The following guest article is written by DG.
Baron, Caravan, Coppersmith, Council Room, Haven, Masquerade, Remodel, Smithy, Thief, and Throne Room
(Click for enlarged link at dominiondeck.com)
I decided to write an article about this game since it’s short and neat, there are some nice card combinations that I’ll try to explain, and it’s not too complicated. I’m also glad to say that my opponent played a good game that does him justice too. This was a friendly game, one of three very good games I played against Yaron in succession. Here is the isotropic log for the game log.
Yaron’s strategy
I’ll talk about Yaron’s game first since it’s both good and simple. Use the Masquerade to draw treasures to buy gold, removing the least valuable cards from the deck at the same time. Once there’s enough gold in the deck just buy provinces. This shows the power of the Masquerade as a starting card to quickly cycle a deck with extra spending. The Caravans simply add more of the same. He plays the Masquerade 5 times in 12 turns while buying 4 provinces. Nice play! Dominion doesn’t have to be complicated.
My strategy
Editor’s note: This is a great example of reading the board and coming up with a thorough plan..
When I chose my strategy the key card was the Remodel. It may sound obvious but this was based on the presence of good kingdom cards for the estates and coppers to be remodelled into. Remodel tends to work best along the 2-4-6-8 cost cards and this is how my deck would have worked depending upon the draws: copper to haven, estate to Caravan/Coppersmith, coppersmith to duchy/gold, gold to province. I had this map already in mind when sizing up the kingdom. The presence of a handy 2 cost haven in the kingdom was influential even though I didn’t actually use it much in this game. Draws like 4 Coppers and a remodel will turn up and if there’s nothing useful you can do then it is bad planning as well as bad luck.
The Caravans were an excellent partner to the Remodel. First of all they only cost four so I didn’t need much treasure in my deck to buy them, which was a good job since remodel decks don’t start out very rich. Caravans are not terminal actions. Remodelling them from the Estates was a double win for me since I removed a poor card that clogged up my deck and added something which sped up my deck too. Speed was important since it can often be a long chain of turns to get the benefits from remodelled cards. Play Remodel and get Caravan – shuffle – play Caravan – (maybe shuffle) – get extra card next turn and buy Gold – shuffle – play Gold to get Province. This doesn’t seem so bad for the first remodelled card but for the second and third cards we’re getting towards the end of the game. In this game I managed to play the remodel 5 times in 12 turns, which shows again how fast the deck cycled.
Another advantage of Remodels with Caravans is the enlarged hand size. For all actions where you choose cards from hand, having a bigger hand gives you more/better choices. In this game I was able to Remodel all my Estates into Caravans quickly. This could have been luck but it could have been due to big Caravan-sized hands giving me more choice. If anyone has wondered why their Forge has always worked badly, I’d take a guess at the Forge never having enough cards in hand to work with.
Now lets look at the Coppersmith. When I bought this guy I planned to Remodel all the Estates out of the deck. It was getting close to an ideal situation where the only other terminal cards in the deck would be 7 Copper, a Silver, a Remodel. This would make it highly likely that the Coppersmith would be better than Silver! Unfortunately reality has to set in and the reality is that a Coppersmith never gets perfect copper filled decks to work with. We fill the decks with pollutants like Gold and Provinces. Nevertheless this Coppersmith was able to find 3 coppers each time I played him (average draws) and that was good enough. Again, the enlarged hand size from the Caravans helped the Coppersmith enormously.
There’s another card in the deck that the caravans help. They help themselves. Drawing a big hand at the start of your turn means that you’re likely to draw more caravans, play more Caravans, get a bigger hand next turn, and so on. In this game, Yaron and myself only had one each turn where we didn’t play a caravan from turn 5 through to the end of the game. Considering that these are duration cards (so that some of them are on the table and not in the draw pile) this is good going. This positive feedback from drawing cards can also be seen with wharves and labs and isn’t a trivial consideration.
The last action card in my deck was a Haven that I bought in preference to Silver. This little card kept my deck cycling quickly and gave me choices such as what cards to put with my Coppersmith (copper obviously) and what cards to put with my Remodel. It also slightly reduced the risk of my two terminal actions clashing. It also had the chance to pull a universally good card like a caravan from the deck (next hand) and get it played sooner.
