Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[florrat]

When cycling is not good

Cycling has two meanings:
- discard cards from your deck (like Cartographer and Spy);
- draw and discard cards from your hand (like Warehouse and Minion).
 
We've all been told that cycling is good. However, in this article I want to argue that cycling is definitely now always good, and sometimes very bad. I definitely don't want to argue that cellar and warehouse are bad: they provide "good cycling." However, people have been defending Advisor for "providing cycling" or alleviating mucking attacks (Spy, Fortune Teller, ...) since they provide cycling, and this is wrong. These cards provide "bad cycling," and this will hurt you, not help you.

So let's distinguish between three types of cycling, good cycling, neutral cycling and bad cycling (I'm not very good with making up original names).
- Good cycling is cycling when you discard bad cards. This typically happens when you choose which cards to discard or when a card is designed to automatically discard bad cards. Examples include Cartographer, Cellar, Warehouse, the non-attack part of Spy, Farming Village, Sage, Journeyman, Wandering Minstrel, Chancellor, ...
- Neutral cycling is cycling which discards average cards. This often happens when digging. Examples include: Rebuild, Loan, Harvest, Golem, the attack part of Minion, ...
- Bad cycling happens when you discard good cards. This typically (but not always) happens when you're attacked. Examples include: Fortune Teller, Advisor, Envoy, the attack part of Spy, ...

Most examples of cycling are "good cycling," which is why we say that cycling is good. However, this does not mean that the cycling provided in the other two categories are good. Let's discuss the three types in more detail.

Good cycling is good, because by discarding your bad cards you'll see your good cards more often, and your bad cards less often. This is why Farming Village is better than village, Cellar is better than Ruined Village and Sage is better than Pearl Diver. We knew this already, so I don't have much to add here.

Neutral cycling happens when you discard arbitrary cards from your deck. This won't help you at all (but it won't hurt you either). The simplest way to see this: suppose you have 17 cards in your deck + discard at some arbitrary point during the game, and you haven't done any cycling this shuffle (so that the cards in your deck are on average as good as the cards in your discard pile). Now suppose we execute the following effect:
Do this 17 times: discard the top card of your deck.
After you've done this, you have the same the same number of cards in your deck as you initially had. So this effect hasn't done anything (sure, the cards in your deck are different, but are as good on average). You won't see your good cards more often or your bad cards less often. This means that if you execute the effect "discard the top card of your deck" once, it will - on average - also neither hurt nor help you. This is also the case with the examples provided above (*).
I want to spend a few more words on Chancellor. Chancellor provides good cycling, because you have the choice to discard your deck: you can discard your deck when you've seen more good cards than expected and leave it on your deck when you've seen fewer. If chancellor forced you to always discard your deck, I'd put it under "Neutral cycling". However, even in that case it would still have a very little (usually negligible) positive effect, because you'll see the card you've just bought faster on average, and during the biggest part of the game you're buying good cards. Also, sometimes it can be good because you have cards which do something with your discard pile (like Hermit, Counting House, Inn, or indirectly Stash)

Bad cycling is bad in the same way that good cycling is good. When good cards are discarded from your deck, you'll see them less often, and hence the bad cards more often. Advisor can be described as "opponent discards good card from your deck + Lab." Clearly Advisor is worse than lab, and this disadvantage is NOT alleviated because Advisor cycles. Advisor is worse, BECAUSE it cycles - in a bad way. This is why opening Advisor is typically bad: you'll be less likely to see the other card you've bought on your second reshuffle. Similarly, when an opponent discards a good card from your deck with Scrying Pool you shouldn't think "well, at least I've cycled": what happened was bad because you cycled, again in a bad way.

As final remark: think of Scheme and the on-gain effect of Inn as anti-"bad cycling": you'll put good cards back on your deck. Clearly the effects of Scheme and Inn are good, hence the opposite effect, discarding good cards, are bad.

---

(*) Mathematically speaking, Rebuild provides good cycling and Golem provides bad cycling, although for all practical purposes these effects are negligible. To see this: suppose you play rebuild and you have exactly 1 victory card in your deck+discard. Then after playing rebuild, your deck is guaranteed to be without victory cards, hence your deck has better cards on average. The same holds with Golem if you have 2 action cards in deck+discard: after playing Golem your deck is guaranteed to not have action cards. However, these effects become much smaller when you have more victory/action cards in your deck+discard. If you have n victory cards in your deck, every victory card has 1/n chance to get discarded, while all non-victory cards have 1/(n+1) chance to get discarded.

