26
Puzzles and Challenges / Re: Simultaneity
« on: March 12, 2012, 03:53:56 pm »
Tournament?
Not to turn this into a Game Reports thread or anything, but I can't help but think that opening Silver / Silver is better for the kind of megaturn that you were looking for.
My favorite example of this:
http://councilroom.com/game?game_id=game-20120215-091334-e85fae18.html#GewoonD-show-turn-10
My opponent plays a Governor for remodeling, and I gain an Inn on his turn with an empty deck, shuffling 6 Actions from the discard pile.
I did this recently. And immediately after, my opponent played a Fortune Teller; discarding all my carefully stacked cards.
In general, a player does not benefit during someone else's turn. It can happen, though.
On the other hand, at a cost of $6 this card competes with Gold, whereas Lab is frequently a great pickup in those hands that fall just short of $6. And if you end up using this card as a village very often, then boy was that an expensive village.
What about for a classic Goons Engine deck? I presume thats harder to simulate, as you need to be able to react and create the 3 pile ending when you're ahead. But I've always played that I don't buy copper till I think the game is going to end on this turn or the next. Is this sensible?
It's not quite that simple- if you've got lots of draw and trash you can often handle some coppers that'll just get discarded to your opponents' Goons anyway- but you're definitely on the right track. In practice, most players seem to dilute their Goons engines with copper and green far too soon, and my reluctance to do that is (I suspect) the major reason Goons is one of my best cards. I hesitate to give more detailed advice than that, because obviously it depends, but as a general rule of thumb you should be willing to forego more of your extra buys than you'd expect, in the engine-building stage at least.
When to Venture
A single Venture is strictly better than Silver in a pure money deck since you are guaranteed to hit $2 with every play (the $1 from Venture and Venture will find at least a Copper) and can be worth as much as $4 (assuming no Platinum of course).
Finally: why didn't this earn PileDriver for curses? I had all 10, and I won. Other cards give PileDriver even if you trash them later on, right?
Maybe it only counts it if you actually buy them, rather than have them given to you.
So say there's 1 Curse left in the pile, and you buy an Estate, which has 2 Embargo tokens on it. You have Trader in hand. Can you gain 2 Silvers?Yes. There are two things happening to the same player, so the player whose turn it is orders them. They are the same thing so whatever. Each one resolves with you picking to take Silver instead of Curse; you gain two Silvers.
Cache is not actually different. Even though Cache tells you to gain two Coppers, you can't gain two Coppers simultaneously; one has to get gained first, because things can happen when cards are gained that have to be ordered, and in any case they go to your discard pile in some order. So you can buy Cache with only one Copper in the supply and reveal Trader twice to get two Silvers.
I'd probably recommend using a mixture of random and fixed. To explore basic concepts, pick a card and make sure it gets at least one friend in the random bunch. So say, "this game is going to be a Gardens game, but we're going to reroll unless there's a +buy that doesn't trash".
"This game is going to explore goons engines, so we're rerolling until we get goons and a village."
Nice. I once gained all 5 prizes in a single turn with a single Province (the turn immediately after buying that Province)!
How would that not add any more information?The level is the mean skill minus three times the standard deviation. The upper bound would be the mean plus three times the standard deviation. If the figure would show both the mean and the level (=the lower bound) we would already know both the mean and the standard deviation, and hence could infer the upper bound. Drawing it separately would hence not add information as such.
I love isotropic, but my biggest gripe is that the logs don't let you fully reconstruct the game. (You play a Smithy and draw three cards -- which three cards?!)
It clearly isn't an issue of preventing the opponent from seeing private information: already, the game logs show e.g. what each player draws at the end of their turn.
Full game reconstruction is a minimum standard I think we should ask of the logs. I've been waiting for the official app and assuming it will offer this feature. If not, I think we should keep bugging them until they put it in.
CouncilRoom will be hugely affected, naturally, by the kind of logs outputted. It could be better than Isotropic, and it could also be worse.
Aaaargh, QWERTY, don't even get me started on that one. I hate not having learnt Dvorak when I first started using a computer. As a software developer, I type quite a lot and I'm sure Dvorak instead of Qwerty would lead to faster development time and less errors, but where do I find the time to switch over?
I really like the graphs -- now I can stop tracking my own trend manually.
Do you think the Leaderboard history curve would look too messy if it included also the mean TrueSkill rating in addition to the lower bound (=level)? Ideally I would want to see also the upper bound for the skill as I would find it more visually pleasing to see a curve with associated credibility intervals, but that would obviously not add any more information. The mean, however, would tell something that is not apparently visible in the level.