Print

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

[blueblimp]

One interesting topic to come up in Goko-related discussions (on BGG mainly, not here) is asynchronous play. The usual answer is "it wouldn't work", which is of course correct with the current rules, because Minion-Moat-type games would be a nightmare. But maybe varying the rules could make async play work.

I think that async play ideally wants each player's turn to require no decisions from his opponent. You could ask for more maybe, like being able to specify multiple turns at once (especially in the opening), but this is good for a start.

So, I made a list of cards that currently have a player make a decision when it's not their turn. These fall into two categories: reactions and opponent-choice. First the list, then a discussion of how they might be changed.

Reactions
Base: Moat
Intrigue: Secret Chamber
Prosperity: Watchtower
Cornucopia: Horse Traders
Hinterlands: Fool's Gold, Trader, Tunnel

Opponent Choice
Base: Bureaucrat, Militia
Intrigue: Masquerade, Saboteur, Torturer
Seaside: Ghost Ship
Prosperity: Contraband, Mountebank, Rabble, Vault, Goons
Cornucopia: Tournament, Young Witch, Followers
Hinterlands: Duchess, Oracle, Margrave
Promo: Envoy, Governor

Overview of Discussion

There are two main ways to adapt cards: one is to make the decision automatic, and the other is to delay the decision to the beginning of next turn. Making a decision automatic loses some flexibility compared to the current rules, but otherwise is faithful.

Delaying the decision can make a pretty big difference compared to current rules, although it often gets played this way IRL in situations where it doesn't matter. Sometimes it matters a lot, e.g. playing a discard attack to help your Tournaments, which is an interaction that would be lost.

If you're playing asynchronously online, the server could choose to report delayed effects one-by-one, to help preserve the frustration of not knowing how many times you'll get hit by Torturer.

Discussion of Reactions

Moat and Horse Traders are easy: just force them to always be revealed to attacks. Although you don't always want to do this, usually you do. IRL, it's difficult to make this accountable (because your opponent can't tell the difference between you failing to reveal Moat and simply not having one in hand), but that's not a problem on a computer.

Fool's Gold's reaction is fine to delay.

Secret Chamber, Watchtower, and Trader are difficult to deal with. Trader could perhaps be auto-revealed to anything your opponent gives you, because almost never will they give you something better than a Silver. Watchtower is tough, because sometimes you really want to trash (Curse, Copper) and sometimes you really want to top-deck (Silver from Governor or from revealed Trader)--maybe it's best to delay the decision, so that in the meantime, the card is set aside. Secret Chamber seems almost impossible to adapt, because delaying the decision would make it useless against deck attacks, so the card would need a redesign.

Tunnel is obvious. Reveal it always. Has anybody ever not revealed Tunnel when they can? There are some problems with accountability here again, since in principle you can discard in a way that hides the Tunnel from your opponent.

An example of an interesting card interaction lost by these changes: not revealing Moat to a Mountebank so that you can reveal Trader instead.

Discussion of Opponent Choice

Some of these are no-brainers: for Tournament and Young Witch, auto-reveal always. Like with reactions, there would be accountability problems IRL. Mountebank is less clear-cut, but it's usually good to reveal against it, so auto-reveal for that card too.

Oracle and Rabble sometimes require re-ordering cards that go on your deck, but it rarely matters, so they could be put back in original order. (Remember that the reason re-ordering is usually allowed is that, IRL, it's hard to keep track of the original order. This isn't a problem on a computer.)

Discarding and put-back-on-deck effects (Bureaucrat, Militia, Torturer, Ghost Ship, Goons, Followers, Margrave, maybe Vault) can be delayed.

(Then there's a question of whether opponent-draws-a-card effects should also be moved to the beginning of opponent turn. I think they should, because the interaction of opponent-draws-a-card effects and discard attacks is important pretty often.)

Saboteur's gain, Governor's opponent-remodel, and Duchess's self-spy can all be delayed. For Governor and Saboteur, your opponent should get to resolve these effects even if the game end condition is fulfilled before their turn starts. (Note that moving Governor's effect can hurt a lot if you are defensively remodeling $7's into Provinces, so this loses some interesting situations.)

Masquerade, Contraband, and Envoy all seem impossible to adapt. With Masquerade, the pass could be removed, which makes the card much less interesting but at least preserves part of it. I don't see any way to deal with Contraband and Envoy.