[silverspawn]

ahm, no. you missunderstand the whole purpose of cycling: to see your new cards faster. more specifically, to draw the card that you bought this turn more quickly.

Do this 17 times: discard the top card of your deck.
After you've done this, you have the same the same number of cards in your deck as you initially had. So this effect hasn't done anything (sure, the cards in your deck are different, but are as good on average).

yes, because you cycled beyond the point of an empty drawing pile. if you have 10 cards in your draw pile and 7 in your discard pile, you want to "cycle" through 10 cards, so that the card you bought this turn will be in your drawing pile the following turn. at the moment where you draw your 11th card, you ruin the effect and jump to worst possible situation: having a fresh drawing pile and the longest possible time for your new card to kick in.

chancellor is good (well, he is not good, but you know what i mean) because he gets your deck exactly where you want it: an empty drawing pile.

the aspect you are focusing on is just the average quality of cards in your deck and discard pile, which is the everlasting side effect of cycling. but as you explained, if you discard one card and don't trigger a reshuffle, it's good on average, because now your new cards will make it into your deck faster, and the average quality of cards in your drawing pile hasn't changed.

of course, there are situations where you want to trigger a reshuffle, because having the new card in your drawing pile is less important than having all the bad cards in your hand not in your drawing pile. in the same manner, there are situations where you don't want to play chancellor. but on average cycling is a good thing for as long as you are improving your deck, and it becomes a bad thing once you buy green cards, i.e. downgrade your deck.

and yes, fortune teller is in fact bad because it speeds up your opponents cycling. if you play a shelter game and your opponent has his overgrown estate in his hand, you are playing chancellor for him. this is not neutral, it means that his new card will kick in several turns earlier.

[Awaclus]

Bad cycling isn't bad for you because it's cycling, it's bad for you because it's bad. Advisor does two things in addition to a Lab: it makes the remaining cards in your deck worse on average and fewer in number. Making them worse on average is bad, making them fewer in number might be good or bad depending on the situation.

The cycling itself, without the "bad" part, does the following things:
1) it makes you see the cards in your current hand and past hands after your last reshuffle faster
2) it makes you see the cards you will buy/have bought this turn and past turns after your last reshuffle faster
3) it makes you see the Advisor itself faster

1) is good when your hands have been better than average and bad when your hands have been worse than average. On average, it's not good or bad. 2) is good when you have been buying good cards and bad when you have been buying bad cards (meaning cards that you don't want to see in your hand, i.e. green). In the early game, it's good, in the late game, it's bad. 3) is good pretty much always, unless Advisor is a card that you don't want to see (and why would you buy it in that case?).

Most of the time, the "cycling" part is actually good for you. The "bad" part is bad for you, and usually more significant.

yes, because you cycled beyond the point of an empty drawing pile. if you have 10 cards in your draw pile and 7 in your discard pile, you want to "cycle" through 10 cards, so that the card you bought this turn will be in your drawing pile the following turn. at the moment where you draw your 11th card, you ruin the effect and jump to worst possible situation: having a fresh drawing pile and the longest possible time for your new card to kick in.

No, that's not right. If you have 10 cards in your draw pile and 7 in your discard and you draw your 11th card, the 7 cards in your discard are exactly where you want them to be: in your draw pile. If you cycle 6 cards more, there's a chance that they will be in your discard again or still in your draw pile.

[silverspawn]

No, that's not right. If you have 10 cards in your draw pile and 7 in your discard and you draw your 11th card, the 7 cards in your discard are exactly where you want them to be: in your draw pile. If you cycle 6 cards more, there's a chance that they will be in your discard again or still in your draw pile.

yea, if you draw the 11th, the cards are in your draw pile. but why would you want them to be there?

[Awaclus]

No, that's not right. If you have 10 cards in your draw pile and 7 in your discard and you draw your 11th card, the 7 cards in your discard are exactly where you want them to be: in your draw pile. If you cycle 6 cards more, there's a chance that they will be in your discard again or still in your draw pile.

yea, if you draw the 11th, the cards are in your draw pile. but why would you want them to be there?

Because they're pretty good cards and you want good cards to be in your draw pile. You know, so that you can draw them.

[silverspawn]

No, that's not right. If you have 10 cards in your draw pile and 7 in your discard and you draw your 11th card, the 7 cards in your discard are exactly where you want them to be: in your draw pile. If you cycle 6 cards more, there's a chance that they will be in your discard again or still in your draw pile.

yea, if you draw the 11th, the cards are in your draw pile. but why would you want them to be there?

