26
Show Posts
This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
29
Council Room Feedback / Tentative time councilroom comes up, July 5th
« on: July 02, 2012, 02:26:01 pm »
If all goes well, it should be up Thursday evening.
If any enterprising individual wants to run the code and host a publically available web server (even on just the last week of logs), I'll happily point the councilroom.com DNS entry at their server in the interim. It really shouldn't be too hard to get running on a standard ubuntu machine.
If any enterprising individual wants to run the code and host a publically available web server (even on just the last week of logs), I'll happily point the councilroom.com DNS entry at their server in the interim. It really shouldn't be too hard to get running on a standard ubuntu machine.
30
Feedback / This working?
« on: June 26, 2012, 10:58:15 am »
I think I've successfully migrated the forum to a server not in my apartment, so this should be much more reliable for you folks with my coming move.
31
Dominion General Discussion / Trying to infer card similarities, some late night councilroom hacking
« on: June 18, 2012, 02:19:09 am »
First, two very big, fairly similar looking pictures.
http://mine.councilroom.com/~rrenaud/grouped_by_prob_wgag.png
http://mine.councilroom.com/~rrenaud/grouped_by_wgag.png
And then some code.
https://github.com/rrenaud/dominionstats/commit/0a97f05263bc709b1f410639cb6843b62e8c5288
Basic idea, form vectors out of the card conditional win stats. Feed these vectors to hierarchical clustering algorithm. Look at pictures. Decide if groupings are sane (sometimes, throne room, king's court) or crazy (throne room, king's court, colony?).
I could perhaps compare this against Empathy's method here:
http://forum.dominionstrategy.com/index.php?topic=647.msg8951#msg8951
or even try combining them to get card features and play features.
Is this going to be useful for anything? I am not sure. It could perhaps help in defining useful encoding of game states for machine learning algorithms.
http://mine.councilroom.com/~rrenaud/grouped_by_prob_wgag.png
http://mine.councilroom.com/~rrenaud/grouped_by_wgag.png
And then some code.
https://github.com/rrenaud/dominionstats/commit/0a97f05263bc709b1f410639cb6843b62e8c5288
Basic idea, form vectors out of the card conditional win stats. Feed these vectors to hierarchical clustering algorithm. Look at pictures. Decide if groupings are sane (sometimes, throne room, king's court) or crazy (throne room, king's court, colony?).
I could perhaps compare this against Empathy's method here:
http://forum.dominionstrategy.com/index.php?topic=647.msg8951#msg8951
or even try combining them to get card features and play features.
Is this going to be useful for anything? I am not sure. It could perhaps help in defining useful encoding of game states for machine learning algorithms.
32
Dominion General Discussion / A little running joke between me and theory
« on: May 25, 2012, 05:10:32 pm »
Every once in awhile, when he has a good idea related to X, I tell him he should run Xstrategy.com site. When he told me how he proposed, I told him he should run ProposalStrategy.com.
But he got one upped, this dude should run DominionProposalStrategy.com.
http://boardgamegeek.com/thread/807953/how-i-proposed-using-dominion
But he got one upped, this dude should run DominionProposalStrategy.com.
http://boardgamegeek.com/thread/807953/how-i-proposed-using-dominion
33
Feedback / Recommended web hosting for forum.dominionstrategy.com
« on: April 24, 2012, 11:47:10 pm »
I am going to move in a couple of months, and part of this move will involve me lacking internet access for awhile.
Since forum.dominionstrategy.com is served from my home machine, this is bad for you guys.
So I'd like to get some reasonable web hosting for the forum and transfer the forum there so you don't have to suffer downtime (councilroom will go down, just cause I really like having total control of the machine serving it, unless someone wants to set up a temporary mirror). I think all we need is a conventional PHP/mysql/email hosting setup, and we don't really use that much bandwidth, since the site performance is mostly fine and it's served off a cable modem. I have a strong personal preference towards Linux, or at least Unix based hosting. Anyone have any good and reasonable hosting providers?
Since forum.dominionstrategy.com is served from my home machine, this is bad for you guys.
