Dominion Strategy Forum

Please login or register.

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

Author Topic: Buying cards based on expected value of later hands in a shuffle  (Read 1941 times)

0 Members and 1 Guest are viewing this topic.

Davio

  • 2012 Dutch Champion
  • *
  • Offline Offline
  • Posts: 4787
  • Respect: +3412
    • View Profile
+6

Granted, the title is somewhat vague, but here's an example of what I mean.

Let's say I'm playing a money-ish game and I have $6 after a shuffle, there's some $5 I really want one copy of right now, but not two (just yet).
I'll often go ahead and buy the $5 with my $6 hand (fearing I won't have $5 later this shuffle), only having to buy a Silver with my next $5 hand.
If the situation was reversed, $5 first, $6 second, I could have gotten the same $5 with a Gold.

But of course when I buy a Gold with that $6 hand, I end up with $4's for the rest of the shuffle and won't be able to buy the $5 I so desperately want.

Something I can improve on is to think shuffle to shuffle instead of hand to hand. With any given hand, I tend to think only about the spending power of the current hand and getting the cards I would like to have right now (not next shuffle). So for example if I have too much terminals right now, I might be more inclined to overpay for a Village.

At the start of the shuffle, think: what does my deck look like now and what do I need for the next shuffle? Then try and get those cards in the order that optimizes your chances of getting all of them.

What are you guys' thoughts on this? Do you think about it this way, do you have the same problems as me?
Logged

BSG: Cagprezimal Adama
Mage Knight: Arythea

WanderingWinder

  • Adventurer
  • ******
  • Offline Offline
  • Posts: 5275
  • ...doesn't really matter to me
  • Respect: +4384
    • View Profile
    • WanderingWinder YouTube Page
Re: Buying cards based on expected value of later hands in a shuffle
« Reply #1 on: January 12, 2016, 08:39:50 am »
+6

Well, this is more of a problem in BM games than engines, and it doesn't come up much except early on. It does come up in the first few shuffles, and it's one of the trickier things. You need to know how likely you are to get back to 5 (or whatever price point), as well as how much better the e.g. 6-cost is for you, and how bad missing 5 is.

In general, I tend to play it safe, because in most of these situations, or at least the ones I find myself in, getting back to my price point isn't likely enough, and it tends to be disastrous to miss it.

AdamH

  • Margrave
  • *****
  • Offline Offline
  • Posts: 2833
  • Shuffle iT Username: Adam Horton
  • You make your own shuffle luck
  • Respect: +3879
    • View Profile
    • My Dominion Videos
Re: Buying cards based on expected value of later hands in a shuffle
« Reply #2 on: January 12, 2016, 08:49:54 am »
0


<3 <3 <3 <3 <3 <3 <3 <3 <3 <3 <3

Uhh, to answer your question, I generally only do this when I have one turn left before I shuffle and I know exactly (or close enough) what I'm going to have to spend that turn. Otherwise, I just buy what I think is the best card for my deck.
Logged
Visit my blog for links to a whole bunch of Dominion content I've made.

aku_chi

  • Minion
  • *****
  • Offline Offline
  • Posts: 622
  • Shuffle iT Username: aku chi
  • Respect: +1436
    • View Profile
Re: Buying cards based on expected value of later hands in a shuffle
« Reply #3 on: January 12, 2016, 08:59:54 am »
0

I figured other people played this way, which is why I lose more often on money-ish boards than on engine boards.  ;D

I understand cognitively that this is the optimal way to play, but I often lack the diligence to keep track of all the data I need (the total money capacity in my deck, how much of that money I've seen since the last shuffle, and how much money missed the shuffle).  It probably isn't much more difficult than tracking the exact contents of my deck in a deck-drawing engine, but it's more work for a lower payoff.

I find that this kind of tracking is easiest and most rewarding in big money deck with some reliability: like Gear and Courtyard.  With Gear big money, it's optimal to open Gear + Silver, because that results in 9 total coins in deck that can be drawn in two turns: Gear + Gold before the next shuffle is the goal.  Gear + Gold is guaranteed if you draw your Gear on turn 3, or have exactly 3 or 6 coins on turn 3.  Gear + Gold results in 12 coins in deck that you can draw in 2 turns.  If you have a Gear on turn 5, you have a good chance of getting Gold + Gold into your shuffle after turn 6.  The challenge occurs when this best-case plan doesn't manifest.  Then, to play optimally, you need to keep track of where all of your money and Gears are.  And if there is a 5-cost card worth getting (there often will be), that's an additional variable to keep track of.  Courtyard and, especially, Gear reward this kind of thinking because of how much control you have over: (a) distributing coins between hands in a given shuffle and (b) determining what misses the shuffle.
Logged

Davio

  • 2012 Dutch Champion
  • *
  • Offline Offline
  • Posts: 4787
  • Respect: +3412
    • View Profile
Re: Buying cards based on expected value of later hands in a shuffle
« Reply #4 on: January 12, 2016, 09:24:54 am »
0

Well, this is more of a problem in BM games than engines, and it doesn't come up much except early on. It does come up in the first few shuffles, and it's one of the trickier things. You need to know how likely you are to get back to 5 (or whatever price point), as well as how much better the e.g. 6-cost is for you, and how bad missing 5 is.

In general, I tend to play it safe, because in most of these situations, or at least the ones I find myself in, getting back to my price point isn't likely enough, and it tends to be disastrous to miss it.
I guess it's a simple issue of risk vs reward.

The reward of having a Gold over a Silver isn't as big as missing the crucial $5.
Still I always feel a little frustrated when I could have gotten more out of a shuffle. :)
Logged

BSG: Cagprezimal Adama
Mage Knight: Arythea

DG

  • Governor
  • *****
  • Offline Offline
  • Posts: 4074
  • Respect: +2624
    • View Profile
Re: Buying cards based on expected value of later hands in a shuffle
« Reply #5 on: January 12, 2016, 11:26:35 am »
0

Yes. If you can make a worthwhile estimate of future turns you should use it.
Logged

Titandrake

  • Mountebank
  • *****
  • Offline Offline
  • Posts: 2210
  • Respect: +2856
    • View Profile
Re: Buying cards based on expected value of later hands in a shuffle
« Reply #6 on: January 13, 2016, 03:16:21 am »
0

I definitely do this. Sometimes it costs me, sometimes it doesn't. I don't remember often enough either way to judge whether it's a net benefit or not.

I try to think shuffle to shuffle, and will sometimes play this way very aggressively. Ex, picking up a terminal draw that costs $5 when my deck is already action starved, because I expect to hit enough hands to pick up the Villages I need. One downside is that it can lock in your buys: if your opponent does something unexpected, it's hard to react to it that shuffle because you've already got plans for the next 2-3 hands.

Is this a net good thing? Maybe? I don't know if I do it well enough, my style of play is to think for a long time, choose the risky decision more often, and lose in very dumb ways. Even if it's net neutral or net loss, I'll still do it to get practice on shuffle to shuffle thinking.
Logged
I have a blog! It's called Sorta Insightful. Check it out?
Pages: [1]
 

Page created in 0.074 seconds with 20 queries.