There's two main plans here: Jacks into Duchy/Duke slog, or Engine into see-how-you-need-to-adjust. I want to stress that the engine player should not commit himself, in his own mind, to any kind of VP plan. That would be throwing away one of the biggest advantages of the engine: its flexibility. Ultimately, the decision of how and when to green should be a tactical one, based on the matchup and, more significantly, the game-state.
He goes for the slog plan. I go for the engine. I open Jack/Village. Jack in an engine is really quite reasonable, and it makes me think of my friend Adam Horton, who is a skilled proponent of this kind of strategy. Now, opening Village is something I rarely do (in fact, I think it has been years and several thousand games since it's happened), but there are several factors that lead me to it here, the lack of a $5 I want early is the biggest one. I am also eventually going to want LOTS of villages pretty clearly, and while Silver is a fine card for my deck, I figure I will get enough from Jack.
I am not going to talk too much about how my opponent played here - he got a stash before his second Jack, which seems pretty wrong to me, but other than that, he plays quite reasonably. He definitely did not have the best of luck down the stretch, but for the most part, his play seems quite textbook.
On my side of things, there are a few big decision points I would like to cover.
I get a Forge pretty quickly, despite having a fairly thin deck, no estates left, and Moneylender to take care of all the rest of my Coppers within the next 5 turns or so. The big reason for the Forge is that 1) it speeds me up a little in finishing to trash down, and 2) it lets me go Silver+Silver->Hunting Grounds, which is something I will want in order to keep my deck-drawing going, and potentially with Jack, in a nice draw-positive way. In general, it also gives me some control over how things are going.
I'm going to italicize the next three paragraphs, because I think they're really critical for wider strategy.
But the big decision-point I want to focus on is whether or not to contest Duchies. This is always a pretty key point. I actually think that, as it turns out, I would have won this game either way. But that isn't always the case, and so I want to look at the factors that go into it. First of all, the dangers. Contesting the green early can really choke your deck down and increase your chances to misfire. Furthermore, it can sometimes simultaneously hasten the end of the game on piles, which is generally not in favor of the engine, which is better suited for the long game. On the other hand, there are benefits as well: The Slog's game-plan often involves getting an insurmountable lead, and contesting can allow you to break up that unbeatable matrix of VP.
In this case, stealing three duchies is enough to mean that ALL the Dukes isn't enough for the Slog player to lock it up (as is generally the case with Duchy/Duke) and there likely isn't a three pile coming until AFTER the Dukes are gone - certainly not if the Engine player is careful, as there just isn't any pile the slog player can realistically empty quickly. So the question becomes whether or not taking these three Duchies is too much trouble for it. To know that, we need to look at the other side. If we don't contest Duchies at all, they would typically need 11 5-cost VP cards to set up their matrix. In this case, though, we have the ability to grab Estates via trashing extra Hunting Grounds. So realistically, they'll need 12 of them. Whereas if we get the three Duchies, then 13 will be insufficient - they either need to empty 13 and then a third pile, or get 13 with a Province as well. So with any amount of realism, you're looking at 14-16. Basically, the question is whether this disrupts you enough to forego contesting or not. Getting 2 more 5s and an 8 or 1 more 5 and emptying a pile will cost them some amount of time. The question is, will that amount of time be more significant than the amount of time you are investing as the engine player? At the time of the game, I decided it was worth it to contest on turn 9. Looking back on things now, I think I would lean the other way. But it's definitely a close call.
And this is where a lot of the interplay comes in. The less the Engine player wants to contest, the more incentivized the Slog player is to build her deck up before turning for the green, so that she will be able to power through to her completed VP lockout more quickly. But the more the Slog player is building, the more incentivized the Engine player is to contest, since in this case, it hurts him less (since he's already built up more to be able to sustain green) AND it is more damaging to the Slog Player's VP Matrix. And all of these decisions have to be made based on the game-state as unfolding, but before the game is really clear - in my opinion, probably the most interesting part of a game.
Okay. So after the dust settled from that decision, following 11, our decks look liked this:
At this point, it's going to take him at least 7 turns to finish out the Dukes, after which he needs to empty a third pile (or at any point he can, take a Province). It's of course highly unrealistic for him to be hitting this every turn, though, so more realistically, I look to have a dozen turns or more to empty out the Provinces. Incidentally, his best bet is probably to spike a Province himself, so if he is ever hitting 6 soon, he can seriously consider taking a Gold, or possibly even a Hunting Grounds.
For my part, I ought to be able to get that Province-running task done with relative ease. The trick is to build my economy a bit whilst simultaneously keeping my deck at least as reliable in drawing as it is now. It is just barely drawing the deck now, if all falls perfect; it will fall well fairly often, but it will fail pretty often too, which means I would not actually consider it very consistent as is. So if you go through the log, you can see that I make a number of small errors from this point, mainly in not maintaining quite enough draw to be as reliable as I ought to be.
Nevertheless, a mere 6 turns later, following turn 17, the situation looks like this: