Print

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

[sherwinpr]

When it comes to opening and early game strategy, one of the few things I'm still pretty unclear about is which cursing card to buy when there is more than one option.  The main cursers are:

Witch
Torturer
Sea Hag
Familiar
Mountebank
Young Witch

I think I rarely go for young witch when there's something better on the board (especially if the bane card is strong such as fishing village, ambassador, or even lighthouse -- because that means the lighthouse can defend you twice as often).  However when it comes to others, I'm not so sure. 

• Would you always take a Sea Hag on a 4/3 start when Witch and/or Mountebank are out there?
• When is the Potion -> Familiar track worth the investment with Witch and/or Mountebank in the mix?  Does this depend on the presence of other useful potion-cost cards such as Scrying Pool or Golem?
• Does the presence of other good cursers pretty much cause Torturer to be an overpriced Smithy, or are torturer chains still worth building?
• When it comes to Witch vs Mountebank, which will do the most damage?  I feel the Mountebank's +$2 is more useful early on than the Witch's +2 cards when there are so many weak cards (coppers and curses) floating around in the game (at least with no trashing), but Mountebank won't always hit, maybe letting the Witch to allow the curses to split 6/4 (in favor of the Witch player), and with a little luck 7/3, but even then the Mountebank is always useful as it can hand out coppers even when the curse stack is depleted.  Then again, I doubt Witch can be played as often as Mountebank, as the Witch player's deck has more weak cards in it.  Therefore, intuitively, it seems to me that Mountebank is the way to go, at least more often than not.
• What cards on the board favor certain cursers over others?  Perhaps Counting House makes Mountebank a liability, or Fishing Villages make Witch more appropriate than Familiars?

If there's an article or post on this topic already out there, please let me know, and I apologize for the redundancy.

[ARTjoMS]

MB just improves in 3-4p games, because opponent on your right might have played MB already and other opponents might have already discarded their curses. In 2p games witches somewhat somewhat have an edge, especially with 5-2 starting split and witch-nothing starting buy. Those coppers you get from MB are not that bad.

Torturer needs +actions, preferably village, but it will be too slow vs other curse attacks.

Familiar is a card that i recently have started to hate and think that it is too much luck based, it might be too slow vs other curse attacks, but not always.

Sea hag usually is a solid opener, but it feels to me like it increases variance quite a bit, not only your key card might be discarded from top, but also there is increased chance that one might not be able to buy a silver after opponents sea hag turn. There is quite a big difference between Curse -Copper-Copper-Copper-Estate and Curse-Copper-Copper-Estate-Estate.

[philosophyguy]

One of the disadvantages of Torturer is that its attack loses its power if the Curses run out, and that's more likely to happen if there are more cursing attacks on the board. So, that's another strike against Torturer on a multi-cursing board.

[Thisisnotasmile]

One of the disadvantages of Torturer is that its attack loses its power if the Curses run out, and that's more likely to happen if there are more cursing attacks on the board. So, that's another strike against Torturer on a multi-cursing board.

I'd still rather have a Torturer than a Witch, Sea Hag, Young Witch and in some situations Familiar and Mountebank after the Curses have run out. Most of the others are better at running out the curses though, but I'm not complaining if my opponent isn't taking Curses to my Torturers. Just means they'll be effective for longer.

[guided]

It depends on the board, but if I had to rank them:

1. Witch
2. Sea Hag
3. Familiar
4. Mountebank
852. Young Witch

If there's another plausible Cursing attack in play that people are buying, Torturer is terrible unless there are simply no other useful $5 cards to buy and you desperately need the +3 cards. If there are no Village-type cards in play, Torturer is still terrible, though there are boards where it can be decent to get one copy of it even without Villages. Without another cursing attack on the board, and with a Village available, Torturer is of course one of the very best cards in the game.

The top 4 are very close. If Witch and Sea Hag are both on the board and I have a 4/3, I might very well open with Sea Hag and look to get a Witch with my first $5 buy. If Mountebank and Witch are available I'll often buy Mountebank first, then get a Witch. If I have a 5/2 I'd probably take Mountebank over Witch pretty much every time. I rank Witch higher because I rarely want 2 Mountebanks (whereas I'll happily take 2 Witches), and I rank Sea Hag higher because it's quite strong with a 4/3 opening even against Mountebank.

I don't recall having had to choose between Familiar and Mountebank. That choice is extremely board-dependent. Any other use for a Potion would skew me toward Familiar, but I might still take Mountebank.

[kn1tt3r]

Young Witch really depends. With a useless Bane card like Chancellor, Embargo or such, it is most certainly better than Sea Hag. Otherwise, it's worse - but it still has some value once the curses are gone.

