Print

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

[terminalCopper]

I would prefer almost any other junking attack.

If so, you are probably underestimating the topdecking. For example, I think, in average, Sea Hag is stronger than a card saying "Each opponent gains a curse and discards down to three" would be.

A comparison to other junk attacks:

- I can hardly imagine a game where I would prefer Marauder or Young Witch, even without a bane.
- If Soothsayer/Witch/IGGs is on board, om 4/3, I might open Hag anyway, and get the "better version" at the next opportunity.
- if there are no other potion cards, I prefer to get two hags than a potion and three familiars.

The ones I agree are Mountebank and Cultist.
Cultist is so terrible that getting 5$ has absolute priority. Mountebank is imho slightly weaker, but it antisynergizes with hags cursing, so it's probably wrong to get both.

Man, I don't think that's more than 50% of all Sea Hag boards. Not even close.

 

You still actually have to buy it well over half the time, to avoid ending up in situation (1)

Let me try to provide an analysis as a starting point:

If I skip the Hag

95% in the presence of Lookout, masq, forager, amb, jack, mountebank, cultist (7 cards),

otherwise, 50% in the presence of hermit, remake, junk dealer, upgrade, Urchin, Trader, Lighthouse (7 cards),

otherwise, 20% in the presence of Soothsayer, Witch, familiar, Watchtower, Moat (5 cards),

and in 5% of the rest of all kingdoms,

according to this -barring edge cases with black market or young witch-

I will skip the hag in

0.95* (1-( (197 choose 9) / (204 choose 9))
+ 0.5* ( (197 choose 9 / 204 choose 9) )* (1- (190 choose 9 / 197 choose 9))
+ 0.2* (190 choose 9 / 204 choose 9) *  (1- (185 choose 9 / 190 choose 9))
+0.05 * (185 choose 9 / 204 choose 9)

=0.95*27.43+0.5*20.52+0.2*11.32+0.05*40.72= 40.6 % of all games <=>

buy at least one hag in 59.4 % of all games.

That makes me curious to know: where are the major disagreements?

[c4master]

Good point.

I would also skip Sea Hag on boards with Marauder or Young Witch, maybe Trading Post.

And on boards with Junk Dealer, Hermit, Upgrade or Steward I'm more likely than 50% to skip the Hag. But still, this would be roughly 50% of all boards with Hag, right?

So I guess, I had a wrong estimation of the frequency of these boards.

edit: And there's the opening. Yet, I'm not sure whether opening splits encourage or disencourage Sea Hag.

[silverspawn]

sea hag is kind of terrible at what it does, but what it does is dealing out curses, and that alone is enough to domninate quite a few boards. so, in terms of this list, it's still fairly high, but i refuse to call it a power card. if it's good, it's only good because the board is msising crucial elements.

I guess, if "good" means "above average among all 4$ cards" and if "above average" means "you buy it more often than a random other 4$ card", then it's a good card, slave to the defition.

[Tables]

Sea Hag is not 'kind of terrible' at what it does. It's kind of incredible at what it does, which is choking up the opponent, second to probably only Mountebank. The thing to argue against it compared to other cursers is that it gives you no benefit.

[Alexmf]

I would also skip Sea Hag on boards with Marauder or Young Witch, maybe Trading Post.

Why would you skip Sea Hag because of Marauder?

[c4master]

I would also skip Sea Hag on boards with Marauder or Young Witch, maybe Trading Post.

Why would you skip Sea Hag because of Marauder?

Marauder gives me a bonus (Spoils), Sea Hag doesn't.
Also, the Sea Hag player is not faster than the Marauder player, because Sea Hag discards one card.
So they are both at equal Speed except that the Marauder player has the  pretty good chance to hit $5 or $6 with his Spoils on turns 5-7 or so. Maybe I'm wrong, but my intuition says, that's usually better than the Hag.

----

edit:

Of course, Sea Hag could also skip Marauder or the Spoils in less then 20% of all cases. That would be really bad, espacially if the Marauder was skipped. But I'd go for the risk and be pretty save in more than 80% of these cases. On the other hand, Sea Hag could skip the curse which is very good for the Marauder player.
Spoils being a one-shot-gold is actually helping the Marauder player to play his Marauder more often. If it was a gold, he might be more weakened by not junking the opponent than strengthened by the gold. Again, I'm not sure about that.

[terminalCopper]

So, let's have a look at "Marauder vs. Hag".

If you play one marauder per turn, it gives you "+3$" at the expense of one random card each turn - starting with the next turn. Hag immediately discards a random card and replaces it with a useless curse (Ignoring the huge variance between discarding marauder or a curse, which probably helps the hag player by a margin).
Taking the different junk attacks into account, one might summarize the differences as follows:

1.) Marauder has +3 instead of two random cards

2.) Maurauder has slight speed downsides  (+3 is delayed by one turn, while one of the random cards is discarded right now)

3.) Marauder doesn't deal curses but ruins

To analyze these effects, consider two extreme cases, the very good and the very bad kingdom.