Because they're pretty good cards and you want good cards to be in your draw pile. You know, so that you can draw them.

but they are not good cards. on average, the cards in your (hand and in play) are above average, and the cards in your discard pile are below average, because you put bad cards there with sifting.

[florrat]

Okay, I agree that on average, discarding a random card from your deck is good. However, this effect is very small (the chancellor effect is not that big, and discarding just 1 card from your deck is n times as small (where n is your deck size)). Also, I think that a good part of chancellor's strength comes from the option to discard your deck. I think that a "forced chancellor" (where you're forced to discard your deck every play) is closer in strength to terminal silver than to chancellor.

However: all effects I listed under "neutral cycling" do not care whether you're about to reshuffle (and neither does fortune teller, except in the edge case that your opponent has no green in their deck): they're just as likely to cause a reshuffle than to get you closer to a reshuffle, so even that evens out. In particular: in addition to the 3 points Awaclus mentions about what Advisor does in addition to lab is
4) it gives a higher chance to reshuffle
This evens out 2) and 3). Sure, you can choose not to play your advisor, but usually this is not worth it (assuming you don't have sources of "good cycling" in your deck, such that your discard pile is full of crap).

[florrat]

but they are not good cards. on average, the cards in your (hand and in play) are above average, and the cards in your discard pile are below average, because you put bad cards there with sifting.

Which leads to another drawback of Chancellor when sifting is present: if you discard your deck, you'll see your the bad cards you sifted past earlier again. Really, I think that discarding your deck is not very good, and mostly not very noticable

[silverspawn]

 

However, this effect is very small (the chancellor effect is not that big, and discarding just 1 card from your deck is n times as small (where n is your deck size))

discarding one card is small. cycling itself is not. the problem with chancellor is not that it's a bad card, it's that it has to compeat with other terminal silvers, or more generally other action cards, and some of them are quite good.

an early catacombs cycles like crazy if you get the right draws, and it can absolutely decide a game.

[Awaclus]

but they are not good cards. on average, the cards in your (hand and in play) are above average, and the cards in your discard pile are below average, because you put bad cards there with sifting.

If you put bad cards there with sifting. Sifting isn't always present in every game, but buying good cards is.

[DG]

I'd agree that the article seems to be about sifting and selection rather than cycling as such. Cycling can be about playing your buys and gains sooner but it can also be about playing key cards more often. A moneylender that is 'cycled' into play more frequently can trash more copper. This sort of change to a deck can be more important as it can improve drawing even more with positive feedback.

Even in terms of sifting though, I think it is important to discuss types of cards. A gold will usually be a better card than a duchess but if you can choose your cards through sifting and cycling you might discard your gold and play a throne room on your duchess. There many more complex examples and the types of cards (or specific combos) are important.

[silverspawn]

but they are not good cards. on average, the cards in your (hand and in play) are above average, and the cards in your discard pile are below average, because you put bad cards there with sifting.

If you put bad cards there with sifting. Sifting isn't always present in every game, but buying good cards is.

yes, but even if there is zero sifting, the cards in your former discard pile are just as good as the cards in your former draw pile, and both of them are worse than the cards that you have in play. so why is it better to get these cards in your draw pile instead of discarding everything?

No, that's not right. If you have 10 cards in your draw pile and 7 in your discard and you draw your 11th card, the 7 cards in your discard are exactly where you want them to be: in your draw pile. If you cycle 6 cards more, there's a chance that they will be in your discard again or still in your draw pile.

[Awaclus]

yes, but even if there is zero sifting, the cards in your former discard pile are just as good as the cards in your former draw pile, and both of them are worse than the cards that you have in play.

One of those cards is the card you bought last turn.

[silverspawn]

yes, but even if there is zero sifting, the cards in your former discard pile are just as good as the cards in your former draw pile, and both of them are worse than the cards that you have in play.

One of those cards is the card you bought last turn.

so in other words: you are willing to have all cards in play, which are exclusively cards you bought, come delayed, in exchange for drawing the cards in your former discard pile faster, because there is one card that you bought last turn? that doesn't seem right.

assuming there is no sifting. if there is sifting, they are even worse.

[Awaclus]

so in other words: you are willing to have all cards in play, which are exclusively cards you bought

No, only one of the cards that miss the reshuffle because of this is a card you bought this reshuffle.

