Print

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

[Collatz]

Code: [Select]

Vineyard, Beggar, Farming Village, Feast, Gardens, Talisman, Graverobber, Royal Seal, Upgrade, Wharf
Log: http://www.gokosalvager.com/static/logprettifier.html?/20150127/log.5065fd6ba2e6f4356e5c0907.1422373084435.txt

So my opponent (lv 13) went Beggar-Gardens, while I went Farming Village-Wharf-Upgrade-Vineyards.

I end up denying my opponent 5 of the Gardens (and use Upgrade to trash them for Wharves and Upgrades) and am able to ramp up my Vineyards to 6 points each and barely eke out a win (his Gardens were worth 7 at the end, and he bought most of the Duchies).

That said, I'm not sure whether I made the right choice here. My opponent wasn't aggressive enough in buying Gardens (he didn't buy any gardens until he had 9 beggars!) which is what allowed me to snipe so many. I on the other hand probably should have bought more Upgrades sooner in anticipation that I'm not going to be drawing my deck until I upgrade away the gardens I snipe anyway (and embarassingly I buy a Graverobber at one point - don't ask).

This board lacked cheap spammable actions in an engine, despite the fact that the actions that were present were strong.

If both I and my opponent played this right, who'd win?

[TheOthin]

Trashing Gardens doesn't sound like a great idea when Graverobber is available to take them back. It doesn't seem like your opponent picked up on that though. Also a big enough Vineyards deck could make some use of them itself.

[liopoil]

You got too many gardens. I think if you focus on making vineyards worth at least as much as gardens you can easily win.

[pubby]

I'm surprised nobody bought talisman here.

[Throwaway_bicycling]

I'm surprised nobody bought talisman here.

And that was my first thought. Especially for the Beggar/Gardens person, since Wharf is the only other +Buy (or equivalent). Without looking at the log in depth, the three-pile I would be aiming for here is Farming Village/Gardens/Talisman

[liopoil]

Nah, Talisman is bad here. On $5+ I'd rather get wharf, on $4 I'll get gardens. If it's too early for gardens, I'll get a beggar. That's if I'm going gardens; and I wouldn't go gardens. In the not-gardens deck talisman is still bad, because it's not worth it to spam farming villages if I can't hit $5. Talisman is a weak card in general. Here talisman, royal seal, beggar, feast, and gardens should all be ignored except maybe at the very very end.

[Throwaway_bicycling]

Nah, Talisman is bad here. On $5+ I'd rather get wharf, on $4 I'll get gardens. If it's too early for gardens, I'll get a beggar. That's if I'm going gardens; and I wouldn't go gardens. In the not-gardens deck talisman is still bad, because it's not worth it to spam farming villages if I can't hit $5. Talisman is a weak card in general. Here talisman, royal seal, beggar, feast, and gardens should all be ignored except maybe at the very very end.

Okay, so I actually just read the last line on Talisman and noted the "that is not a victory card" part. Which really very much reduces the enthusiasm I had for it in a Beggars-Garden rush. So...never mind. (Have to say I don't consider buying Talisman very often so didn't realize the VP card prohibition, but I should have done so before posting.)

[pubby]

The benefit of Talisman is that you can end the game on piles much faster. You can't beat Vineyards if you don't rush fast enough.

[Burning Skull]

The benefit of Talisman is that you can end the game on piles much faster.

For a Beggar/Gardens player at every point of this game if you consider buying a Talisman, you would be better served with either Gardens or Beggar.

You can't beat Vineyards if you don't rush fast enough.

That's the problem with Beggar/Gardens here. But Talisman doesn't help it in any way.

[WanderingWinder]

The benefit of Talisman is that you can end the game on piles much faster.

For a Beggar/Gardens player at every point of this game if you consider buying a Talisman, you would be better served with either Gardens or Beggar.

You can't beat Vineyards if you don't rush fast enough.

That's the problem with Beggar/Gardens here. But Talisman doesn't help it in any way.

Are you sure that a Turn 1 Talisman to get multi-beggars is no good? I'm not.

[Burning Skull]

According to simulation, Talisman in Beggar/Gardens is a no go.

Code: [Select]

{
  name: 'BeggarGardensTalisman'
  requires: ['Beggar', 'Gardens', 'Talisman'] 
 
  gainPriority: (state, my) -> [
    "Talisman" if state.countInSupply("Talisman") == 10
    "Gardens"
    "Duchy"
    "Estate" if state.gainsToEndGame() <= 4
    "Beggar"
    "Silver"
    "Estate"
    "Copper"
  ]
}


Code: [Select]

{
  name: 'BeggarGardensNoTalisman' 
  requires: ['Beggar', 'Gardens'] 
 
  gainPriority: (state, my) -> [
    "Gardens"
    "Duchy"
    "Estate" if state.gainsToEndGame() <= 4
    "Beggar"
    "Silver"
    "Estate"
    "Copper"
  ]
}

Player with no Talisman wins 62.2% in 10000 games.

[WanderingWinder]

According to simulation, Talisman in Beggar/Gardens is a no go.

Code: [Select]

{
  name: 'BeggarGardensTalisman'
  requires: ['Beggar', 'Gardens', 'Talisman'] 
 
  gainPriority: (state, my) -> [
    "Talisman" if state.countInSupply("Talisman") == 10
    "Gardens"
    "Duchy"
    "Estate" if state.gainsToEndGame() <= 4
    "Beggar"
    "Silver"
    "Estate"
    "Copper"
  ]
}


Code: [Select]

{
  name: 'BeggarGardensNoTalisman' 
  requires: ['Beggar', 'Gardens'] 
 
  gainPriority: (state, my) -> [
    "Gardens"
    "Duchy"
    "Estate" if state.gainsToEndGame() <= 4
    "Beggar"
    "Silver"
    "Estate"
    "Copper"
  ]
}

Player with no Talisman wins 62.2% in 10000 games.

That simulation shows very very little, because the bots play very badly. For instance, you shouldn't open Gardens.....

[Burning Skull]

That simulation shows very very little, because the bots play very badly. For instance, you shouldn't open Gardens.....

It seems that exact strategy is a perfect target for simulation: you only need to specify the buy order, and play Beggar every time it's in hand. No other decisions involved.
As for opening Gardens, yes. Let us change bots so they only start buying green after having two Beggars in deck:

Code: [Select]

{
  name: 'BeggarGardensTalisman'
  requires: ['Beggar', 'Gardens', 'Talisman'] 
 
  gainPriority: (state, my) -> [
    "Talisman" if state.countInSupply("Talisman") == 10
    "Gardens" if my.countInDeck("Beggar") > 1
    "Duchy" if my.countInDeck("Beggar") > 1
    "Estate" if state.gainsToEndGame() <= 4
    "Beggar"
    "Silver"
    "Estate"
    "Copper"
  ]
}


Code: [Select]

{
  name: 'BeggarGardensNoTalisman' 
  requires: ['Beggar', 'Gardens', 'Talisman'] 
 
  gainPriority: (state, my) -> [
    "Gardens" if my.countInDeck("Beggar") > 1
    "Duchy" if my.countInDeck("Beggar") > 1
    "Estate" if state.gainsToEndGame() <= 4
    "Beggar"
    "Silver"
    "Estate"
    "Copper"
  ]
}

Result remains effectively the same: 63.2%.

[WanderingWinder]

That simulation shows very very little, because the bots play very badly. For instance, you shouldn't open Gardens.....