In a very strong kingdom, both players tend to have strong decks with an average above 1.5$ / card. Thus, point 1 is enough to prefer the hag.

In a very weak kingdom - example given, a two-card kingdom with nothing but the hag and the marauder - it's a little more complicated. Looking exclusively at point 1 favors marauder, as the average buying power might stay below 1/card during the whole game.
On the other hand, there's two big downsides.
The first one is that there won't be many shuffles, the spoils obtained in the last turn will be useless, and if e.g. 2 out of 10 spoils you gain throughout the game are dead, this decreases their value by 20% (or even more than 20% because of the missed compound interest dominion offers).
The second one is, that in a slog the difference between curses and ruins is BIG. On the one hand, ruins can provide small advantages; and more important, 1 VP in a slog is A LOT. Just compare duchess to monument ... or take into account the following example: If one player gets -10 VP, and he ends the game with 6 duchies, while his opponent grabs only 2 duchies and 3 estates, he will be down by one point anyway.

I agree that marauder's "+3 next turn" might have more economic impact than two missed random cards; and it is possible, that this economic advantage is even stronger than the downsides by
"-20% on spoils" and "curses never help, ruins do so occasionally". But all in all, I can hardly believe that the small remaining economic advantage for the marauder player compensates 10 VP.

In a middling kingdom, I can't fully guarantee that the interpolation between "very strong" and "very weak" kingdom works correctly in favor of Sea Hag; there might be some rare exceptions.

But in most scenarios, I think that the Hag is clearly better than Marauder.

[c4master]

To deal all the curses/ruins in this example, you would have to play like 30 turns or so. Apparently this doesn't happen very often. Even if it does, it would easily mean a lead of 1-2 Provinces for the Marauder player.

So, we have to rather compare kingdoms where you don't play just a junk and money slog. I just feel better about the economy provided by Marauder. You could buy such a nice $5 card as a Carthographer, for example. This will make a huge difference. Other examples are INN, Stables, Labs, or maybe even Vault or Hunting Party. Most of them are really crucial to get ASAP, either to play your Marauder more often or to get use of your junk.

It's not easy to evaluate, but I feel like hitting $5 or more is very important in many kingdoms.

[silverspawn]

So, let's have a look at "Marauder vs. Hag".

If you play one marauder per turn, it gives you "+3$" at the expense of one random card each turn - starting with the next turn. Hag immediately discards a random card and replaces it with a useless curse (Ignoring the huge variance between discarding marauder or a curse, which probably helps the hag player by a margin).
Taking the different junk attacks into account, one might summarize the differences as follows:

1.) Marauder has +3 instead of two random cards

2.) Maurauder has slight speed downsides  (+3 is delayed by one turn, while one of the random cards is discarded right now)

3.) Marauder doesn't deal curses but ruins

To analyze these effects, consider two extreme cases, the very good and the very bad kingdom.

In a very strong kingdom, both players tend to have strong decks with an average above 1.5$ / card. Thus, point 1 is enough to prefer the hag.

In a very weak kingdom - example given, a two-card kingdom with nothing but the hag and the marauder - it's a little more complicated. Looking exclusively at point 1 favors marauder, as the average buying power might stay below 1/card during the whole game.
On the other hand, there's two big downsides.
The first one is that there won't be many shuffles, the spoils obtained in the last turn will be useless, and if e.g. 2 out of 10 spoils you gain throughout the game are dead, this decreases their value by 20% (or even more than 20% because of the missed compound interest dominion offers).
The second one is, that in a slog the difference between curses and ruins is BIG. On the one hand, ruins can provide small advantages; and more important, 1 VP in a slog is A LOT. Just compare duchess to monument ... or take into account the following example: If one player gets -10 VP, and he ends the game with 6 duchies, while his opponent grabs only 2 duchies and 3 estates, he will be down by one point anyway.

I agree that marauder's "+3 next turn" might have more economic impact than two missed random cards; and it is possible, that this economic advantage is even stronger than the downsides by
"-20% on spoils" and "curses never help, ruins do so occasionally". But all in all, I can hardly believe that the small remaining economic advantage for the marauder player compensates 10 VP.

In a middling kingdom, I can't fully guarantee that the interpolation between "very strong" and "very weak" kingdom works correctly in favor of Sea Hag; there might be some rare exceptions.

But in most scenarios, I think that the Hag is clearly better than Marauder.

yea, it's a nice try, but not a complete abstraction of real games. firstly, the difference between ruins and curses can be anywhere between huge and non existent, depending on how you deal with them and how much usage you can take out of the ruins. but more importantly, the "2 random cards for $3" thing just doesn't work; with marauder you get to 5$ faster and then things become way more complicated

[terminalCopper]

yea, it's a nice try,

Thx

but not a complete abstraction of real games.

