Print

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

[DG]

Hermit/silver is likely to have high variance rather than any second player bias.

[liopoil]

I'm surprised bishop/silver doesn't show up for P2, and surprised militia/silver doesn't show up for P1. My nonsense theory for hermit-silver: these game tend towards fast three piles, which has more ties than average, which favors P2.

I'm really surprised that there are mirror openings where P1 has a 70% winrate... that means there are kingdoms where P1 is heavily favored.

[ben_king]

I'm surprised bishop/silver doesn't show up for P2, and surprised militia/silver doesn't show up for P1. My nonsense theory for hermit-silver: these game tend towards fast three piles, which has more ties than average, which favors P2.

I'm really surprised that there are mirror openings where P1 has a 70% winrate... that means there are kingdoms where P1 is heavily favored.

Bishop-Silver is up there for P2 with a P2 win % of 47% (the overall expected win percentage for P2 is 43%), but I don't have enough data to say for sure that this is a real effect.  Surprisingly, Militia-Silver seems pretty close to average with a P2 win % of 44%.

There are actually several openings that have a much higher P1 % than 70% in the mirror, but I don't have enough data to be totally confident about them. 

• Haggler-<nothing> (81%)
• Wharf-Silver (80%)
• Monument-Forager (79%)
• Horse Traders-Chapel (77%)
• Salvager-Market Square (76%)

[c4master]

I'm surprised bishop/silver doesn't show up for P2, and surprised militia/silver doesn't show up for P1. My nonsense theory for hermit-silver: these game tend towards fast three piles, which has more ties than average, which favors P2.

I'm really surprised that there are mirror openings where P1 has a 70% winrate... that means there are kingdoms where P1 is heavily favored.

I think you are wrong with you last implication.

What it says, is, that there are card combinations, that are bad to mirror as P2. It does not say, whether there can be other cards to minimalize this effect (because now we're looking at mirrors). For example, if your P1 opponent opens Swindler/Silver you could try something else like Remake/Silver, so that the advantage of P1 goes down by a few percentages. Maybe, if my opponent opens with an attack card, I am more likely to invest into defense like a moat or something, and maybe this can be better than mirroring what P1 does.

Btw. I am surprised not to see any Witch-openings in the list of P1 advantages.

[liopoil]

So you're saying that as P2, when your opponent opens tournament/ambassador, you won't mirror? Good luck...

[T-hawk]

Putting it in a binomial distribution calculator (http://stattrek.com/online-calculator/binomial.aspx), you can see that if you assume it's random who wins (50% chance for each player), the probability of getting 60 or more wins out of 90 is 0.103%, about 1/1000.  So obviously you want to say that these results are meaningful (99.9% chance it's not coincidence), but man, I have no idea why Rebuild/Silver would favor P2.

This is how stats works, right?  Someone who knows this stuff better than me should check, but this feels right to me.

This is a multiple endpoints fallacy.  What's happening is not that Rebuild/Silver is a 1/1000 outlier.  What's happening is that we looked over the population of all possible openings and cherrypicked the one that deviated the most.  If there are 1000 possible openings in Dominion, then one of them is going to be a 1/1000 outlier.

We didn't try one candidate and get a 99.9% unlikely result.  We tried 1000 candidates, of which one displayed a 1/1000 result.  That is perfectly ordinary.

In other words, there's nothing special about that 0.1% chance that Rebuild/Silver displays those results, because we also have a population who fills out the 99.9% of unexceptional results.  We just didn't pay any attention to the other ones and cherrypicked the outlier.

As usual, XKCD nails it: http://xkcd.com/882/

[GeoLib]

Putting it in a binomial distribution calculator (http://stattrek.com/online-calculator/binomial.aspx), you can see that if you assume it's random who wins (50% chance for each player), the probability of getting 60 or more wins out of 90 is 0.103%, about 1/1000.  So obviously you want to say that these results are meaningful (99.9% chance it's not coincidence), but man, I have no idea why Rebuild/Silver would favor P2.

This is how stats works, right?  Someone who knows this stuff better than me should check, but this feels right to me.

This is a multiple endpoints fallacy.  What's happening is not that Rebuild/Silver is a 1/1000 outlier.  What's happening is that we looked over the population of all possible openings and cherrypicked the one that deviated the most.  If there are 1000 possible openings in Dominion, then one of them is going to be a 1/1000 outlier.

We didn't try one candidate and get a 99.9% unlikely result.  We tried 1000 candidates, of which one displayed a 1/1000 result.  That is perfectly ordinary.

In other words, there's nothing special about that 0.1% chance that Rebuild/Silver displays those results, because we also have a population who fills out the 99.9% of unexceptional results.  We just didn't pay any attention to the other ones and cherrypicked the outlier.

As usual, XKCD nails it: http://xkcd.com/882/

Yeah this is the same thing I was talking about earlier, just by another name.

