Dominion > Dominion General Discussion

Early Deck Cycling Revisited

(1/1)

philosophyguy:
There were recently a number of threads about the statistics of playing a Venture or other cycling card and how that impacts the deck distribution, if at all. Before that, there were threads discussing cycling in general, with the general consensus that cycling early is good because it gets new (and presumably stronger) cards into your deck, while cycling once you start greening is disadvantageous because you're adding dead cards to the deck.

All of this analysis seems to overlook something really important: namely, that the impact of lucky or unlucky cycling is not equal, even if the odds are equal for having a lucky or unlucky cycle. The point was driven home to me in a recent game where my Loan flipped my opening buy on the first play and set me behind for turns 4 and 5, and then the rest of the game. It's the same reason that having a Sea Hag flip a Sea Hag is often gg. If you cycle in a lucky manner, you may gain at most a turn advantage. If you cycle unluckily, however, you can be set back a full reshuffle.

I think this suggests that early deck cycling (especially blind cycling, like Loan or Fortune Teller) is more dangerous than usually believed. Cycling effects are a high-variance strategy early on, with a much bigger downside than upside in most cases.

What this implies, in terms of strategy:

* Self-cycling cards like Loan are extremely risky as openers
* In multi-player games and games with a high first player advantage, cycling attacks like Fortune Teller or Pirate Ship become better buys for players in later turn order
Thoughts?

ecq:
Let's call this "blind cycling" since anything that draws cards (e.g. Smithy) will cycle the deck.

I think the risk of opening with Loan would depend a lot on the kingdom, but my hunch is that the odds of really screwing yourself are pretty low and mitigated by the fact that you're cycling with the Loan, so if you do discard your other opening buy, you'll find it faster.

I say that, but I can't tell you how many times Loan has failed to find a Copper in my early hands.

There's a pretty disgusting case worth noting here, where you draw Loan on turn 4 with an Estate and opening non-treasure in the draw pile.  Either you play Loan and make your other opening buy miss 2 shuffles (and your Loan miss one shuffle) or you treat the Loan as a dead card in your current hand and still miss 1 shuffle with your opening buy.

WanderingWinder:
The venture thread showed pretty conclusively that the mean and variance of your next hand, indeed the whole distribution, remain pretty unchanged, in terms of treasure, as long as you don't reshuffle. Of course, you do reshuffle and there are action cards to work with, and so the cycling adds some variance. But risky != bad. Risky is only bad for you if you're more than likely to win to start with. And good for you if you're way behind. So you shouldn't really avoid a play because it's risky, at least out of the opening. You just want to play what's good. (Later on, if you have a big lead, you play less risky, and if you're behind, you play more risky)

DG:
If you feel that your loan turning over your 4 cost card is disaster, three times more unlucky than your loan turning over an estate (or two estates or three estates) then perhaps you should avoid the loan. This however comes under different theory: don't expect to get full value from action cards in a loan/venture deck since all your treasures get played in each cycle but some other cards might be discarded. Similarly, don't expect to get full value from treasure in a golem deck since all your actions get played each cycle but some other cards will probably be discarded.

jomini:

--- Quote ---If you feel that your loan turning over your 4 cost card is disaster, three times more unlucky than your loan turning over an estate (or two estates or three estates) then perhaps you should avoid the loan. This however comes under different theory: don't expect to get full value from action cards in a loan/venture deck since all your treasures get played in each cycle but some other cards might be discarded. Similarly, don't expect to get full value from treasure in a golem deck since all your actions get played each cycle but some other cards will probably be discarded.
--- End quote ---

I'm not sure I buy this, yes you may not get as many plays of your actions per shuffle, but your shuffles come sooner. The real measure of utility is how many hands you will play the action (or treasure) in the game. For instance, with loan flipping your green, decreasing your deck size by trashing copper, and just cycling other treasure, the time between reshuffles goes down. Further loan's increase to your cycling rate is a geometric progression with just its copper trashing (the first shuffle is hit with fewer cards played, the second with three, the third with six, the fourth with ten, etc.). I think it is going to depend upon deck composition and the length of the game if on net you will play power cards more or less with loan.

Navigation

[0] Message Index