Conclusion

Using the changes suggested above, nearly every card can be adapted somewhat reasonably to asynchronous play, with the loss of some interesting interactions. The exceptions are Masquerade, which needs its pass removed, and Contraband, Envoy, and Secret Chamber, which seem impossible to adapt.

Compiling this list made me realize how much worse Dominion would be without interaction on opponent turns. Sure, in most games it doesn't make any difference, but the games where it does matter are some of the most interesting games.

That said, the core of Dominion would still be there. It'd still be a fun game, albeit a little less so than normal.

[blueblimp]

Whoops, just noticed that I made a pretty big mistake: if effects like Ghost Ship's attack and Council Room's card draw are delayed to start-of-next-turn and reported one-by-one, then how can that be displayed reasonably in the interface, given that attacks like Rabble and Swindler actually occurred when played?

[Captain_Frisk]

Masquerade would still be a strong card without the passing.  Envoy could have programmed rules, but would get a boost which it most certainly does not need.  (If you have actions remaining - discard your highest costing action card.  If you have no action cards, or have no actions remaining - discard your highest costing treasure). 

Contraband would be really tough.  If you really wanted something similar - you could just say "You may not buy Gold, Platinum, Provinces or Colonies when this card is in play", but this card would be worse than contraband most of the time, and significantly better in the right situations (duke)

As you pointed out with the others - most of the time its ok.

Something you'd have to remember associated with delaying things like b-crat and ghost ship is the impact that it would have on cards like tribute and or jester.

Oh - and saboteur is one of the cases where auto reveal of trader would be a major kick in the teeth - Province-> Duchy -> Sorry you are gaining a silver sucka!

[eHalcyon]

I think you missed some interactions.  Bishop isn't on your list at the moment.

There are always edge cases (and maybe some not-so-edge cases) that make delaying reactions less acceptable.  I'll give a few examples.

You say FG reaction is fine to delay.  But what if the opponent has Thief?  OK, well, what the Thief gains could be decided asynchronously, at the start of the the target's turn (and the decision of which treasure to trash/gain is generally quite easy and could probably be automated).  But then there is the case where the attacker is able to draw that newly gained treasure on the same turn!  And in that case, it would make a difference whether you chose to use the FG reaction or not, which means the delay no longer works.

Governor remodel is probably not good to delay.  You might have a $7 in hand to gain a Province, which is a big deal.  Or you might have a $4 in hand to turn into a Duchy, which is also rather important for the current player to know.  (And reading onward, I see you also mention this a bit...)



For Contraband, maybe there could be a way to set up automatic rules that you can change on your turn (so early game you'd usually set it to ban Gold, later it would ban Province).  Set up several tiers in case they play multiple Contrbands.  Would be crude though.



Theese are some pretty interesting things to think about.  You know, the easiest way to do asynch Dominion is to simply ban all the cards that don't automate or delay well.  But would that make games really uninteresting?  Not sure.  But hey, those non-interactive boards come up in regular play too.  And some reactions (Moat, HT) can still be included.

[Kelume]

It seems that the main conflict occurs when the top of the opponent's deck changes non-mandatorily (FG, Ghost Ship, etc) followed by an attack that depends upon the top of the opponent's deck (Spy, Swindler, Thief, etc).

In that case, after the second attack is played, the game could present the turn to the opposing player (player B) as though it is his turn, asking for the decisions to be made, and then give control back to player A to finish his turn. This should occur more infrequently than just the single attack case.

The other issue is opponent gains or trashes which may have to be handled similarly; if the players want true async play, they'd have to ban those cards, I suppose.

[Captain_Frisk]

Cases where you might gain cards and use them are pretty rare - and while neat - are hardly game breaking if they weren't included.  The power levels of the cards would shift around some - but it would still be recognizable as dominion.

[blueblimp]

Masquerade would still be a strong card without the passing.  Envoy could have programmed rules, but would get a boost which it most certainly does not need.  (If you have actions remaining - discard your highest costing action card.  If you have no action cards, or have no actions remaining - discard your highest costing treasure).

Yep, Masquerade is still good without the passing, and maybe better overall. The reason is that if you have Masquerade and your opponent doesn't, your deck may be more trimmed, so the cards you send over may be worse than the ones your opponent sends you. On the other hand, you can't discard-then-Masq.

I'm not a fan of making complicated play rules. I think it's better to simply remove Envoy (since it's a promo anyway).

