Print

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

[tlloyd]

My opponent opened Envoy/Silver, so I tried Oracle/Silver with a second Oracle soon after.

Envoy obviously draws more, which can lead to early Gold, but it also allows the opponent to nerf the best card you draw, which can start to hurt toward the end when you draw more than one green. But even if your best cards are getting discarded, Envoy still cycles through your deck like mad, which in general helps you ramp up quickly to $8 hands.

Oracle is in some sense the opposite of Envoy. It only draws two cards, but at times allows you to avoid drawing bad cards. It also allows you to - again - keep your opponent's best cards out of his hand. And while Envoy's huge draw creates a significant threat of collision if you buy more than one, Oracle's built-in defense against collisions allows you to buy more than one.

So my thought was that multiple Oracles should match up to a single Envoy in terms of draw power, and the fact that both my Oracles and his Envoy give me opportunities to interfere with his drawing should give me the edge.

Here's the log: http://dominion.isotropic.org/gamelog/201202/20/game-20120220-162906-e6335563.html

As it turned out, Oracle's "check first" ability was very useful, as I more than once discarded two green cards. Even more significant was the fact that I actually discarded his Envoy more than once, which really slowed him down.

So did I get lucky, or does Oracle have the edge over Envoy?

[Forge!!!]

Got rid of Envoy at two important province times and then got a province in 4 out of the last 5 turns. I'm going to go with lucky, but we shall see what the simulators say.

[HiveMindEmulator]

I think I and WW have independently checked this with Geronimoo's simulator. This was actually my first thought when I read the Oracle card: Oracle+BM might be able to beat stuff that relies on a single good card. It's close (like 42-50 vs Envoy), but it doesn't actually work.

[DG]

The simulator prefers envoy by a small percentage. The simulator misses two advantages of the oracle though. It always rates copper as a bad card even on turn 3 so gets no advantage from the spy ability on early turns. It also misses opportunities to manage your shuffles and possibly your opponent's shuffle as well.

[tlloyd]

Got rid of Envoy at two important province times and then got a province in 4 out of the last 5 turns. I'm going to go with lucky, but we shall see what the simulators say.

First of all, the chance of discarding your opponent's good cards (including Envoy) is one of the functions of Oracle. So just saying "how lucky! you discarded his good cards!" is not very careful analysis.

Second, I hit and discarded his Envoy on (my) turns 6, 9 and 11. This can only affect his next hand, not the one he is currently holding. And because he is first player, that means I affected his hands on (his) turns 8, 11, and 13 (he is already holding his hand for his turn 7 when I play Oracle on my turn 6 and discard the Envoy that he would have drawn on his turn 8 ). Interestingly enough, he bought a Province on all three of those turns! So my discarding his Envoy never kept him from buying a Province. It may have slowed his cycling, but I'm not sure whether that hurt him given that he was already greening.

Finally, I did buy Provinces consistently at the end -- that is, after all, the objective. But did my ability to do so stem from Oracle? It certainly appears so. On turns 9 and 11 I discarded two green cards with Oracle and then drew enough for a Province. So I wouldn't be so quick to attribute the win to luck, rather than to the benefits of an Oracle strategy.

EDIT: did anyone notice the inadvertent from "on his turn 8 )"?

[tlloyd]

The simulator prefers envoy by a small percentage. The simulator misses two advantages of the oracle though. It always rates copper as a bad card even on turn 3 so gets no advantage from the spy ability on early turns. It also misses opportunities to manage your shuffles and possibly your opponent's shuffle as well.

I think this is key. Deck management is not my strongest skill, but notice on turn 4 I drew a copper and estate with $5 already in hand. I kept the copper and estate because that would allow me to buy a gold without triggering a reshuffle, so my gold (and Oracle) could show up again quickly. I'm guessing the simulator would have chucked the copper and estate.

[HiveMindEmulator]

It's hard to say how much of a difference the sub-optimal play makes. Stuff like managing reshuffles beyond turn 10 is pretty negligible, but stuff that happens on the first shuffle can be a big deal. The thing is, even if it's not 42-50 and is instead 47-45 or something, it's at least not significantly superior, if superior at all. You have to have a little more else going on for Oracle's ability to "counter" Envoy to make a real tangible difference.

[tlloyd]

It's hard to say how much of a difference the sub-optimal play makes. Stuff like managing reshuffles beyond turn 10 is pretty negligible, but stuff that happens on the first shuffle can be a big deal. The thing is, even if it's not 42-50 and is instead 47-45 or something, it's at least not significantly superior, if superior at all. You have to have a little more else going on for Oracle's ability to "counter" Envoy to make a real tangible difference.

So I know nothing about simulators, but I'll take your word for it. Couple questions though: first, how many of Envoy and Oracle do the simulators buy? I also wonder if Envoy is higher variance, given that you want one or at most two envoys, and therefore there is always a chance of going through your whole deck before getting to your (probably single) envoy. Whereas with Oracle you can buy two or even three, making it much less likely that you'll go many turns without an Oracle.

[Forge!!!]

Got rid of Envoy at two important province times and then got a province in 4 out of the last 5 turns. I'm going to go with lucky, but we shall see what the simulators say.

First of all, the chance of discarding your opponent's good cards (including Envoy) is one of the functions of Oracle. So just saying "how lucky! you discarded his good cards!" is not very careful analysis.

Second, I hit and discarded his Envoy on (my) turns 6, 9 and 11. This can only effect his next hand, not the one he is currently holding. And because he is first player, that means I effected his hands on (his) turns 8, 11, and 13 (he is already holding his hand for his turn 7 when I play Oracle on my turn 6 and discard the Envoy that he would have drawn on his turn 8 ). Interestingly enough, he bought a Province on all three of those turns! So my discarding his Envoy never kept him from buying a Province. It may have slowed his cycling, but I'm not sure whether that hurt him given that he was already greening.