Finally I’d like to give an important mention to Gold. Gold was very important in the deck. Having nice action cards with lots of synergy is nice and fine but you need some payoff and the Gold provided the payoff. Two of my provinces were bought with gold and two were remodelled from Gold. At the end of the game a Remodel can stop creating your ‘engine’ and start to deliver victory cards. It can do it very well. This game also shows that a Remodel can give you two victory cards in a turn without an extra buy. Even if that extra victory card is only an estate it can still be enough to win.
Dominion Strategy 
A very surprising game for me. I really hope Geronimoo works hard and we can have some simulation data on both strategies.
The coppersmith buy is indeed justified and hard to realize in the first place. In my opinion it is hard to judge when to get one instead of yet another caravan, or a Throne Room. (It is not obvious at all for me how Throne Rooms are totally ignored here…)
On the other hand, the ability to remodel copper to haven just isn’t helping here. It is still an awful idea to waste your remodel on a copper. The statement of using Haven to draw a good card is also flawed as that card would be in your hand have you not had the Haven in the first place.
I think this game shows how Caravan is a powerful card combined with some form of trashing. Without looking at the game record I won’t believe that both strategy can get to 4 provinces in 12 turns. Overall, an very interesting read!
> (It is not obvious at all for me how Throne Rooms are totally ignored here…)
Maybe because there is fierce competition among the $4 cards here. But I am surprised too, as Throne Room is unlikely to be the only Action in a hand given enough Caravans in the deck and moderate Masquerade trashing, and Throne Room + Caravan > 2 Caravans.
but it’s only greater for +1Action, which at least for Yaron is completely useless.
For DG the +1 Action is also not soo important, given the risk that you draw TR alone.
You have the possibility to TR the Coppersmith, which – without +Buy in the Deck – might also be mostly wasted. Playing Coppersmith AND Remodel or 2xRemodel might be a little bit more usefull, but overall it is not so a great winning. There are only 2 draws with 2x Caravan for DG, so I don’t think that the TR really worked out. So given you value playing the Caravan higher than playing a second CS or RM, the TR would have done more harm than good.
After re-reading the game, I figure you may be right, but the caveat is that the interface doesn’t tell you which cards were drawn from last-move Caravans.
A Throne Room would have been much stronger with a longer game (Curses or Colonies), in which the esxtra Action would allow Council Rooms. A 12-move game is definitely to short to take that route.
Not completely useless, even for Yaron, as he might draw another caravan with the masquerade. But mostly useless, definitely. For DG, I would say that only TR on remodel might be a reason to get TR. The extra action is unlikely to be useful for him as well, and doubling a coppersmith but having one less card in hand is almost worthless. I think ignoring TR was the right decision.
“The statement of using Haven to draw a good card is also flawed as that card would be in your hand have you not had the Haven in the first place.”
No, it isn’t — he mentioned that the alternative was buying a silver. Buying nothing would indeed have been another option, but I think it’s pretty clear that Haven is better than that. It’s probably also the case that Silver is better than that, and if so, the ability to cycle is in fact relevant in evaluating Haven against it.
“No, it isn’t — he mentioned that the alternative was buying a silver.”
Just to nitpick, timchen was talking about the hypothetical plan of Remodeling Copper into Haven. So the alternative was having a Copper in place of the Haven.
Just to nitpick even more, no he wasn’t ^^
That statement did come at the end of the paragraph in which he was talking about remodeling copper being a bad idea, but he switched focus slightly (indicated by the “also” in “is also flawed”) and was directly referencing a different statement made by DG:
“The last action card in my deck was a Haven that I bought in preference to Silver. […] It also had the chance to pull a universally good card like a caravan from the deck (next hand) and get it played sooner.”
Sure. For that matters, I just found for me it is more consistent overall to phrase the cycling as a problem of Silver instead of an advantage of Haven. Depending on your point view of course, but having some feature which is already there by having nothing is not what I think worth bragging about.
Or to put it another way: without the +1 card, Haven would be an incredibly weak card. Do you want to exchange a precious card slot in hand for the ability to save one card for the next turn? I doubt it. Same for lots of other self-replacing cards. I just see those card as “invisible”, which means they provides some marginal benefit without occupying a card slot. Emphasizing that they can draw the good cards on the deck just seems awkward for me.