Of course.

firstly, the difference between ruins and curses can be anywhere between huge and non existent, depending on how you deal with them

Yeah, that's the huge question. Whenever it's easy to get rid of curses, Hag is clearly worse than Marauder. In other words:

So, we have to rather compare kingdoms where you don't play just a junk and money slog. I just feel better about the economy provided by Marauder.

If you have Hag and Marauder, but no slog, there has to be strong trashing. Given this, I agree, that Marauder is the better card.

However, I was analysing a two-card-kingdom, or more general, a kingdom without trashing, where Hag and Marauder are the dominant cards. Usually, I will buy two hags asap, and if I don't get mirrored, throw in a third one later to make sure I give you all the curses. I suppose if someone prefers Marauder he will likely do the same.
With such a brutal junking performance, both ruins and curses will be out before T20, and both players will go for the duchies much earlier.

2 random cards for $3" thing just doesn't work; with marauder you get to 5$ faster and then things become way more complicated
 

Yeah, "2 random cards for $3" is just approximating the average, not the variance. I agree that the Marauder player has the higher variance and will yield $5 more often; but on the other hand, hag is excellent in reducing an opponents hand to $2. If you doubt about that just look at

http://forum.dominionstrategy.com/index.php?topic=9913.0
where Robz produces 2$ for 15 turns in a row ...

So, I'd like to show that in a slog, the increased chance of Marauder to get to $5 won't matter a lot, for the following reasons:

- it won't be useful that soon.
Say, after T4 you have gained two silvers, two Marauders, a spoil and a curse. You will use the curse in T5-T7 to get a $5-card, and you will use it for the first time in T8-T11...
if nothing goes wrong. If the spoil or the Marauder misses the shuffle or gets discarded, if the spoil shows up with less than two coppers (not that uncommon with 3 estates, 2 marauders, one curse and a 4-card-hand after being hagged), your first play will be delayed to T12 or higher. Every shuffle takes a long time, briefly: You can't swap the game because of a $5-card you will use three times.

-the chance to hit $5 is not THAT much higher
Example given, the hag player that gains one ruin, 2 hags and 2 silvers in T1-T4 will have a total of  $11 for T5-T7. It sounds way better to have a spoil leading to §14 (for 16 cards); but during T3-T7 you will suffer from three hag attacks. There's so many ways it can damage your economy, but let me just pick out a seemingly harmless examle: If I replace a copper in T3 by a curse, what will that do? If your hand goes down from $5 to $4, it sucks. If it goes down from $3 to $2, you don't get that silver. If you have $3 instead of $4 you might not get your second marauder ...

- the Hag player will get $3 more reliably
The simple reason is that he will never play a 4card-hand. But there's some other little helping friends, namely abandoned mine, ruined library, survivors or even ruined market (for a free copper).
Well, if you get to 5$ on one hand, but only 2$ on the other, this might be fine in an engine. But in a slog, having 2 silvers is often stronger than a 5$ card, and if it's weaker, not by that much.

I agree that if it wasn't for the curse/ruins thing, Marauder would be the better card. But in a slog, even if there remains a small economic advantage, i am sure that it won't be enough to compensate -10 VP.
 

[c4master]

@terminalCopper:
1) The example-game has been discussed and I think Robz has made some mistakes (too less coins, too many action cards). Plus, what really hurt was the Lookout that only trashed good cards (copper) instead of bad ones. With a marauder, this would have been no big deal, since you would have some economy due to the spoils. So I guess, this example shows nothing about Sea Hag.

2)Let's assume a slog like you did. I'm not quite sure about that second marauder, but I guess it will help. By turn 5 you have 11 coppers, 2 silver and a spoils totalling $19 of buying power in your deck. Your deck probably has 16 cards, maybe 17 if the Hag player got a bit lucky. That's more than $1 per card! I really don't want to calculate the variance here, but getting $5 or more within the next 3 turns should be more than likely at least two times. I really can't believe that getting two $5 cards by turn 7-9 won't change A LOT compared to two silvers (also the hag player might stall into $2 hands...but let's not assume this). For convenience, just name five $5 action cards that won't change this slog game enough, if you get two and your opponent none.

3) You will need $5 hands (or better) to win the game unless there is $3 or $4 VP cards. Of course, the hag player could aggressively buy estates to end the game befor turn 20. But then, the Marauder player could just stop playing his marauders when there are 4 ruins left. These extra turns will probably be enough to spike two provinces later on. Even worse: the hag player probably needs 3 hags to deal all 10 curses before turn 20 if there is no support.

With any kind of sifting, the marauder will do even better: Discarding real crap (curses) into some maybe really good cards (spoils) is much better than discarding mostly crap (ruins) into some decent cards like silver.

