Dominion Strategy Forum

Please login or register.

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

Author Topic: Silverspawn's trashing formula  (Read 1848 times)

0 Members and 1 Guest are viewing this topic.

Aidan Millow

  • Bishop
  • ****
  • Offline Offline
  • Posts: 110
  • Respect: +116
    • View Profile
Silverspawn's trashing formula
« on: December 13, 2013, 04:25:39 pm »
+1

I was still only lurking here when Silverspawn started a fight with half of the forum by suggesting that lookout was a better trasher for 3 than ambassador but I watched it with amusement and interest. One thing that interested me was his 'ratio' for the reduction in deck size vs the damage done to your current hand (which no-one actually refuted by the way) and I decided to reprise this here.

Silverspawn's 'ratios' were off as he treated all cards in hand as being of equal worth when they're clearly not; estates are worth nothing financially so removing them does nothing to harm your current hand while the spot taken in your hand by the trasher is likely worth 2 as it could have been a silver.

Here are my revised 'ratios'.

Lookout:
1/2

Ambassador (treating junking your opponent one card as equivalent to trashing one):
1/2 (does not return anything, just junks your opponent), 2/3 (returns a single copper), 3/4 (returns 2 coppers), 2/2 (returns a single estate), 3/2 (returns 2 estates)

Obviously Ambassador is superior here but I thought it would be interesting extending this to other trashers that cost 4 or less (for simplicity's sake I'm only going to consider gaining money and trashing copper or estates even though most of these cards are worse than normal in such a situation).

Chapel (obviously has many more possibilities but I'm going to assume you trash all the estates and coppers you see):
3/5-3/2 (collides with your other opener), 4/6-4/3 (doesn't)

Moneylender:
1/0

Remodel:
0/2 (draws without estates and doesn't trash a copper), 0/-1 (trashes estate, grains silver (yes I am aware that gaining a card worth 3 is not equivalent to the worth of 3 money in your hand))

Masquerade (early on +2 cards vs 2 money is a wash):
0/-1 (passes estate for copper, doesn't trash), 1/0 (passes estate, trashes whatever's recieved), 1/-1 (passes estate for copper, trashes estate), 0/0 (passes copper for copper, doesn't trash), 1/1 (passes copper, trashes)

Steward:
0/0 (doesn't trash), 2/4-2/2 (does)

Island:
1/3-1/2

Loan:
0/1 (hits silver), 1/1 (hits copper)

Trade Route:
1/2, 1/3

Remake:
2/4 (cc), 1/0 (ce), 0/-4(ee)

Develop:
1/3, 0/-3

Jack:
0/-1 (trashes), -1/0 (doesn't)

Spice Merchant:
1/1

Trader:
1/3, -1/-6

Forager:
1/2 (trashes copper or estate (having not trashed a treasure earlier)), 1/1 (trashes estate after trashing copper)

Hermit:
0/-3 (trashes), -1/-3 (doesn't)

Rats:
0/0-0/1

Doctor:
0/2-3/2

What do people think: how much value does such an analysis actually have?
« Last Edit: December 13, 2013, 05:54:49 pm by Aidan Millow »
Logged

qmech

  • Torturer
  • *****
  • Offline Offline
  • Posts: 1918
  • Shuffle iT Username: qmech
  • What year is it?
  • Respect: +2320
    • View Profile
Re: Silverspawn's trashing formula
« Reply #1 on: December 13, 2013, 04:37:54 pm »
+1

Logged

markusin

  • Cartographer
  • *****
  • Offline Offline
  • Posts: 3846
  • Shuffle iT Username: markusin
  • I also switched from Starcraft
  • Respect: +2437
    • View Profile
Re: Silverspawn's trashing formula
« Reply #2 on: December 13, 2013, 06:39:26 pm »
0

I remember that argument. I forgot to bring this up, but I wanted to know what people think of an Ambassador/Lookout opening. I think there are boards where it's better than both Amb/Amb and Lookout/Lookout. I know I've won with that opening against Amb/Amb on a board that had very little engine potential.

Nice to see a compilation of 'ratios' though.
Logged

RD

  • Thief
  • ****
  • Offline Offline
  • Posts: 93
  • Respect: +70
    • View Profile
Re: Silverspawn's trashing formula
« Reply #3 on: December 13, 2013, 11:33:38 pm »
+2

Well, first off it's pretty clear these aren't (or shouldn't be) "ratios" in the normal sense. The utility of money, handsize, trashing, etc are all highly nonlinear, to say nothing of board-dependent, so there's no reason to assume that for instance the "1/2" of Lookout is equivalent to the "2/4" of Remaking 2 Coppers. The only thing we can really conclude is that 3/1 is better than 2/1, or 1/1 is better than 1/2, but that was already obvious.

I think the core idea is indisputable: trashing usually comes with an opportunity cost, and all else equal you would want that exchange to be as favorable as possible. (I also agree the opportunity cost mostly consists of the economic damage in the Buy phase of the turn where the trasher comes up, although there are others. The buy spent on the trasher is also big of course, but we can ignore that part if we're comparing Lookout to Ambassador as it's assumed we will buy one or the other.)  The question seems to be whether you can quantify this tradeoff, in a meaningful much less Kingdom-independent way, and unfortunately it sounds like a very tall order.
« Last Edit: December 14, 2013, 10:38:36 am by RD »
Logged

EvanC

  • Pearl Diver
  • **
  • Offline Offline
  • Posts: 11
  • Respect: +9
    • View Profile
Re: Silverspawn's trashing formula
« Reply #4 on: December 13, 2013, 11:40:27 pm »
0

I remember that argument. I forgot to bring this up, but I wanted to know what people think of an Ambassador/Lookout opening. I think there are boards where it's better than both Amb/Amb and Lookout/Lookout. I know I've won with that opening against Amb/Amb on a board that had very little engine potential.

Nice to see a compilation of 'ratios' though.

In this specific situation I could see this opening being stronger. If there are no plus actions to keep the ambassadors from colliding once you trim your deck down, going amb/lookout means your two trashers will never collide. Once your deck becomes rich enough you don't want to lookout, it becomes an ambassador target.
Logged
Pages: [1]
 

Page created in 0.04 seconds with 20 queries.