If you want to see what the havens can do here I suggest you try the hand with a haven/remodel opening and never buy a silver. This small change propels the turns the deck into the ‘coppersmith utopia’ where you become dependent upon the coppersmith to buy provinces and the coppersmith is dependent upon the havens and caravans to feed him copper and draw him every turn. This deck does work but is probably a turn slower than opening with a silver.
So the conclusion is…? Are you suggesting a single haven is comparable to a silver in this setting? It’s conceivable, but the point here I want to make is that if two or more of the remodels are used on coppers->havens, this strategy is probably going to lose. Turning a copper into a haven is itself great, but wasting 2 cards in hand to do so seems too slow. Therefore, it seems to me the presence of haven is not a major reason to consider an initial remodel purchase.
The lack of a good 2 value card in a kingdom can severely weaken a remodel strategy. As soon as you remodel out some of the estates you become less likely to draw future hands with both remodels and estates and more likely to draw remodels with copper. In this deck I liked having an option to remodel a copper into a haven. It’s far better than having no option but to suffer a bad draw.
The haven is also of value in this deck. The deck was effective because it put the right cards together in combination rather than just having a bulk of good cards. If the haven splits the remodel and coppersmith, or delivers the right treasures to the action cards, then it is worth the loss of one copper.
I, personally was wondering if you could make good use of the baron utilising caravans and havens, figured that would make the baron strong throughout the middle game when it usually becomes weak
Yes, Baron + Haven was the start I had in mind too (is this to slow in terms of cycling? It should give more early buying power). A deck without +buy or remodel might be fast enough, but lacks flexibility.
from Annotade Game #4:
“Editor’s note: one of these days we’ll feature a game with yaron that he wins. Honest.”
Maybe, I’ve missed this one 😉
Was Yaron correct not to follow the penultimate province rule?
At this stage of the game certainly yes, as he has no +buys with 3-3splitted provinces, if you want to win you have to gain the 4th. If he would not have bought it DG had 4 Province,1 Estate anyway and Yaron stick with 3 Provinces, there are not many possibilities how you can win then without +buy. Masquerading into a 4xProvince+1Estate Hand from DG and then buy a Province would be the only possibility, and I think getting the 4th province, hoping that the opponent doesn’t grap one in his next turn and buying the 5th the turn after has a higher probability than this one, even if it’s small.
Strictly speaking, he didn’t break it. That rule only applies when you’re behind, and they were tied. (Unless I’m very mistaken, it would have been a tie if DG had simply bought the Province, since they would each have played 12 turns).
With that said, buying the penultimate province obviously turned out poorly for Yaron. It was certainly risky, since DG had the capacity to take the tiebreaker point with Remodel.
Yaron buying a Duchy might have failed immediately too. DG could have countered by remodeling Province -> Province and buying the last Province. But suppose you’re Yaron, and you assume the worst-case scenario that DG is capable of doing this next turn. I suppose this is practically impossible unless he has a Gold in his hand as well as a Remodel. (He needs either Gold or Coppersmith to buy a Province, and can’t play Coppersmith and Remodel in the same turn). In this case, buying a Province won’t save Yaron either! DG can Remodel Gold into Province and probably buy at least an Estate. (With two extra Caravan cards, this is almost assured).
So here’s how I see things: by buying the penultimate province, Yaron probably couldn’t have hoped for much better than a draw (the probable outcome if DG didn’t get his Remodel, which was about a 50/50 chance). There was a slim chance that DG would effectively draw a seven-card hand of Copper and Provinces, allowing Yaron to snag the last Province, but I don’t think it was very good.
If Yaron had bought a Duchy, things become less clear. As I said, he hasn’t really increased the chance of being defeated next turn (as any hand that could eat through both Provinces would probably have beaten him anyway). But it’s hard to say whether he’s increased his chances of winning or just delayed the inevitable and maybe even decreased the chance of tying the game. The odds would probably be against him in the long run; his Masquerade is weak in the endgame, while Remodel remains potent. But the odds are already against him, so who can say which the lesser evil is?