[terminalCopper]

By turn 5 you have 11 coppers, 2 silver and a spoils totalling $19 of buying power in your deck. Your deck probably has 16 cards, maybe 17 if the Hag player got a bit lucky. That's more than $1 per card! I really don't want to calculate the variance here, but getting $5 or more within the next 3 turns should be more than likely at least two times. I really can't believe that getting two $5 cards by turn 7-9 won't change A LOT compared to two silvers (also the hag player might stall into $2 hands...but let's not assume this). For convenience, just name five $5 action cards that won't change this slog game enough, if you get two and your opponent none.

I don't get that ... ?!

[c4master]

By turn 5 you have 11 coppers, 2 silver and a spoils totalling $19 of buying power in your deck. Your deck probably has 16 cards, maybe 17 if the Hag player got a bit lucky. That's more than $1 per card! I really don't want to calculate the variance here, but getting $5 or more within the next 3 turns should be more than likely at least two times. I really can't believe that getting two $5 cards by turn 7-9 won't change A LOT compared to two silvers (also the hag player might stall into $2 hands...but let's not assume this). For convenience, just name five $5 action cards that won't change this slog game enough, if you get two and your opponent none.

I don't get that ... ?!

Math. Let's do it. Well, I just miscalculated. But anyways, it's $18 which is still more than $1 per card.

Btw. There are more than five $5 cards that won't change a lot (e.g. Stash), but still the majority does change things.

[shark_bait]

Both Marauder and Sea Hag are good cards.  In fact, if I were to play a game with both Marauder and Sea Hag I would actually want both of them!  Why both?  If you ignore 1 of them you are likely to be exceptionally more bogged down in junk than your opponent.  With 2 piles then empty your opponent is in a significantly better position to opportunely 3-pile.

In this type of game, I open Marauder.  Why, I want to get that Spoils ASAP.  In most cases it will lead to a power $5 or Gold.  Either of those will help your future buys be better than Silver.  In turns 3 or 4, I get a Hag.  You can't ignore cursing because the -VP is very significant.  If you get it later than turns 3 or 4 you look to lose the curse split quite significantly (assuming opponent opened Marauder).

Marauder and Sea Hag are functionally different in what they bring to a Kingdom but when together they both function to provide endgame 3-pile control.  Marauder both Junks and gives you a spike in buying power.  Sea Hag does nothing to help you but gives your opponent -VP.  If you ignore 1 of them you will probably lose.  Even though they are both $4-cost junkers they are significantly different in what they bring to a deck.  To ignore one would be perilous.

[HiveMindEmulator]

Seems like people are arguing 2 different things. The original statement that started this discussion was that you should skip Hag when Marauder is on the board. This is not the same as Marauder > Hag. I think Marauder is probably better than Hag, but alone is not a reason to skip Hag.

[terminalCopper]

By turn 5 you have 11 coppers, 2 silver and a spoils totalling $19 of buying power in your deck. Your deck probably has 16 cards, maybe 17 if the Hag player got a bit lucky. That's more than $1 per card! I really don't want to calculate the variance here, but getting $5 or more within the next 3 turns should be more than likely at least two times. I really can't believe that getting two $5 cards by turn 7-9 won't change A LOT compared to two silvers (also the hag player might stall into $2 hands...but let's not assume this). For convenience, just name five $5 action cards that won't change this slog game enough, if you get two and your opponent none.

I don't get that ... ?!

Math. Let's do it. Well, I just miscalculated. But anyways, it's $18 which is still more than $1 per card.

Your 1$-error irritated me less than the idea to have 11 coppers at T5. I will probably have to ask Celestial Chameleon how that could happen ... ?!

I want to get that Spoils ASAP.  In most cases it will lead to a power $5 or Gold. 

I know there has been previous discussions about Marauder being slow or not ... as I could not believe that there's a meaningful economic difference between marauding and hagging early on, I went for some calculations again:

The chance on hitting $5 till T3/4


a) Assume that P1 opens marauder/silver and P2 opens hag/silver. Obviously, on T3/4 the economics are completely symmetric.
b) Assume the same with P1 being the hag player. Now, there's a 5/12 chance that a card P2 wants to play will be replaced by a curse, which disables gaining a 5$, if

- a silver is discarded, and there are 3 coppers (probability: 1/12 * [   (7 choose 3)* (4 choose 1) ] / (11 choose 4)) = 0.03535
- a silver is discarded, and there are 4 coppers (probability: 1/12 * [   (7 choose 4)* (4 choose 0) ] / (11 choose 4)) =  0.0088
- a copper is discarded, and there are 4 coppers (probability: 7/12 * [  (6 choose 4)* (5 choose 0) ] / (11 choose 4)) = 0.0265
- a copper is discarded, and there is silver +2 coppers (probability: 7/12 * [  (6 choose 2)* (4 choose 1) ] / (11 choose 4)) = 0.106

all in all, it's 5/12 *(0.03535+0.0088+0.0265+0.106) =0.0736 =>
the chance of P1 to disable gaining a $5 on T3 with the hag is 7.36%. In other words, if P1 goes for the hag, his chance to hit $5 on T3/4 is 7.36% higher.