[c4master]

So you're saying that as P2, when your opponent opens tournament/ambassador, you won't mirror? Good luck...

There might be openings that do better against tournament/ambassador.

Maybe it's worth to take the risk of collision with double ambassador? Maybe you could try to overcome the copper flood with Apothetecary. Maybe there's some other attack like Swindler or Knights to kill the ambassador...

If your chances are already down 30-70, you could just as well try and go for something weird, meaning: something with high variance.

[liopoil]

So you're saying that as P2, when your opponent opens tournament/ambassador, you won't mirror? Good luck...

There might be openings that do better against tournament/ambassador.

Maybe it's worth to take the risk of collision with double ambassador? Maybe you could try to overcome the copper flood with Apothetecary. Maybe there's some other attack like Swindler or Knights to kill the ambassador...

If your chances are already down 30-70, you could just as well try and go for something weird, meaning: something with high variance.

Double amb could work, but I don't know if that would score much better. Anything else is probably just bad; if it were good your opponent would have gone that route.

[WanderingWinder]

I'm surprised bishop/silver doesn't show up for P2, and surprised militia/silver doesn't show up for P1. My nonsense theory for hermit-silver: these game tend towards fast three piles, which has more ties than average, which favors P2.

I'm really surprised that there are mirror openings where P1 has a 70% winrate... that means there are kingdoms where P1 is heavily favored.

I think you are wrong with you last implication.

What it says, is, that there are card combinations, that are bad to mirror as P2. It does not say, whether there can be other cards to minimalize this effect (because now we're looking at mirrors). For example, if your P1 opponent opens Swindler/Silver you could try something else like Remake/Silver, so that the advantage of P1 goes down by a few percentages. Maybe, if my opponent opens with an attack card, I am more likely to invest into defense like a moat or something, and maybe this can be better than mirroring what P1 does.

Btw. I am surprised not to see any Witch-openings in the list of P1 advantages.

The idea that you should do something different as player 2 because you're player 2 is almost always just wrong. There's usually a best thing to do, and you should just do that. And when this isn't the case, you're usually adapting because some strategy is a little better against what they did, and it was close to start. You're really almost never never switching just because you're p2.

[DG]

The idea that you should do something different as player 2 because you're player 2 is almost always just wrong.

There are some exceptions where the second player can avoid strategies that reinforce first player advantage. These tend to be games with rapidly increasing momentum where first player is likely to have first shot at a big finish, collecting key cards, or a three pile ending. There still needs to be a competitive counter strategy that can break up the play though and it might not start with a different opening.

[c4master]

I'm surprised bishop/silver doesn't show up for P2, and surprised militia/silver doesn't show up for P1. My nonsense theory for hermit-silver: these game tend towards fast three piles, which has more ties than average, which favors P2.

I'm really surprised that there are mirror openings where P1 has a 70% winrate... that means there are kingdoms where P1 is heavily favored.

I think you are wrong with you last implication.

What it says, is, that there are card combinations, that are bad to mirror as P2. It does not say, whether there can be other cards to minimalize this effect (because now we're looking at mirrors). For example, if your P1 opponent opens Swindler/Silver you could try something else like Remake/Silver, so that the advantage of P1 goes down by a few percentages. Maybe, if my opponent opens with an attack card, I am more likely to invest into defense like a moat or something, and maybe this can be better than mirroring what P1 does.

Btw. I am surprised not to see any Witch-openings in the list of P1 advantages.

The idea that you should do something different as player 2 because you're player 2 is almost always just wrong. There's usually a best thing to do, and you should just do that. And when this isn't the case, you're usually adapting because some strategy is a little better against what they did, and it was close to start. You're really almost never never switching just because you're p2.

I don't say, you should always chose a different opening. But when you're already behind because of first player advantage you should try to find something that helps you more than just mirroring. I'm referring to Stef's article about taking more risks as P2.

Obviously, not every board with Ambassadoe und Tournament offers some other viable opening (except maybe double Ambassador). But you should not automatically mirror P1 because now it's pretty known that your chances will go down by more than 10% because of the opening.

[DG]

Obviously, not every board with Ambassadoe und Tournament offers some other viable opening (except maybe double Ambassador). But you should not automatically mirror P1 because now it's pretty known that your chances will go down by more than 10% because of the opening.

I think you are drawing the wrong conclusion. There is no evidence that alternative openings are more successful against ambassador/tournament. In fact, the council room stats for the old isotropic server show that Ambassador/tournament was one of the strongest openings (and I seem to remember those stats are specifically for non-mirror matches).

[Awaclus]

Isn't Ironmonger/Ambassador a better opening though?

[blueblimp]

I'm surprised bishop/silver doesn't show up for P2, and surprised militia/silver doesn't show up for P1. My nonsense theory for hermit-silver: these game tend towards fast three piles, which has more ties than average, which favors P2.