It seems that exact strategy is a perfect target for simulation: you only need to specify the buy order, and play Beggar every time it's in hand. No other decisions involved.
As for opening Gardens, yes. Let us change bots so they only start buying green after having two Beggars in deck:

Code: [Select]

{
  name: 'BeggarGardensTalisman'
  requires: ['Beggar', 'Gardens', 'Talisman'] 
 
  gainPriority: (state, my) -> [
    "Talisman" if state.countInSupply("Talisman") == 10
    "Gardens" if my.countInDeck("Beggar") > 1
    "Duchy" if my.countInDeck("Beggar") > 1
    "Estate" if state.gainsToEndGame() <= 4
    "Beggar"
    "Silver"
    "Estate"
    "Copper"
  ]
}


Code: [Select]

{
  name: 'BeggarGardensNoTalisman' 
  requires: ['Beggar', 'Gardens', 'Talisman'] 
 
  gainPriority: (state, my) -> [
    "Gardens" if my.countInDeck("Beggar") > 1
    "Duchy" if my.countInDeck("Beggar") > 1
    "Estate" if state.gainsToEndGame() <= 4
    "Beggar"
    "Silver"
    "Estate"
    "Copper"
  ]
}

Result remains effectively the same: 63.2%.

There are a LOT of other problems with it. It plays the endgame pretty badly right now. More importantly, the whole point of buying talisman is to get extra cards with it. Here, the way you're simulating it, it will very very often just be an expensive copper.

I played several practice games, and I'm not sure whether talisman is better or not - it's not going to be much better in the mirror, certainly, and is probably a tiny bit worse, but I imagine it's MUCH better in the non-mirror.

[Burning Skull]

Here, the way you're simulating it, it will very very often just be an expensive copper.

I've tried adding a rule like "don't buy Gardens if you have Talisman in hand and there are X+ Gardens left in Supply", and I varied X all the way from 1 to 8. Did not change the result significantly.


Anyway, I see your point about non-mirror. This simulation can only prove something in B/G vs B/G games.
To see whether Talisman improves B/G chances versus Vineyards player you need to simulate Vineyards strategy on this board as well.

[WanderingWinder]

Here, the way you're simulating it, it will very very often just be an expensive copper.

I've tried adding a rule like "don't buy Gardens if you have Talisman in hand and there are X+ Gardens left in Supply", and I varied X all the way from 1 to 8. Did not change the result significantly.


Anyway, I see your point about non-mirror. This simulation can only prove something in B/G vs B/G games.
To see whether Talisman improves B/G chances versus Vineyards player you need to simulate Vineyards strategy on this board as well.

I don't know how Dominiate handles things - shouldn't you mean "talisman in play"? Surely it should actually change the result significantly at some point - it's just clearly WRONG to do at certain points.

But the problem is, a rule like that is going to be too simple. You need several changes, at least, and you need to change how you handle endgames and greening.

Trying to simulate Vineyards? Fuhgeddaboudit. When I won the simulation tournament for 2 card combos, it was with a script of dozens, if not hundreds (pretty sure it was hundreds) of rules. And that was only with 2 kingdom cards. For a strategy as complex as Vineyards here, it would need tons (you also need good play rules). And anyway, you have to adjust a lot to what the opponent is doing and how the shuffles go, so for all but the simplest strategies, simulation is just not where you want to be.

[ehunt]

How is big money wharf going to go on this board?

[Burning Skull]

shouldn't you mean "talisman in play"?

Yes, that's what I was trying to say. But really, the fluctuation of result was below 3%.

Trying to simulate Vineyards? Fuhgeddaboudit. When I won the simulation tournament for 2 card combos, it was with a script of dozens, if not hundreds (pretty sure it was hundreds) of rules. And that was only with 2 kingdom cards. For a strategy as complex as Vineyards here, it would need tons (you also need good play rules). And anyway, you have to adjust a lot to what the opponent is doing and how the shuffles go, so for all but the simplest strategies, simulation is just not where you want to be.

With that I absolutely agree, that's why I'm not even trying to implement it.

[Burning Skull]

How is big money wharf going to go on this board?

It will lose badly against Beggar/Gardens with or without Talisman.

[Davio]

There's no need to pick between gardens and vineyards.

A wharf engine can take gardens, vineyards and provinces. Timing and pile control is more important here. It's easy to pick up a few gardens so you can neutralize the gardens player quite a bit. And then you can grab the lead with provinces and vineyards and just end on piles.