So I'd like to get some reasonable web hosting for the forum and transfer the forum there so you don't have to suffer downtime (councilroom will go down, just cause I really like having total control of the machine serving it, unless someone wants to set up a temporary mirror). I think all we need is a conventional PHP/mysql/email hosting setup, and we don't really use that much bandwidth, since the site performance is mostly fine and it's served off a cable modem. I have a strong personal preference towards Linux, or at least Unix based hosting. Anyone have any good and reasonable hosting providers?
34
Dominion General Discussion / Help a stats loving brother out, list combos, counters, nombos
« on: April 20, 2012, 03:35:41 pm »
So this is inspired by wanting to make councilroom better for this purpose:
http://forum.dominionstrategy.com/index.php?topic=2301.0
If I had some training data, I could try to optimize the card (combo) ranking function. The existing ranking is based on QVist's single card data.
This is sort of in the spirit of the QVist community rankings, except without anyone doing the hard work of aggregating and building a uniform ranking and providing commentary. Of course, if someone wanted to take something akin to his role, I'd happily let this morph into that if someone wants to take the job.
Basically, list a card, followed by : +followed by cards that you think make it better, ?followed by cards that you think are mostly independent/non-interactive, - followed by cards that you think make it worse.
For example,
Mountebank: +King's Court, Scheme, Throne Room, Chapel, Village ? Caravan, Apprentice, -Counting House, Gardens, Ambassador, Trader
So I like Mountebank more in games with KC, Scheme, Throne Room, and Chapel, Village, I am mostly indifferent to the presence of Caravan or Apprentice, and I like it less in Counting House, Gardens, Ambassador, and Trader games.
This is just overall, in aggregate. Don't try to stretch to hard to find some conceivable reason for the indifferent cards to matter. For example, KC/Caravan/Mountebank might work better than just KC/Mountebank by itself, but mostly Caravan won't influence my decision for Mountebank, so I list it as indifferent.
Broad coverage is more useful than different people telling me 50x that KC and Bridge are awesome together. It's also important to know about the more generic and less interesting combinations. Eg, University/Bazaar is a fine and good combo, even if it's not exciting.
It's also useful for you to be unbiased. Try to come up with your own thing without looking at others feedback or councilroom.
http://forum.dominionstrategy.com/index.php?topic=2301.0
If I had some training data, I could try to optimize the card (combo) ranking function. The existing ranking is based on QVist's single card data.
This is sort of in the spirit of the QVist community rankings, except without anyone doing the hard work of aggregating and building a uniform ranking and providing commentary. Of course, if someone wanted to take something akin to his role, I'd happily let this morph into that if someone wants to take the job.
Basically, list a card, followed by : +followed by cards that you think make it better, ?followed by cards that you think are mostly independent/non-interactive, - followed by cards that you think make it worse.
For example,
Mountebank: +King's Court, Scheme, Throne Room, Chapel, Village ? Caravan, Apprentice, -Counting House, Gardens, Ambassador, Trader
So I like Mountebank more in games with KC, Scheme, Throne Room, and Chapel, Village, I am mostly indifferent to the presence of Caravan or Apprentice, and I like it less in Counting House, Gardens, Ambassador, and Trader games.
This is just overall, in aggregate. Don't try to stretch to hard to find some conceivable reason for the indifferent cards to matter. For example, KC/Caravan/Mountebank might work better than just KC/Mountebank by itself, but mostly Caravan won't influence my decision for Mountebank, so I list it as indifferent.
Broad coverage is more useful than different people telling me 50x that KC and Bridge are awesome together. It's also important to know about the more generic and less interesting combinations. Eg, University/Bazaar is a fine and good combo, even if it's not exciting.
It's also useful for you to be unbiased. Try to come up with your own thing without looking at others feedback or councilroom.
35
Feedback / Translated versions of dominionstrategy on the web
« on: April 09, 2012, 05:08:03 pm »
I don't know if this is cool, shady, or an opportunity for us.
http://tesera.ru/article/dominion_guide_base/
Instead of complaining, maybe we should ask the community for and host translated versions of the articles?
http://tesera.ru/article/dominion_guide_base/
Instead of complaining, maybe we should ask the community for and host translated versions of the articles?