Contraband would be really tough.  If you really wanted something similar - you could just say "You may not buy Gold, Platinum, Provinces or Colonies when this card is in play", but this card would be worse than contraband most of the time, and significantly better in the right situations (duke)

This seems like a reasonable Contraband variant. It's not really Contraband anymore, just a card similar to Contraband, but that's probably OK.

Something you'd have to remember associated with delaying things like b-crat and ghost ship is the impact that it would have on cards like tribute and or jester.

Yes, the idea here is to give up these interactions. It's nearly impossible to preserve them, at least for Ghost Ship/Tribute. There's no way to choose the GS put-back automatically, and your opponent really can't play until they know what the Tribute hit.

The only way I see to do it is to transfer control to your opponent temporarily to choose what to put back for Ghost Ship. This is fine most of the time, but then Ghost Ship/Governor games become nearly unplayable. My goal here was to ensure every game is playable, at the expense of not faithfully following the rules.

Oh - and saboteur is one of the cases where auto reveal of trader would be a major kick in the teeth - Province-> Duchy -> Sorry you are gaining a silver sucka!

For Saboteur, since the choice of new card would be delayed to the start of your next turn, you'd have control over Trader. So this should be OK.

I think you missed some interactions.  Bishop isn't on your list at the moment.

Thanks, I missed Bishop. Seems like it can safely be delayed along with discards, draws, etc.

There are always edge cases (and maybe some not-so-edge cases) that make delaying reactions less acceptable.  I'll give a few examples.

You say FG reaction is fine to delay.  But what if the opponent has Thief?  OK, well, what the Thief gains could be decided asynchronously, at the start of the the target's turn (and the decision of which treasure to trash/gain is generally quite easy and could probably be automated).  But then there is the case where the attacker is able to draw that newly gained treasure on the same turn!  And in that case, it would make a difference whether you chose to use the FG reaction or not, which means the delay no longer works.

My general feeling is that it's okay for these sorts of interactions to break. They are really cool, but also pretty rare. It's nice to know what your Thief did when you played it, so delaying it is not so great.

Governor remodel is probably not good to delay.  You might have a $7 in hand to gain a Province, which is a big deal.  Or you might have a $4 in hand to turn into a Duchy, which is also rather important for the current player to know.  (And reading onward, I see you also mention this a bit...)

Yeah, this sometimes disadvantages the current player in Governor play, because of not knowing how many points the opponent will have. It also sometimes disadvantages the opponent if the current player takes all the remaining Provinces and now there aren't any left to upgrade $7's to. I don't think it should unbalance the card, but it does make it play differently.

For Contraband, maybe there could be a way to set up automatic rules that you can change on your turn (so early game you'd usually set it to ban Gold, later it would ban Province).  Set up several tiers in case they play multiple Contrbands.  Would be crude though.

Yes, this could be done. I prefer to avoid this sort of thing though, because choosing rules for things that might happen is pretty difficult and boring. It becomes comparable to writing simulator bots, which is more difficult and subtle than just playing the game.

Theese are some pretty interesting things to think about.  You know, the easiest way to do asynch Dominion is to simply ban all the cards that don't automate or delay well.  But would that make games really uninteresting?  Not sure.  But hey, those non-interactive boards come up in regular play too.  And some reactions (Moat, HT) can still be included.

Yeah, that's a good point. The number of cards I listed in the original post, plus Bishop, is only 27, so even if all of them got banned, there's still a lot of game left, although it's a little weird to never play with discard attacks. But as you point out, sometimes boards lack these cards regardless.

As a practical solution, this is probably better than changing the card rules.

It seems that the main conflict occurs when the top of the opponent's deck changes non-mandatorily (FG, Ghost Ship, etc) followed by an attack that depends upon the top of the opponent's deck (Spy, Swindler, Thief, etc).

In that case, after the second attack is played, the game could present the turn to the opposing player (player B) as though it is his turn, asking for the decisions to be made, and then give control back to player A to finish his turn. This should occur more infrequently than just the single attack case.

The other issue is opponent gains or trashes which may have to be handled similarly; if the players want true async play, they'd have to ban those cards, I suppose.

Yeah, the thing is here that you run into the risk of having unplayable games. It's not bad if it just transfers control once per turn (although even then, that could as much as double the number of control transfers in the game). But if you're doing something like alternating Governors and Ghost Ships, or playing a Margrave chain, then the game becomes really awful.