The chance of hitting $5 before T8 (after the second reshuffle)

It's way too complicated to wade through all the imaginable outcomes of T1-T4, so, for the sake of simplicity, let's say that no hag or marauder will collide or be discarded, and one player gained 2hags + 2silvers + a ruin, the other gained 2 marauders + 2silvers + a curse+ a spoil.
That is a total economy
for the hag player: 11$ in 15 cards = 0.733.. per card
for the marauder player: 14$ in 16 cards = 0.875 per card.

Sounds like a huge difference, right?

Well, stopping the analysis for T3/4 right here would be almost as wrong as analysing militia without the attack. Once again, the discard attack should not be mitigated... if the hags don't collide  in 3 turns(we've made that assumption for the marauders as well, so that's quite fair), 2 out of 15 cards will be discarded. That means that every card has a chance to "survive" of roughly 13/15, and so, the marauder player will rather face an average buying power of 0.875 * 13/15 = 0.75833.. per card. Yeah, that's a little more than the hag player has - namely 3.79.. vs. 3.666.. coins per hand...

but we haven't finished yet. By now, we have only taken into account the risk of destroying a $5-purchase at T3/4, not the  risk of discarding the marauder in the first shuffle (not huge, but 8.33% is more than 0). If a hag on T3 cuts down a hand from $3 to $2, calculations would have worked differently as well ... and we also didn't look at the small advantages abandoned mine or survivors might provide (even ruined library could help, if it's worth to take the risk). I am not really sure how well the sum of these subtle differences compete the difference between 3.79.. vs. 3.666. If you want me to guess,  the risk of a discarded marauder on its own is more than enough economic compensation for the small difference in average buying power without these subtleties.

Conclusion:
Before T8, two marauders don't have better buying power against two hags. If P1 has the hags, he is even more likely to hit $5 before T8. In a nutshell: Marauder's benefit is slow, Sea Hags topdecking is immediate.

After the third reshuffle

Finally, if no marauder was discarded, the marauder player will start this round with one more spoil. This adds +3$ at the expense of one slot, so it might add +2$ to the hand. So yeah, this might give a $5-card instead of a silver...

Just remember, there's very heavy junking, about 22 cards have been shuffled, and before the next shuffle, it's going to be like 27....thus, before the game ends, there might be only 2 or 3 occasions to benefit from that spoil; and only 1 or 2 occasions to play a $5 instead of a silver. That won't compensate -10 VP, regardless how strong the 5$-card is.

3) You will need $5 hands (or better) to win the game

Of course, the hag player is able to get $5 on its own. Example given, as mentioned, we have $11 in 15 cards, and the only way to distribute 11 coins among 3 hands without getting a 5$ is to split them 4$-4$-3$. Hitting 5$ gets even more probable with every shuffle.

  In fact, if I were to play a game with both Marauder and Sea Hag I would actually want both of them!

You might be right that getting one of both is the best strategy. And you are definitely right, that if both players open hag, the next thing to purchase is a marauder (and vice versa). However, the original discussion arose from c4masters thesis that one should skip Sea Hag on (most) boards with marauder.

[c4master]

@terminalCopper: Your calculations are pretty well and they nearly convinved me...
I am still missing the high variance vs. low variance thing. And the Hag player being P1 makes a significant difference. If it was P2, things would change in favor of P1 again.

I do accept your point: The game might end on a three-pile when the Marauder player has still not enough points to catch up.

You should absolutely be aware of that.

[shark_bait]

.... lots of math

For Marauder, it's more about Variance and less about a measely $0.1/card higher than Hag.  The entirety of that difference is in a single Spoils.  Yeah, the average isn't much higher but you can be sure that whatever hand has the Spoils will be better than any hand in the Sea Hag's deck.

And yes, I actually didn't read through everything.  I skimmed and saw tons of discussion on Marauder/Sea Hag stuff.  So I decided to just chime in.

[c4master]

Now that I've got the "thyme" to answer not as short as last time, I'll explain a little bit.

First, my error counting the $s was a result of this: You've got 7 Coppers and two silvers, that's $11. I actually counted the silvers twice by thinking these $11 would all be Coppers. Well, and after that, I just added $8 instaed of $7 for two silvers and a spoils.

Second, I'm not sure whether I would get a second Marauder over a third silver. Nor is it clear, that I want that silver instead of any $2 or $3 action card. I think, a Marauder player is more likely to do well with a sifting card such as Cellar or Warehouse than a Hag player. So strictly saying, both players get 2 attack cards and two silvers is surely not the best idea.
the Hag player really needs to buy at least two silver in order to get his economy running. The Marauder player doesn't necessarily have to do so. So, if there is cool stuff at a level of $3 or $2, the Marauder player might be favorised. If there is no easy card, I want to get, hitting $5 is still easier for the Marauder because then he buys only one Marauder and 3 silvers  netting him roughly 90% chance to hit $5 at least once before turn 8.

