Tunnel is a card that I think is pretty poorly understood. I certainly know that my friend Adam is not a big fan of the card, and he has said that he feels confused by it often. So hopefully this little article will help to clear things up a bit. Let's dive right in, and look at the different modes Tunnel presents.
Tunnel as VP
The biggest thing to remember about Tunnel is that it provides 2 VP for 3 coins. In terms of coins, that's a more efficient rate than Duchy even, but of course cards tend to be more important than coins, and this is less efficient cards-wise. That hurts you both in that you have to use two buys on this rather than just one, and moreover because you will have to draw two junky cards every shuffle rather than one. Still, sometimes you just need points, and sometimes this is upping the total amount available to you.
I don't have any specific example games here, because it's just too ubiquitous and non-specific to really have - you get the idea though, it's just a green card like Estate or Duchy, just a bit between.
Tunnel as Defense
This is perhaps the biggest trap of all. Your opponent has discard attacks, and you think, aha, I can get tunnels, and then when they attack me, I will gain golds. The problem here is that gaining golds just isn't very good very often in this scenario. Essentially, what you are saying when you do this is, "If I gain enough Golds, my three card hands can beat you". Well, this is usually just a losing proposition. Of course, there are probably some exceptions, where there is just nothing to do, no way to draw, the game will be very long, and you can also leverage the victory part of the Tunnel. But these are going to be exceedingly rare, and for the most part, you would just rather be on the side of the attack, and in general, just having better stuff than a do-nothing green card. There is also, of course, no guarantee that any given hand gets attacked, so you have a pretty good chance of not getting your benefit - unless, of course, your opponent has built a consistent engine. In that case, though, you would almost always have been better served to build such an engine yourself! I will note that using Tunnel as a defense is probably more realistic in multiplayer, where you're much more likely to get attacked every turn - but I still wouldn't expect it to be good often.
I also don't have examples here, because I can't find any recent games of mine where anyone really did this. The word is more or less out now, at least at the high levels, but it wasn't always so, and you will still see quite a number of players falling for this idea.
The Headlong Rush for Gold-Flood
This one is ever-popular. The basic idea is to get Tunnels and enablers (discard outlets) en masse, to then acquire golds en masse, to then buy things. This strategy consistently gets way overrated. The first problem is that you need to get things to line up in the right order for this to work at all - and if you build it right, they usually will, but you're still hanging some non-trivial percentage for disaster. The next problem is that this is slow. First, you have to get Tunnels, which are cards that actively hurt your deck. The next shuffle, you get to start gaining gold, but you won't reap the benefits of that gold until the shuffle after that. So it takes a long time. Further compounding this, you actually need to get a lot of gold before you can do what you want. Typically, your enablers are going to leave you down a card. That means you'll need two gold and two copper in order to buy your province. It takes quite a lot of cards in order to get that set up reliably, even with sifting benefits. Keep in mind one alternative is always buy silver to buy gold, and that is going to be a lot easier on you to reach 8 coins in terms of needing golds. In order to alleviate this problem, you are going to need really exceptional enablers. The best by far are Storeroom and Embassy. They just let you see so many cards, and they provide additional economic benefits to lessen the burden of needing to get Gold Gold Copper Copper exactly. Even with these cards, the strategy is only a reasonably good baseline, not anything super amazing. Young Witch is another mention, but this is not because it's really a good enabler, but because the attack synergizes with the slog-like nature of the Gold Flood strategy to start with. With worse enablers, you should be quite leery indeed.
When you're building this deck, the emphasis should really be on the enablers far more than the Tunnels. The quick thing is that the enablers do something for your deck if they don't collide, whereas the Tunnels don't. They also accelerate your shuffles even when they 'miss', which is a big deal in this kind of deck. Furthermore, you just don't really need that many Tunnels. Getting a couple Golds per shuffle reliably and getting back round to them faster is simply a much bigger deal than potentially getting a higher number of golds per shuffle.
I have one example game here: http://www.gokosalvager.com/static/logprettifier.html?/20150611/log.50a94f0ce4b07a80ae66d119.1434063461261.txt
You'll see I focus on getting my Embassies up much moreso than the Tunnels, and it pays very good dividends (of course, I was also absurdly lucky here). My opponent actually also plays pretty well (perhaps his first Duchy should be another Embassy - in general this is true anyway - but it is hard to criticize as he is so far behind and clearly needs to get rather lucky to have a chance), and his draws are more reasonable as a baseline here.
Tunnel as Payload
This is something I don't see very often, but it's I think pretty significantly the best use of the card. The concept is pretty simple: you build up a big draw engine, then you use a Tunnel to gain Golds to use as the economic finish. The nice thing about this is that a lot of your enablers are cards you already want to grease the wheels of your engine, you can use a single Tunnel multiple times in a turn, and you can use all your buys going forward on engine pieces, without needing to waste any just making money. Also, because engines draw so much, the speed both at which you acquire golds and at which you are able to reap the benefits of them is quite high.
The key to playing this kind of deck is to get your engine up and running first, and add the Tunnel as a payload card. You also want to think about potentially limiting the activations of your tunnel - Gold is a card that is generally pretty good when it's free, but you don't always want more. It's important to keep your deck-size under control and your reliability up - and there's little point in making more money than you need. Furthermore, there are often going to be more opportunities on later turns.
Let's look at a couple of example games I played here:
Here, I get lots of enablers, because they will help get my engine firing anyway. I do take an early Tunnel and a couple of gold, as I need to get my economy up to get the Hunting Grounds I need. But I don't activate it every time, at least until I am getting into draw-my-deck territory. Because I have a high number of sifters, my deck is reliable, and when I get enough Hunting Grounds on-line, my deck explodes, closing out 6 Provinces over 2 turns.
This game is a similar story. One Tunnel soon (in this case, I got to trash Hovel, which had me pulling the trigger sooner than I normally would). Pick up a forge to start coalescing, and then go to town. Forging extra Golds into Hunting Grounds is a big deal here, and while I do go a bit overboard in getting a zillion golds at the end (probably not really the best play), it doesn't really matter, as the game is just ending.