All I know for sure is that this is a much more complicated situation than the standard penultimate-province rule can handle.
I, for one, think he might fare better had he bought a Duchy. If two provinces are gone next turn so be it. If not, it would then be not so easy for DG to decide to get the second-to-last Province, if he’s not able to grab a Duchy at the same time. Remodel may be weaker if this turns into a Duchy race.
There’s another place that fast deck cycling played a critical role here, which hasn’t been mentioned.
DG got his first Province by Remodeling Gold on turn 8. This is something I would have been terrified to do, especially since he only had one other Gold in his deck at the time, and was immediately forced to reshuffle before he could get another one!
He would never have had the buying power to keep up with yaron, except for his incredible turn 9. Having reshuffled, he drew all but four cards from his deck that turn and bought another Gold, almost immediately shuffling it into his draw pile (and proceeding to use both Golds to buy Provinces on the next two turns). I wonder if DG would have chosen to Remodel his Gold, except that he had just played three Caravans on turn 8 and was therefore assured of a quick reshuffle?
As things turn out, the first Gold was among the four cards in his deck that weren’t drawn on turn 9. But if he had drawn it, at first it seems he should have set it aside with Haven, since without it he could still buy the second Gold which was critical here, and then use the Gold to buy a Province next turn. But maybe not! If the first Gold had been used to buy a Province on turn 9, it would have been shuffled back into the deck and reused on turn 10 or 11. Using the same Gold to buy a Province on turn 9 and again on turn 10/11 would have been just as good as what happened in the game (where a second Gold was bought on turn 9 so that one could be used on turn 10 and the other on turn 11)!
I would really love to see an article about the intricacies of reshuffles and planning around them, going beyond the virtues of fast deck cycling. After all, as I just described, DG needed two Golds to do the work of one Gold because of a reshuffling quirk that really can’t be explained in terms of fast/slow cycling. And with cards like Haven or Royal Seal, it’s very possible to manipulate this, whereas of course even Smithy can be used to deliberately trigger a reshuffle on the rare occasion that this might be important.
A very good and delicate point here. It’s almost the hardest aspect of the game IMO.
Incidentally, the play in turn 8 is also questionable: is it obvious better to play this way rather than getting yet another gold? With the same draw in turn 9, he would have bought the province, and with 1 more gold vs. 1 more haven in the original play.
When you consider that the game was over by turn 12 then gaining the first province on turn 8 is not ridiculous. Yaron also bought his first province on turn 8 and I knew my deck had to deliver provinces through the coppersmith and remodel. My deck was not going to present these draws to me every single turn so I needed to take the chances as they were presented.
Remaining competitive in vp is very important in short games too or else ‘scoreboard pressure’ can mount quickly. If I’d been able to gain a small advantage then the option to remodel a province to a province, denying points to my opponent, could have come into play.
Yeah I understand this reasoning perfectly, and I would probably do the same. But the real question is whether this is really better. For the record we have, if we get another gold in the turn, after the next turn, we would have the same number of Provinces but one more gold and one less Haven. (Well, strictly speaking, this means if you use $6 that turn to buy a haven, you get the same result as remodeling and buy a haven after the next turn.) So for this particular draw remodel the gold is not a particularly strong move. The question is, what about random draws? I do have a feeling that getting a gold there in average is actually better.
Or think in another way, one gold and one province in the deck is never enough. After remodeling the gold, it is most likely that you will have to spend another turn to buy another back. If you get a gold that turn right away, presumably at that particular later turn you can get the province.
I wish the log could tell us how many times yaron drew a Caravan with Masquerade.
Wait, nevermind. It does. Not sure what I was thinking there.
No, I was right the first time. It doesn’t. It also doesn’t tell us which cards are passed when Masquerade is played.
This (the masq problem) is probably the biggest problem with the logs. But I can’t complain too much, ’cause they’re really good.
I imagine it should be relatively simple to modify the code to record the passed card as well. (As well as cards hidden by Haven, card drawn in a turn.) I wonder why it is not implemented?
at a guess, the logs report(ed) what is(was) printed in game, and in multiplayer, you don’t know what passes between your opponents, so you can’t print the information in the log.