36
Dominion Isotropic / Log says curses gained on Possesion turn with embargo when supply empty
« on: April 01, 2012, 01:41:21 pm »
Credit to drheld for finding this bug. There aren't any curses left to be gained on the possession turn.
http://councilroom.com/game?game_id=game-20120329-201642-64791f44.html#aredoubles-show-poss-turn-18-1
curse 1
--- aredoubles's turn 6 ---
aredoubles plays a Mountebank.
... getting +$2.
... dingding gains a Copper and a Curse.
curse 2
--- dingding's turn 6 ---
dingding plays a Familiar.
... drawing 1 card and getting +1 action.
... aredoubles gains a Curse.
curse 3
--- aredoubles's turn 8 ---
aredoubles plays a Familiar.
... drawing 1 card and getting +1 action.
... dingding gains a Curse.
curse 4 and 5
--- dingding's turn 8 ---
dingding plays a Bazaar.
... drawing 1 card and getting +2 actions and +$1.
dingding plays a Familiar.
... drawing 1 card and getting +1 action.
... aredoubles gains a Curse.
dingding plays a Familiar.
... drawing 1 card and getting +1 action.
... aredoubles gains a Curse.
curse 6
--- aredoubles's turn 10 ---
aredoubles plays a Bazaar.
... drawing 1 card and getting +2 actions and +$1.
aredoubles plays a Familiar.
... drawing 1 card and getting +1 action.
... dingding gains a Curse.
curse 7
--- dingding's turn 10 ---
dingding plays a Familiar.
... drawing 1 card and getting +1 action.
... aredoubles gains a Curse.
curse 8
--- dingding's turn 11 ---
dingding plays a Familiar.
... drawing 1 card and getting +1 action.
... aredoubles gains a Curse.
curse 9
--- dingding's turn 12 ---
dingding plays a Familiar.
... drawing 1 card and getting +1 action.
... aredoubles gains a Curse.
curse 10
--- dingding's turn 14 ---
dingding plays a Familiar.
... drawing 1 card and getting +1 action.
... aredoubles gains a Curse.
and then the bug
--- dingding's turn (possessed by aredoubles) ---
dingding plays a Market.
... drawing 1 card and getting +1 action, +1 buy, and +$1.
dingding plays 3 Coppers.
dingding buys a Smithy.
... ... aredoubles gains the Curse.
... aredoubles gains the Smithy.
(dingding reshuffles.)
(dingding draws: a Copper, a Familiar, 2 Silvers, and a Curse.)
http://councilroom.com/game?game_id=game-20120329-201642-64791f44.html#aredoubles-show-poss-turn-18-1
curse 1
--- aredoubles's turn 6 ---
aredoubles plays a Mountebank.
... getting +$2.
... dingding gains a Copper and a Curse.
curse 2
--- dingding's turn 6 ---
dingding plays a Familiar.
... drawing 1 card and getting +1 action.
... aredoubles gains a Curse.
curse 3
--- aredoubles's turn 8 ---
aredoubles plays a Familiar.
... drawing 1 card and getting +1 action.
... dingding gains a Curse.
curse 4 and 5
--- dingding's turn 8 ---
dingding plays a Bazaar.
... drawing 1 card and getting +2 actions and +$1.
dingding plays a Familiar.
... drawing 1 card and getting +1 action.
... aredoubles gains a Curse.
dingding plays a Familiar.
... drawing 1 card and getting +1 action.
... aredoubles gains a Curse.
curse 6
--- aredoubles's turn 10 ---
aredoubles plays a Bazaar.
... drawing 1 card and getting +2 actions and +$1.
aredoubles plays a Familiar.
... drawing 1 card and getting +1 action.
... dingding gains a Curse.
curse 7
--- dingding's turn 10 ---
dingding plays a Familiar.
... drawing 1 card and getting +1 action.
... aredoubles gains a Curse.
curse 8
--- dingding's turn 11 ---
dingding plays a Familiar.
... drawing 1 card and getting +1 action.