Third, why is the Hag player going first? Most of it's advantage is due to first player advantage, that he would get no matter what P2 does. I don't believe that mirroring would be a good idea as it basically settles this advantage.

As shark bait pointed out, there is still the comparison of high variance vs. low variance.

[terminalCopper]

Second, I'm not sure whether I would get a second Marauder over a third silver.

Almost in any scenario, you should. Falling back in the junking race voluntarily sounds very weak.

So strictly saying, both players get 2 attack cards and two silvers is surely not the best idea.

Well, I just picked the most common example...

the Hag player really needs to buy at least two silver in order to get his economy running. The Marauder player doesn't necessarily have to do so. So, if there is cool stuff at a level of $3 or $2, the Marauder player might be favorised. If there is no easy card, I want to get, hitting $5 is still easier for the Marauder because then he buys only one Marauder and 3 silvers  netting him roughly 90% chance to hit $5 at least once before turn 8.

Almost any of the 203 other cards changes smth in favor of marauder or hag, but where's the significant relevance?

Third, why is the Hag player going first?

He does so in every second game. If the Marauder player goes first, there is just no difference in the chance of hitting 5$, I said so (see a)).

As shark bait pointed out, there is still the comparison of high variance vs. low variance.

Yeah, but I am almost sure, that doesn't change a lot ... maybe I'll explain that later.

I don't want to calculate the exact chance of hitting 5 neither for marauder nor the hag, that's way too complicated. Though, I am persuaded that variance won't help the marauder player much, let me try to construct some kind of upper bound:

i) The overall chance to hit $5 is very similar

I guess that before T8 the odds...
for the marauder player to get $5 (at least once) are close to 85%, at most 90%.
for the hag player to get $5 (at least once) are close to 80%.

I am almost sure that the difference is lower than 10%.

ii) If you agree with my argumentation that the average economy is pretty similar before T8, a huge variance would mean both more strong and more weak hands

Even if the difference wasn't 10% but 15%, it would have a reason, namely some other bad moves the marauder player has to expect. This includes the disastrous 1/12 chance that the marauder gets discarded before the first play, but the most common accident will be a $2-hand. Thus, for every time the marauder player gets a $5 instead of a silver, he will often get a weak card instead of a silver in another hand.

iii) If so: in a slog, two silvers is a brilliant thing

Now, let's stick to the expected difference of 15%, and let's say that in these cases, the Marauder player gets 5-2, while the hag player gets 4-3. Will that change the game? Usually, it won't:
4-3 provides at least two silvers, which is great in a slog; to get advantages from bigger variance,
there has to be strong 2's and 5's, but no strong 4's and 3's.
If so, you will probably use that advantage only up to three times; compound interest is virtually negligeable in a slog.
Now multiply these advantages by an optimistic three and  an exaggerated  0.15, and even then, the resulting advantage won't be impressive:

0.45 * (strength of two cards costing 5$/2$ - strength of two cards costing 4$/$3).

That's either negative or almost nothing.

PS: There's also a kind of meta-argument that makes me so certain about the hag's strength in this scenario. Thinking about good game design, comparing two 4$-cards, there has to be some moments where the one outshines the other and vice versa. Now, whenever the junk can be trashed, marauder is better than the hag; example given, in the presence of forager, upgrade, junk dealer ...
So, there has to be some other moments where the hag outperforms marauder - and i really wonder:

Where do you expect the hag to be stronger than marauder, if not in a game without trashing?

[c4master]

Second, I'm not sure whether I would get a second Marauder over a third silver.

Almost in any scenario, you should. Falling back in the junking race voluntarily sounds very weak.

Even if the Hag player archieves 2 attacks per shuffle, he needs 5 shuffles (plus two initial ones to get hte hags) to deal all the curses. That's 7 shuffles. Without trashing or sifting that is quite impossible. Assuming an average of 25 cards (10 starting cards, 4-5 junks, 2 sea hags, rest silver and other bought cards) that would be 7*25/5=35 turns.
So, I'm pointing out, that without any kind of sifting or trashing, you would need to much time, to deal all the curses. The difference between dealing 4 curses and dealing 7 curses is not that amazing, I think, especially, since your not dealing them more quickly with 3 hags than with 2 and only slightly more quickly with 2 than with only 1.
I've played many games, where I found myself buying one junk card, but not the second. Even if I lose the split 7-3, it might be worth it not haven a dead card in your deck.
This said, I might be tempted to get a hag after the Marauder, just in order to not get all the curses, if that is possible in the first place (say, there is sifting by Cellar).

So strictly saying, both players get 2 attack cards and two silvers is surely not the best idea.