[silverspawn]

so in other words: you are willing to have all cards in play, which are exclusively cards you bought

No, only one of the cards that miss the reshuffle because of this is a card you bought this reshuffle.

well but it doesn't matter if you bought them this turn, the cards in play are usually engine parts, and you don't want engine parts to miss the reshuffle. in pretty much any game where you have any sort of draw the cards in play will be much better than the cards in your discard or drawing pile, becaue copper, estates, other victory cards and curses are never in play, and while your hand might be slighty worse than your discard pile, because the latter contains some freshly bought cards, I can't imagine that that's enough to make them better than the cards in play plus the cards on your hand.

if you have 10/6 draw pile/discard pile and you manage to draw these 10, forcing a reshuffle to have the discard pile become your new draw pile seems like a terrible move, unless you have enough steam left to draw your discard pile too, and in that case you will have everything in your discard pile anyway.

[Awaclus]

so in other words: you are willing to have all cards in play, which are exclusively cards you bought

No, only one of the cards that miss the reshuffle because of this is a card you bought this reshuffle.

well but it doesn't matter if you bought them this turn, the cards in play are usually engine parts, and you don't want engine parts to miss the reshuffle. in pretty much any game where you have any sort of draw the cards in play will be much better than the cards in your discard or drawing pile, becaue copper, estates, other victory cards and curses are never in play, and while your hand might be slighty worse than your discard pile, because the latter contains some freshly bought cards, I can't imagine that that's enough to make them better than the cards in play plus the cards on your hand.

if you have 10/6 draw pile/discard pile and you manage to draw these 10, forcing a reshuffle to have the discard pile become your new draw pile seems like a terrible move, unless you have enough steam left to draw your discard pile too, and in that case you will have everything in your discard pile anyway.

But if you can shuffle your discard pile into your deck and discard X random cards from your deck where X is the number of cards that were in your old discard pile, it's a good move.

[MarkowKette]

I think you should also mention that the good cycling effect gets reversed once you start greening. (and yes please dont throw all the edge cases there are for this statement at me now. I know there are plenty)

[c4master]

so in other words: you are willing to have all cards in play, which are exclusively cards you bought

No, only one of the cards that miss the reshuffle because of this is a card you bought this reshuffle.

well but it doesn't matter if you bought them this turn, the cards in play are usually engine parts, and you don't want engine parts to miss the reshuffle. in pretty much any game where you have any sort of draw the cards in play will be much better than the cards in your discard or drawing pile, becaue copper, estates, other victory cards and curses are never in play, and while your hand might be slighty worse than your discard pile, because the latter contains some freshly bought cards, I can't imagine that that's enough to make them better than the cards in play plus the cards on your hand.

if you have 10/6 draw pile/discard pile and you manage to draw these 10, forcing a reshuffle to have the discard pile become your new draw pile seems like a terrible move, unless you have enough steam left to draw your discard pile too, and in that case you will have everything in your discard pile anyway.

But if you can shuffle your discard pile into your deck and discard X random cards from your deck where X is the number of cards that were in your old discard pile, it's a good move.

I can't see a way to do this with a guarantee not to discard your good card(s).

In the beginning, cycling is obviously very good because you want to see your newly bought cards more often. But randomly cycling a 10/6 seperation into some other random 12/4 seems to have nearly no effect at all. Yes, there is a chance for your good card to be in the draw pile, but your cards in play AND the card you buy on this turn will take longer until they reach your hand.

Of course, if you are not doing a "random" thing like cycling*, but a more sophisticated sifting, things are different.

----

* Even cycling isn't totally random in Dominion. There is no card, that says: "Discard the top card(s) of your draw pile." Closest may be Chancelor whose cycling effect is positive, but marginal in many games.

[Awaclus]

I can't see a way to do this with a guarantee not to discard your good card(s).

In the beginning, cycling is obviously very good because you want to see your newly bought cards more often. But randomly cycling a 10/6 seperation into some other random 12/4 seems to have nearly no effect at all. Yes, there is a chance for your good card to be in the draw pile, but your cards in play AND the card you buy on this turn will take longer until they reach your hand.

Of course, if you are not doing a "random" thing like cycling*, but a more sophisticated sifting, things are different.

----

* Even cycling isn't totally random in Dominion. There is no card, that says: "Discard the top card(s) of your draw pile." Closest may be Chancelor whose cycling effect is positive, but marginal in many games.

You wouldn't be cycling a 10/6 separation into a 12/4 separation but another 10/6 separation.

You don't need a guarantee not to discard your good cards. The average card in your draw pile still improves on average if your discard pile contains better cards than your draw pile, which often is the case since the cards you just bought are there.