... aredoubles gains a Curse.
curse 9
--- dingding's turn 12 ---
dingding plays a Familiar.
... drawing 1 card and getting +1 action.
... aredoubles gains a Curse.
curse 10
--- dingding's turn 14 ---
dingding plays a Familiar.
... drawing 1 card and getting +1 action.
... aredoubles gains a Curse.
and then the bug
--- dingding's turn (possessed by aredoubles) ---
dingding plays a Market.
... drawing 1 card and getting +1 action, +1 buy, and +$1.
dingding plays 3 Coppers.
dingding buys a Smithy.
... ... aredoubles gains the Curse.
... aredoubles gains the Smithy.
(dingding reshuffles.)
(dingding draws: a Copper, a Familiar, 2 Silvers, and a Curse.)
37
Dominion General Discussion / Ambassador Chan is a certain one - curse pat!
« on: March 25, 2012, 10:26:17 pm »
I have this strange liking to read foreign language web pages that link to councilroom after they have been Google translated to English.
"What surprised me was that the Curse is not zero Buys
I'm probably not miss operation"
"When you run out of mountain'd buy 3 "
http://kohada.2ch.net/test/read.cgi/cgame/1331366540/
"What surprised me was that the Curse is not zero Buys
I'm probably not miss operation"
"When you run out of mountain'd buy 3 "
http://kohada.2ch.net/test/read.cgi/cgame/1331366540/
38
Council Room Feedback / Steal drheld's point counter deck tracking UI for game pages?
« on: March 09, 2012, 02:57:25 pm »
drheld's modified UI for the iso text pages is pretty nice. It's probably better than what I have now on the game pages.
I should probably steal it.
I should probably steal it.
39
Council Room Feedback / New Councilroom feature: goal page and some new goals
« on: February 12, 2012, 01:00:58 am »40
Council Room Feedback / New councilroom feature: historical level/#games graphs
« on: February 11, 2012, 12:12:48 am »41
Council Room Feedback / Correct for amount money earned bias in win by turn graphs
« on: January 27, 2012, 12:54:31 am »
Basically, $3s and $4s look bad in mid game not because they are bad, but because earning only $3 or $4 in the mid game is itself a bad sign. Compute the win rate given turn # and money earned, and provide the option to correct for this factor on the graph.
42
Dominion General Discussion / MOVED: Vanity Accomplishments
« on: January 26, 2012, 11:34:18 am »
This topic has been moved to Council Room Feedback.
http://forum.dominionstrategy.com/index.php?topic=1570.0
http://forum.dominionstrategy.com/index.php?topic=1570.0
43
Council Room Feedback / Cute idea? Piggy back on qvist's rankings for pop buys/supply win
« on: January 25, 2012, 02:08:27 pm »
Especially once Qvist is done with his report on the community rankings, there is a nice golden dataset for card rankings.
Using that data as the desired outcome, I (we, you?) could do some regression to find weights from features like percentage of games a card was purchased, win rate given any purchased, win rate per purchase, average gains per game. This would determine how important each feature is to the rankings. Then using those weights, for example, we could compute personalized rankings (rankings given a single players stats) and card synergies (change in score given a new card is available).
Using that data as the desired outcome, I (we, you?) could do some regression to find weights from features like percentage of games a card was purchased, win rate given any purchased, win rate per purchase, average gains per game. This would determine how important each feature is to the rankings. Then using those weights, for example, we could compute personalized rankings (rankings given a single players stats) and card synergies (change in score given a new card is available).
44
Council Room Feedback / New Councilroom feature: optimal card ratios
« on: January 23, 2012, 09:58:14 am »
This feature is due almost entirely to http://councilroom.com/player?player=tlstyer. He came from the internet came and gave me a nice implementation of an idle idea I had a year ago. Yay open source!