Well, I just picked the most common example...

OK. It's most common, but it might not always be correct play.

the Hag player really needs to buy at least two silver in order to get his economy running. The Marauder player doesn't necessarily have to do so. So, if there is cool stuff at a level of $3 or $2, the Marauder player might be favorised. If there is no easy card, I want to get, hitting $5 is still easier for the Marauder because then he buys only one Marauder and 3 silvers  netting him roughly 90% chance to hit $5 at least once before turn 8.

Almost any of the 203 other cards changes smth in favor of marauder or hag, but where's the significant relevance?

You do not play a 2-card-kingdom. You play with 10 (or 11) kingdom cards and I believe, it's hard to find a combination of cards, where both, Hag and Marauder are present and I would not prefer Marauder over Hag.
My point here is: Marauder is more flexible and allows more combination with other kingdom cards than hag does - in general. This is because a Sea Hag doesn't provide any benefit, espacially no economy.

Third, why is the Hag player going first?

He does so in every second game. If the Marauder player goes first, there is just no difference in the chance of hitting 5$, I said so (see a)).

Ok, I got this wrong.
Still, if the second player opened Hag instead of Marauder, there would be no better chance for him to hit $5, but a worse!
If you're trying to explain: "As P1 I would rather get Sea Hag than Marauder because it hurts P2 so much.", I'm pretty fine with that. I believe by now that P1 has kind of an equally well choice - or nearly as good as Marauder at least. Thank you for making this clear.

As shark bait pointed out, there is still the comparison of high variance vs. low variance.

Yeah, but I am almost sure, that doesn't change a lot ... maybe I'll explain that later.

I don't want to calculate the exact chance of hitting 5 neither for marauder nor the hag, that's way too complicated. Though, I am persuaded that variance won't help the marauder player much, let me try to construct some kind of upper bound:

i) The overall chance to hit $5 is very similar

I guess that before T8 the odds...
for the marauder player to get $5 (at least once) are close to 85%, at most 90%.
for the hag player to get $5 (at least once) are close to 80%.

I am almost sure that the difference is lower than 10%.

"At least once" is not the same as "on average". By the way, to counter this:
- Whenever you have a Hag in your hand, you need either two Silvers and at least one Copper or one Silver and three Coppers to hit $5. So, chances are pretty high, that these hands don't provide $5.
- You could also look it at that way: Whenever I have a Spoils in my hand, my chance of hitting $5 is roughly 90-95%. Without a Spoils, my chances are as good as those from the Hag player if I was not attacked the turn before and slightly lower if I have been attacked.
- Getting a good card early can really change anything. I'd say those cards will be important: Laboratory, Library, Witch, Adventurer (Spoils are treasures!), Minion, Trading Post, Upgrade, Ghost Ship, Merchant Ship. Tactician, Wharf, Mountebank, Rabble, Vault, Venture, Goons, Hoard, Hunting Party, Carthographer, Embassy, IGG, Inn, Margrave, Stables, Border Village, Farmland (for trashing and province spiking), Catacombs, Count, Counterfeit, Cultist, Junk Dealer, Rebuild, Altar, some Knights, Baker, Butcher, Journeyman, Merchant Guild and Governor. And these are only those cards, that obviously change things.

ii) If you agree with my argumentation that the average economy is pretty similar before T8, a huge variance would mean both more strong and more weak hands

Even if the difference wasn't 10% but 15%, it would have a reason, namely some other bad moves the marauder player has to expect. This includes the disastrous 1/12 chance that the marauder gets discarded before the first play, but the most common accident will be a $2-hand. Thus, for every time the marauder player gets a $5 instead of a silver, he will often get a weak card instead of a silver in another hand.

While a Cellar is not such a strong card for the Hag, it's awesome for the Maurader. Marauder/Embargo embargoing silver seems nice. Marauder/Duchess is a soft counter for this hag stuff. A few Candlestick Makers don't hurt. Squire, Beggar, even Poor House all seem fine. There's hardly any $2 card I wouldn't get if I missed $3 with a Marauder.

iii) If so: in a slog, two silvers is a brilliant thing

Now, let's stick to the expected difference of 15%, and let's say that in these cases, the Marauder player gets 5-2, while the hag player gets 4-3. Will that change the game? Usually, it won't:
4-3 provides at least two silvers, which is great in a slog; to get advantages from bigger variance,
there has to be strong 2's and 5's, but no strong 4's and 3's.
If so, you will probably use that advantage only up to three times;

That's wrong. Just assume Stables, Carthographer or Adventurer.
And also, 15% is wrong. You'll get both, $5 cards earlier and more of them.

compound interest is virtually negligeable in a slog.
Now multiply these advantages by an optimistic three and  an exaggerated  0.15, and even then, the resulting advantage won't be impressive:

0.45 * (strength of two cards costing 5$/2$ - strength of two cards costing 4$/$3).