Also, Sea Hag when the Curses have run out is a card that says that. It is useful to know when it hurts your opponent and when it doesn't.

[c4master]

I can't see a way to do this with a guarantee not to discard your good card(s).

In the beginning, cycling is obviously very good because you want to see your newly bought cards more often. But randomly cycling a 10/6 seperation into some other random 12/4 seems to have nearly no effect at all. Yes, there is a chance for your good card to be in the draw pile, but your cards in play AND the card you buy on this turn will take longer until they reach your hand.

Of course, if you are not doing a "random" thing like cycling*, but a more sophisticated sifting, things are different.

----

* Even cycling isn't totally random in Dominion. There is no card, that says: "Discard the top card(s) of your draw pile." Closest may be Chancelor whose cycling effect is positive, but marginal in many games.

You wouldn't be cycling a 10/6 separation into a 12/4 separation but another 10/6 separation.

You don't need a guarantee not to discard your good cards. The average card in your draw pile still improves on average if your discard pile contains better cards than your draw pile, which often is the case since the cards you just bought are there.

Also, Sea Hag when the Curses have run out is a card that says that. It is useful to know when it hurts your opponent and when it doesn't.

There is no card that says: "Count the cards in your discard pile (x). Put your draw pile into your discard pile and reshuffle. Discard the top x cards of your draw pile."

I cannot see how this is relevant in Dominion.

Moreover, it must be worse than simple Chancellor since you cannot use your cards in play (and on your hand) until the next shuffle (no Schemes or Inns, please).

Btw: If I get a sea hag at all (which I try to avoid), and if the curses have run out, I am usually willing to feed the hag to any trashing card available. Getting rid of the hag should help me more than cycling one card every now and then could possibly disturb my opponent.

[Awaclus]

There is no card that says: "Count the cards in your discard pile (x). Put your draw pile into your discard pile and reshuffle. Discard the top x cards of your draw pile."

I cannot see how this is relevant in Dominion.

Moreover, it must be worse than simple Chancellor since you cannot use your cards in play (and on your hand) until the next shuffle (no Schemes or Inns, please).

Btw: If I get a sea hag at all (which I try to avoid), and if the curses have run out, I am usually willing to feed the hag to any trashing card available. Getting rid of the hag should help me more than cycling one card every now and then could possibly disturb my opponent.

It's relevant, because you can achieve the same result by playing enough "neutral cycling" cards.

Yes, you should trash the Hag most of the time, especially since before greening, the neutral cycling actually helps your opponent.

[c4master]

Actually only Harvest and maybe Minion are really neutral. I forgot this, Harvest is neutral (meaning: fully random) cycling in Dominion.

Rebuild, for example, discards an arbitrary number of non-green cards and those green cards you named and exactly one other green card, if there is at least one other green card left. I would say, it's slightly positive because it guarantees to discard at least one green card (given that you have one).

Anyways, it's tough to discard exactly the amount of cards you have in your draw pile plus the amount you had in your discard pile. With the given 10/6 example you could play 4 Harvests or 4 Minions, but that is really an edge case and surely only very minor benefit.

[Awaclus]

Actually only Harvest and maybe Minion are really neutral. I forgot this, Harvest is neutral (meaning: fully random) cycling in Dominion.

Rebuild, for example, discards an arbitrary number of non-green cards and those green cards you named and exactly one other green card, if there is at least one other green card left. I would say, it's slightly positive because it guarantees to discard at least one green card (given that you have one).

Anyways, it's tough to discard exactly the amount of cards you have in your draw pile plus the amount you had in your discard pile. With the given 10/6 example you could play 4 Harvests or 4 Minions, but that is really an edge case and surely only very minor benefit.

Rebuild has some positive cycling, but also a random amount of neutral cycling. And that exact number is not relevant, the point is that neutral cycling is good for you if your good cards are in your discard pile and triggering a reshuffle which increases your draw pile is bad for you if your good cards are in your hand, in play, or about to be bought later this turn, and that on average, in the early game, your good cards are in your discard pile or about to be bought later this turn.

[c4master]

But I would still prefer to only discard all my cards from the draw pile and NOT create a new draw pile in order to make my cards in play and my cards to be bought or gained this turn not miss the shuffle.

So, my opinion is: Given the 10/6 example, Chancellor (-->0/16) is better than randomly cycling 16 cards (--> 10/6). Is there anything wrong about that?

What I mean is: This neutral cycling is good in the beginning, but it can become bad if you have to create a new draw pile before the end of the turn.