4:20

"Sometimes I say some things people may think are just outlandish, but I'm going to have the last laugh." -- Riff Raff  Exodus 6:66

ancientcampus

  • Navigator
  • ****
  • Offline Offline
  • Posts: 76
  • Respect: +52
    • View Profile
Re: End Game Example
« Reply #4 on: July 11, 2015, 04:52:22 pm »
0

Wow, this is a useful article. I can say I usually don't think along these lines, save when I'm running a massive engine and plan to end the game on piles from the start. It's definitely something I need to do more to step up the game.

This case clearly showed the need to think of end-game situations considerably earlier than average players would expect. (Here, there were 4 provinces and only one pile missing. Ending the game on piles required 6 gains minimum.)

What I'd like to see explored are two other situations:

-A comprehensive discussion on advanced endgame considerations in 2-player matches, in boards without massive engines. There's the penultimate province rule, duchy-dancing, when to buy estates... I know I could stand to improve there, but I'm wondering what other factors exist that I'm not even considering.

-Endgame considerations in multiplayer games. Piles can run out more quickly, and a lot more decision theory needs to be considered.

Unfortunately, I'm slightly more skilled than most of my RL friends, and when I unexpectedly end a multiplayer game on piles it's less fun for them. Ah, well.

ephesos

  • Explorer
  • *****
  • Offline Offline
  • Posts: 347
  • Shuffle iT Username: Ephesos
  • Respect: +290
    • View Profile
Re: End Game Example
« Reply #5 on: July 12, 2015, 08:30:27 pm »
0

Not sure if my math is wrong here, but Stef played 5 of 6 Bazaars and 3 of 5 Mystics, giving him 11 coins, not 16.

He then played Remake Bazaar Mystic Mystic Bishop Bishop for 7 more coins, bringing him to 18, not 19.

If he had taken the course of action you suggested, he would have played Remake Bazaar Mystic Remake Mystic Mystic Mystic for 9 more coins, bringing him to 20(if he started at 11.)
Logged

markusin

  • Cartographer
  • *****
  • Offline Offline
  • Posts: 3846
  • Shuffle iT Username: markusin
  • I also switched from Starcraft
  • Respect: +2434
    • View Profile
Re: End Game Example
« Reply #6 on: July 12, 2015, 11:24:14 pm »
+2

"You can only win a Dominion game through empty piles: either the province pile or three other piles."

Consideration of three pile endings should really form part of your Dominion though process throughout the entire game, at least to the same extend as Province/Colony endings.

It's not just about spotting 3 pile endings. It's also about preparing your deck to end the game on a win with piles in advance when appropriate. Even if a three pile ending seems far away.

Having even the tiniest of leads can prevent your opponent from diving heavily into engine components to increase their deck's reliability because then you'll empty piles on a win. This idea also highlights why it's powerful to get your share of key card splits first. Imagine you gain 3 Grand Markets on your turn to get to 5 GMs in your deck. Your opponent can empty the GMs on their next turn perhaps, but you'll be the one to take advantage of 5 GMs in deck first, possibly ending the game on piles for the win. If your opponent doesn't empty the GMs, them you can do it yourself next turn to get like a 8-2 split and then win on the turn after that. Being the first to get your share of the engine components gives you the end game control advantage.
« Last Edit: July 12, 2015, 11:29:23 pm by markusin »
Logged

ancientcampus

  • Navigator
  • ****
  • Offline Offline
  • Posts: 76
  • Respect: +52
    • View Profile
Re: End Game Example
« Reply #7 on: July 13, 2015, 09:44:21 pm »
0

Quote
It's not just about spotting 3 pile endings. It's also about preparing your deck to end the game on a win with piles in advance when appropriate. Even if a three pile ending seems far away.

^That

-Stef-

  • 2012 & 2016 DS Champion
  • *
  • Offline Offline
  • Posts: 1574
  • Respect: +4419
    • View Profile
Re: End Game Example
« Reply #8 on: July 19, 2015, 07:18:37 am »
+2

Happy to see this article got some response, even though it's 5 months later :)

There certainly is a lot to say about endgames like these. It's hard to give simple rules that will always help, I think the best you can do to improve is to think hard about your own and study better players. But it requires a lot more effort then some other articles / videos that you can simply "consume". If you only read the text here, and don't study the log yourself, you're missing out on most of it.

Although reducing the number of forced wins I miss - like the one described here - is always a good idea, I don't think that's where I have most to gain. The 1 or 2 turns preceding the forced-win-turn is where it's really complicated. Unfortunately, the rules for those turns are even "softer" and discussing them that much harder.
Logged
Join the Dominion League!
Pages: [1]
 

Page created in 0.045 seconds with 22 queries.