That's either negative or almost nothing.

I don't think so. You're doing this wrong.

PS: There's also a kind of meta-argument that makes me so certain about the hag's strength in this scenario. Thinking about good game design, comparing two 4$-cards, there has to be some moments where the one outshines the other and vice versa. Now, whenever the junk can be trashed, marauder is better than the hag; example given, in the presence of forager, upgrade, junk dealer ...
So, there has to be some other moments where the hag outperforms marauder - and i really wonder:

Where do you expect the hag to be stronger than marauder, if not in a game without trashing?

When is Scout better than Sea Hag? When is Thief better than Marauder? When is Thief better than Noble Brigand?

It was really cool, if these arguments held, but I fear, some cards shine only very rarely.

The answer is: I expect Sea Hag to be almost anytime minor to Marauder. That's just what I try to explain the whole time.

btw: I do have changed my mind a bit: If there is no junking and no other cursing I might pick up a Sea Hag on the second shuffle. Maybe.

[terminalCopper]

@c4master:
Basically, I am saying that you wildly overestimate both the economic difference between +spoil and the topdecking and the impact of a $5 in a slog with multiple junking and no trashing. If you want, don't believe me.

There's always complex details, and so, I can't prove my thesis. The only way of using mathematical approximations in a useful way is to have a kind of common sense which are the important aspects and which aren't. Obviously, we are far from having a common sense about variances, 10 VP's, compound interest in slogs, the values of $5's compared to silver in slogs, and maybe dominion in general...
So please, don't feel offended: but this is one of the cases where I believe that following the discussion is no longer fruitful for the community (probably it already isn't, we just didn't see it yet). I will stop the strategical argumentation about hag-vs-marauder-without-trashing right here, but maybe the meta-argumentation might make you wonder a little:

I expect Sea Hag to be almost anytime minor to Marauder.

It happens that I'm wrong, yeah.
It also happens, that community votings are wrong, that's true.
Theoretically, a community might also vote a card on No.1 that is "almost anytime minor" to another.
It might also happen, that Donald X. designs a card that is too weak compared to another...
And theoretically, it is possible that this underpowered card gets voted on No.1 by a stupid community.

So, either all of these five things happened, or you are wrong.

If you think it's the former, your self-confidence is truly enviable ...

[silverspawn]

#1 It happens that I'm wrong, yeah.
#2 It also happens, that community votings are wrong, that's true.
#3 Theoretically, a community might also vote a card on No.1 that is "almost anytime minor" to another.
#4 It might also happen, that Donald X. designs a card that is too weak compared to another...
#5 And theoretically, it is possible that this underpowered card gets voted on No.1 by a stupid community.

#5 being weaker than another card doesn't make that card underpowered at all. you rarely ever pick journeyman when there's wharf, but journeyman is still really good.
#4 that happens all the time... what about scavanger and chancellor? what about pretty much any terminal silver and swindler? any terminal draw and wharf? any 5$ junker and mount?
#2 that also happens all the time. what about rebuild on the last list?

I don't like these kind of lists, because they make it sound llike your counerpart has to admit that he's wrong unless he wants to be a complete outsider. i peronally don't think that hag is that much worse than marauder, but i don't see why the opinion isn't legit. #3 is a fair point, but also not that hard to believe. procession is... #35 on this list, and i think it's a lot better than sea hag, which is #1, even though they're very different cards. and if that's not enough, WW has voted quarry over sea hag, which is higher than procession, but still lower than marauder

[terminalCopper]

I don't like these kind of lists, because they make it sound llike your counerpart has to admit that he's wrong unless he wants to be a complete outsider.

Uhh, that's not my intention. If it sounds like that, I'm sorry.

What I'd like to emphasize is, that there is a huge difference between being weaker than another card and being (almost) dominated. Example given, it's perfectly legit to think that Hermit is better than masquerade in general/in the majority of all games. But if someone says that hermit is superior to masq in almost every game, I will question myself whether the person has fully understood masq's abilities.

Only few cards face are almost dominated by a second one - you mentioned some good examples. However, I am almost sure that no Top 10-card is ever close to be dominated by another card. Sure, my opinions are wrong in many situations, and Qvist rankings make errors as well- but I think that cards which are almost dominated like thief, scout or chancellor don't appear in the top spots of the rankings. At least, I think that this is very unlikely; that's what my list wants to say, just my two cents.

[c4master]

Hey, I really didn't want to offend anyone, nor did I intend to have such an intense discussion about two cards which are both strong ones.

Let's just say, we have different opinions about these cards and none can prove the other to be false, nor the own opinion to be correct. If it's possible to create a kingdom with only two fixed cards, I might playtest it against the bots just for myself to figure out a bit.

Anyways, I think these lists have fulfilled their task in making us think and discuss about certain cards. So, thanks again, Qvist, for creating them.