I'm really surprised that there are mirror openings where P1 has a 70% winrate... that means there are kingdoms where P1 is heavily favored.

I think you are wrong with you last implication.

What it says, is, that there are card combinations, that are bad to mirror as P2. It does not say, whether there can be other cards to minimalize this effect (because now we're looking at mirrors). For example, if your P1 opponent opens Swindler/Silver you could try something else like Remake/Silver, so that the advantage of P1 goes down by a few percentages. Maybe, if my opponent opens with an attack card, I am more likely to invest into defense like a moat or something, and maybe this can be better than mirroring what P1 does.

Btw. I am surprised not to see any Witch-openings in the list of P1 advantages.

The idea that you should do something different as player 2 because you're player 2 is almost always just wrong. There's usually a best thing to do, and you should just do that. And when this isn't the case, you're usually adapting because some strategy is a little better against what they did, and it was close to start. You're really almost never never switching just because you're p2.

Yes. The thing to keep in mind is that there are common events that can take place on T1-T4 that are comparable in effect to P1 advantage. Examples: bad opening split, bad money distribution on T3/T4, or a key opening buy missing a shuffle. With Tournament/Ambassador, some additional specific unluck examples that can occur are Ambassador hitting 4 coppers without a good $2 on the board, or various mishaps in the process of gaining a Province and connecting it with Tournament. While any of these individually may be of greater or smaller impact than P1 advantage, the point is that P1 advantage can easily be overturned by events within the game.

[c4master]

I'm surprised bishop/silver doesn't show up for P2, and surprised militia/silver doesn't show up for P1. My nonsense theory for hermit-silver: these game tend towards fast three piles, which has more ties than average, which favors P2.

I'm really surprised that there are mirror openings where P1 has a 70% winrate... that means there are kingdoms where P1 is heavily favored.

I think you are wrong with you last implication.

What it says, is, that there are card combinations, that are bad to mirror as P2. It does not say, whether there can be other cards to minimalize this effect (because now we're looking at mirrors). For example, if your P1 opponent opens Swindler/Silver you could try something else like Remake/Silver, so that the advantage of P1 goes down by a few percentages. Maybe, if my opponent opens with an attack card, I am more likely to invest into defense like a moat or something, and maybe this can be better than mirroring what P1 does.

Btw. I am surprised not to see any Witch-openings in the list of P1 advantages.

The idea that you should do something different as player 2 because you're player 2 is almost always just wrong. There's usually a best thing to do, and you should just do that. And when this isn't the case, you're usually adapting because some strategy is a little better against what they did, and it was close to start. You're really almost never never switching just because you're p2.

Yes. The thing to keep in mind is that there are common events that can take place on T1-T4 that are comparable in effect to P1 advantage. Examples: bad opening split, bad money distribution on T3/T4, or a key opening buy missing a shuffle. With Tournament/Ambassador, some additional specific unluck examples that can occur are Ambassador hitting 4 coppers without a good $2 on the board, or various mishaps in the process of gaining a Province and connecting it with Tournament. While any of these individually may be of greater or smaller impact than P1 advantage, the point is that P1 advantage can easily be overturned by events within the game.

I don't think that's true.

1 - bad opening split:
Talking about a board with Ambassador/Tournament and a situation where you could mirror P1 means, that you both have a 3/4 opening. There is no luck involved.

2 - bad money distribution on T3/T4:
I claim, that's no problem. Best thing for your Ambassador turn would be Tournament + Ambassador + 2 Estates on T3. Returning 2 Estates would still mean only $3. Without the Tournament, if you give back 2 cards with Ambassador, you cannot have more than $2.
The worst thing would be Amb + 4 Coppers. If you cannot buy something useful at that price point, you're "thinning" more than the other player who bought maybe a 3-cost.
Best turn with Tournament would be Tournament  + 5 Coppers giving you $6 (or maybe $5 is just as well depending on the board).
It's tough to say, but I think, only the last thing really matters...unlesss...your opening buys miss the shuffle.

3- key opening buy miss the shuffle:
Given that we mirror each other, the worst thing to happen would be to miss the shuffle for both of your cards. Chances are 2/12 * 1/11 = 2/132 = 1,5%. With a cantrip like Tournament, the second worst would be for Ambassador to be the last card and thus miss the shuffle. Chances are 1/12 = 8,3%. So basically, my opponent gets bad luck on ~10% of all those games. Unfortunately, my own luck will be bad also on 10% of all games, thus remaining a ~9% chance for my opponent to stumble and me not to.
So, yes, chances are comparable, but still my disadvantage of ~10% is bigger than my chances to cover up due to luck (9%).

[HiveMindEmulator]

Btw. I am surprised not to see any Witch-openings in the list of P1 advantages.

Witch generally lengthens the game, which makes first player advantage less important. This should counteract the fact that player 1 is more likely to win a favorable Curse split, especially when there is a way to trash the Curses.