[DG]

There are quite of lot of 'depends' going on. I recently chose familiars ahead of a sea hag since there were king's courts in the kingdom and the familiars were the only +1 action card (not that I'd recommend doing it again). I've also seen games where a jester, potential curse giver, was a clearly better choice than buying a second witch.

You generally need to keep an eye on how you're going to increase your spending in a deck that's going to fill up with curses. This can be really important if there are good 5 or 6 cost cards that can turn round a low value deck (such as trading posts or vaults). Sea hags and familiars can be overbought leaving you with poor spending until the game is effectively over.

[guided]

I should say, Young Witch can be quite good if the bane card sucks and there are no other cursing attacks. If the bane card is a cantrip or Fishing Village or something, don't even bother.

[WanderingWinder]

I want to pop in and say that I'm always amused when the bane card is moat.
I also just won a game where I picked up YW over Familiar, but I wouldn't recommend this too often if the Bane is even mediocre; probably even a fortune teller is about good enough - in my case it was a workshop, not a lot of great 3-4s, so I think it was slightly the right call for me, but pretty close.

The best card of these to have is generally familiar, but it's steep cost to get is a little bit prohibitive. Mountebank would be next, except that its drawback is actually pretty significant, and making them gain coppers isn't as bad as you'd normally think. So witch is probably 2nd, though witch/mountebank is a little board dependent, and mountebank is obviously much better once the curses are out, at which point witch is somewhat worse than moat.
Torturer is extremely rarely a good idea with almost any other curse-giver out (exception being YW when the bane card is strong). Basically the attack is at best neutral once the curses run out, and paying so much for a smithy is not so great when you've got so many curses in your deck. There are also a decent number of situations where the attack actually helps opponents - ambassador and occasionally hunting party or vault before the curses run out, and library, watchtower, horse traders, menagerie or secret chamber afterwards.

[sherwinpr]

I think I would still go for Torturers (in addition to other cursers, perhaps) in University games.  Does this seem like a valid exception to the "don't get Torturers in the presence of other strong cursers" rule?

[jonts26]

I ran some simulations using Geronimoo's simulator for witch, mountebank, and sea hag.

The buy rules were big money ultimate looking to get two of the cursing cards in your deck. Witch and mountebank were prioritized over gold while sea hag was not. I simulated both 4/3 openings and 5/2 openings. Here are the results. Percentages are win rates.

4/3 Opening

Witch beats Mountebank 60% to 36%
Witch beats Sea Hag 52% to 44%
Sea Hag beats Mountebank 56% to 41%

5/2 Opening

Witch beats Mountebank 63% to 33%
Witch beats Sea Hag 81% to 17%
Mountebank beats Sea Hag 67% to 30%

Obviously the presence of other cards will affect what is actually best. But these data seem to suggest that you prioritize witch over the other two no matter the split, while you choose Sea Hag/Mountebank depending on the opening.

[WanderingWinder]

I'm not sure how specifically you implemented each of the strategies, but we already knew this, more or less. What's a bit more interesting is whether you should buy a hag and ramp up to mountebank or witch, or whether double silver in an attempt to rush to the witch is better.

[jonts26]

Yeah I realize that was the general consensus here, but I find it nice to have data to back up your intuition.

And I simulated your more interesting cases.

Opening Sea Hag/silver and getting a witch at first opportunity beats Silver/Silver and getting two witches 58% to 37%.

Opening Sea Hag/Silver and ramping up to Mountebank loses to just grabbing the second sea hag 44% to 52%.
Then I wondered whether it would be better, if you happened to draw $5 before getting the second Hag, to go ahead and buy Mountebank or buy down for the second Hag, ignoring Mountebank altogether. So I ran the buy rule of just getting 2 hags ASAP versus getting the first Hag and then buying a Mountebank with 5 or a second hag with 4. These two came out tied at 48% each.

[sherwinpr]

Thanks for running the sims, jonts26, it's good to have some data.

[drg]

It's always board dependent, although mountebank is more annoying early and I would always try to get that first.  If KC/TR are around, mountebank becomes even more potent, as the extra coppers are not helpful.  If you're trying to empty the curse pile, you need to get one of something else unless you are reliably playing mountebank more than once per turn, otherwise it could take a while.

If familiar is the only potion card and one of witch/sea hag is available, avoid the familiar.  I see so many people going for them while I get witches and the witch just wins.  Opening 2 silvers (or a terminal $2/silver) gives you a much better chance at a witch than a potion does at a familiar as it isn't relying on 1 card, you just need $5, and the money is useful after the 3rd time through your deck, while the potion isn't.  If there's another good potion card, by all means go for the potion/familiar.  Sure they end up with more familiars than I do witches... but that's because their only other choice is silver and 2 cursers is plenty, by the time you get 3 or 4, you're lucky if you can play them once each before curses are gone.