Finally, I did buy Provinces consistently at the end -- that is, after all, the objective. But did my ability to do so stem from Oracle? It certainly appears so. On turns 9 and 11 I discarded two green cards with Oracle and then drew enough for a Province. So I wouldn't be so quick to attribute the win to luck, rather than to the benefits of an Oracle strategy.

EDIT: did anyone notice the inadvertent from "on his turn 8 )"?

You played Oracle 6 times during the game, and hit his Envoy half of them. I would assume that this is, if nothing else, at least above-average luck.

Cycling is a big part of an Envoy deck. It works effectively while only buying one because of its cycling ability, and isn't something that can simply be dismissed. Of course, that he was able to buy provinces all three turns was also above-average luck.

It's the objective, that doesn't make it any less lucky that you were able to. Everyone hates the dreaded 7.

And of course, whether or not your game was luckier than usual or not is kind of moot now that we have simulator data. Good old simulators. (question for simulator people: does having 2 envoys instead of 1 help against oracle/bm? i would assume not, but maybe.)

Nothing like hitting quote instead of modify every time. Sorry about that.

[Tahtweasel]

I was the opponent here.

Obviously tlloyd was lucky - there's no question about that. If you get to 5 provinces 4 estates in 13 turns with a Big Money strategy, you've been lucky.

He won by a ton, though - so the question is how lucky he was, not whether he was. I'm sure Double Oracle + BM comes very, very close to Envoy + BM in a typical game. What we're wondering is whether I made the slightly better play, or he did.

[ftl]

Simulator buys a bunch of oracles; up to 3. Buys one envoy early, and possibly a second one late. 

Default bots give a 50-40-10 advantage to envoy. Oracle play is imperfect, so reality is probably closer; if better oracle play makes a difference in, say, 5% of games, that already pushes the percentages to 45-45-10.

Fun fact - the advantage of envoy over oracle in the simulators is less than the first-player advantage. Oracle-BM as first player has the advantage over second-player Envoy-BM.

I'm not quite sure how to go 'beyond simulators' in this other than just playing a bunch of games.

I'm not sure envoy is any higher variance than Oracle. Because while with Envoy there's a lot of luck in where in the shuffle you draw your single envoy, with Oracle there's a lot more luck every time you play it - if the top of your deck is [estate, estate, silver, silver] that's way better than [silver, silver, estate, estate] if you have an oracle in hand, but it's equivalent if you have an envoy instead.

[WanderingWinder]

I believe that oracle is the strongest BM card outside of cursers, probably goons, possibly monument, wharf, probably masquerade, and maybe courtyard. The simulator plays it rather dreadfully actually.

Edit: Oh and quite possibly jack, too.

[HiveMindEmulator]

^I think you may be overrating Oracle a bit. Even though the simulator makes a good deal of wrong plays, I'm somewhat skeptical that it makes the difference between a loss and win that much of the time. It's in the power neighborhood of the $4 terminal draw cards, but not even close to the other cards you mentioned. I don't really see why Courtyard is a "maybe" or why you omit Ghost Ship.

[WanderingWinder]

^I think you may be overrating Oracle a bit. Even though the simulator makes a good deal of wrong plays, I'm somewhat skeptical that it makes the difference between a loss and win that much of the time. It's in the power neighborhood of the $4 terminal draw cards, but not even close to the other cards you mentioned. I don't really see why Courtyard is a "maybe" or why you omit Ghost Ship.

I omit Ghost Ship because, while it is in the realm of those other cards, oracle owns the matchup there. I seriously do think it makes that much of a difference how much it misplays, because it uses the wrong kind of decision-process, not just at a wrong point on the scale. And I really do see oracle as being on the same level as courtyard here.
Anybody want to play me some matches with oracle vs these others, stipulating that we have to play that matchup (obviously if we're fixing those cards, it won't be for rating anyway)?

[HiveMindEmulator]

I thought you were talking about overall goodness with BM, not head-to-head. But even so, I don't think Oracle beats Ghost Ship head-to-head. While Oracle in some sense counters the attack, it's not enough to make it better. You still end up playing 4-card hands vs 6-card hands. So even though you filter 2 cards, the Ghost Ship guy just keeps the 2 extra cards. I don't know though. It's really hard to judge the strength of the Oracle attack. I haven't played with/against it enough to know how I feel about it.

If you want to play-test, you can just open 2 different browsers (not 2 windows, 2 browsers, so they don't share cookies) and play against yourself. But you may have to play a lot of games to get concrete results...

EDIT: I just played a couple games, and obviously I can't conclude much from the small sample size, but the attack is in some ways better than I was thinking against Ghost Ship. If you let him keep Ghost Ships on top and remove other stuff (even if it's something like Silver+Copper), you're more likely to induce dead draws.

EDIT 2: After thinking more about it, I think you're right about Oracle vs Ghost Ship. 4 filtered is better than 6 anti-filtered, provided you do some smart Oracle play.

[WanderingWinder]

Oh, and it may well be weaker than Noble Brigand.
Seriously, we're all massively underrating the brigand for any kind of silver-gold based economy.

[Davio]

I have a love/hate relationship with Brigand.

The point in the game where I want it is when there's an actual chance of hitting a Silver/Gold, but from that point on the Brigand won't see that many plays (greening) and the gained Silver/Gold comes even later. It seems way way better in multiplayer and doesn't scale back to 2p as well as other cards.

I think Brigand is better to attack a BMU-deck dependent on a few Golds than to use one in it.