The trashing level of a board also matters, if it's good, you may choose not to open sea hag and get witches later, if it's none, it's almost surely right to buy a sea hag with a 4/3 opener.

As has been said, torturer is nullified by the presence of another cursing card, and shouldn't be taken into consideration.  It's a beast when there's no other cursing though.

If ambassador is present, you may want to avoid the cursing all together (especially sea hag, ambassador is just better than sea hag, I also avoid sea hag if masquerade is around) depending on how ambassador heavy your opponent goes.  Personally, I prefer ambassadors to cursing cards, as they give your opponents useless cards while removing your own.  If I'm winning the ambassador race, I will often buy a cursing attack later just to add to their misery as they have too much crap to return and can't keep up.

A lot depends on what +actions are available as well.  Witch is much better when they are around, and they help support more than 1 mountebank too.

[drg]

One of the disadvantages of Torturer is that its attack loses its power if the Curses run out, and that's more likely to happen if there are more cursing attacks on the board. So, that's another strike against Torturer on a multi-cursing board.

All the others except mountebank lose their attack ability when the curses run out too (mountebank loses about 60%-70 of it).  The thing about torturer is, when there's no other cursing, it often takes MUCH longer before the curses run out. (as long as there's no watchtower)  Torturer is actually usually the MOST useful of these cards once the curses are gone.  Sea hags have no value at all if there are no curses left.  Sea hags are much better cards when you can apprentice or salvage them later.

[theory]

FWIW, simulation data will definitely underrate Mountebank compared to Witch.  This is because Mountebank screws up combos much more than Witch does, but it doesn't screw up Big Money quite as bad.  So simulations will somewhat understate Mountebank's effect because they're only running Big Money.

[Superdad]

That is a very pertinent point Theory. Many of these simulation results are biased in the setup of the simulation (i.e. choosing the opponents to be big money variants will bias the results towards whatever beats big money variants).

That being said, in a curse-heavy game, I would be pretty apprehensive to build a combo engine - although that again obviously depends on the resilience of the combo deck to bloating. For example, something like Worker's Village/Peddler/Bridge will be completely useless when bloated, but something like Kings Court/Warehouse/Remodel could survive it?

/edit, I can't type today

[DG]

I think a good example would be that upgrades can sometimes play through the cursing attack from witches. It's that much harder to play through mountebanks using upgrades.

[Geronimoo]

Although it's true most simulations you see here involve Big Money variants, the simulator is perfectly suited for building complex engines. It's just harder to get the buy rules right and some cards can/should not be used because the simulator handles them poorly (like Remodel, Forge, Haven...).

For example: the built-in COMBO - Native Village/Bridge (clearly not a big money deck) will lose more often to Mountebank than it does to Witch for the reason theory suggested.

Another one: the built-in NGN - Drunk Marine Students (stupid name, but it involves University, Vineyard and Wharf and is certainly not big money because it doesn't buy any money). This one is equally resistant to both Mountebank and Witch.

....

[DStu]

Another one: the built-in NGN - Drunk Marine Students (stupid name, but it involves University, Vineyard and Wharf and is certainly not big money because it doesn't buy any money). This one is equally resistant to both Mountebank and Witch.

I like the name

[Atto]

Another question concerning the simulation data is: Are we sure we always want two cursing attacks?
I can imagine some situations, where i would prefer to buy only one attack card. For example if the setup includes an important trasher. I can't play three cards in a thin deck and if it is thin, my only attack can be played often. For me this seems to be better than not to trash and buy two attacks.

[DG]

Another question concerning the simulation data is: Are we sure we always want two cursing attacks?

Not at all. The second and third cursing cards have have far less impact in changing the division of the curses. Also in two player games you can assume that every curse you don't pass out will arrive in your deck instead, however in multi-player games this is far from true.

[Atto]

That's true. You can easily see this here.
The first Witch or Mountebank increase your win rate by 0.2, the second only by 0.1. Additional cards even decrease it.
Familiars give you about 0.12 for the first and second and 0.03 for the third. More than three are worse.
Sea Hag and Young Witch are only interesting in single numbers (+0.9 for Sea Hag and +0.6 for YW).

Of course this statistics are based on all games, so a good player may play differently.
Nevertheless i consider this data to be a good basis for designing a strategy. It seems to be clear that buying two Witches or Mountebanks is a successful strategy while buying two Sea Hags is not.
I don't know why it works in the simulations. I have to think about that.

[DStu]

That's true. You can easily see this here.

You can not really see it easily there, as this is the difference in gains compared to your opponent. So it is one more Witch/Hag than your opponent increases your win rate, which could also mean you have 5 and your op has 4.