This tries to answer question, I am going for an X/Y combo, what order should I buy the cards? It contains two heatmaps of how often players in a particular X/Y part of the combo win. One for the final game state (what should my deck look like when this is over), and one for all intermediate game states (how should I get it there).
http://councilroom.com/optimal_card_ratios?card_x=Ambassador&card_y=Silver (everyone's favorite question)
http://councilroom.com/optimal_card_ratios?card_x=Tunnel&card_y=Warehouse
http://councilroom.com/optimal_card_ratios?card_x=Gardens&card_y=Workshop
http://councilroom.com/optimal_card_ratios?card_x=Potion&card_y=Alchemist
This tries to answer question, I am going for an X/Y combo, what order should I buy the cards? It contains two heatmaps of how often players in a particular X/Y part of the combo win. One for the final game state (what should my deck look like when this is over), and one for all intermediate game states (how should I get it there).
http://councilroom.com/optimal_card_ratios?card_x=Ambassador&card_y=Silver (everyone's favorite question)
http://councilroom.com/optimal_card_ratios?card_x=Tunnel&card_y=Warehouse
http://councilroom.com/optimal_card_ratios?card_x=Gardens&card_y=Workshop
http://councilroom.com/optimal_card_ratios?card_x=Potion&card_y=Alchemist
45
Dominion Isotropic / Level 6s pulling 4 provinces in 13 turns
« on: January 17, 2012, 02:47:20 pm »
What is this crap?
The level of play on iso is getting pretty damn good.
The level of play on iso is getting pretty damn good.
46
Council Room Feedback / New Councilroom feature: card conditional winning stats
« on: January 15, 2012, 06:05:43 pm »
This thing currently sucks, but hey, the perfect is the enemy of the crappy. And/or real artists ship. Alternatively, better now than never. You get the point.
http://councilroom.com/supply_win
There are some bugs in the stats with the cards that are in every set (estate, gold, duchy, etc).
The UI is a terrible, painful, undesigned mess.
You can't post stable links to how your page looks.
But the stats might be interesting! And the frontend is all javascript and there is an JSON API to get at the raw data, so some dedicated people could design a useful interface without having to run the councilroom backend.
First insight? It looks like Remodel helps FG more than Mint does, despite Remodel's general mediocrity and the crazy strength of the Mint/FG opening. This supports WW and Fabian's play in one of the tournament games.
I'd love suggestions for a good way to interact with the data.
http://councilroom.com/supply_win
There are some bugs in the stats with the cards that are in every set (estate, gold, duchy, etc).
The UI is a terrible, painful, undesigned mess.
You can't post stable links to how your page looks.
But the stats might be interesting! And the frontend is all javascript and there is an JSON API to get at the raw data, so some dedicated people could design a useful interface without having to run the councilroom backend.
First insight? It looks like Remodel helps FG more than Mint does, despite Remodel's general mediocrity and the crazy strength of the Mint/FG opening. This supports WW and Fabian's play in one of the tournament games.
Name Condition %+ Per gain Any gain Num gained
Fool's Gold None 64.6 1.02 ± 0 0.97 ± 0 2.7
Fool's Gold Remodel 74.9 1.06 ± 0.01 0.98 ± 0.01 3.32
Fool's Gold Mint 69.3 1.06 ± 0.01 0.97 ± 0.01 3.15
I'd love suggestions for a good way to interact with the data.
47
Dominion Videos and Streams / Annotated game logs vs game videos?
« on: January 12, 2012, 02:07:12 pm »
I am sure many of us enjoy WW's videos, but I am wondering how they would compare to textually annotated game logs.
Imagine having an iso log where you keep pressing a key to see the game get played turn by turn, and textual comments made during the video show at the appropriate times. Obviously, the display quality of the text would be much better, and you can see way more of the game history. Likewise, there are no annoying pauses when opponents are making long decisions, and there is a convenient place to anchor discussion (turn 2 discussion occurs right on turn 2).
On the other hand, the logs from iso are limited, you don't get to see as much in the logs as you do during the game (maybe the official version will fix this? fingers crossed). We certainly wouldn't get the stream of consciousness thing going on, which could be a loss. And it's more work, I would still need to write the software, and transcribing the videos would take some work.
Imagine having an iso log where you keep pressing a key to see the game get played turn by turn, and textual comments made during the video show at the appropriate times. Obviously, the display quality of the text would be much better, and you can see way more of the game history. Likewise, there are no annoying pauses when opponents are making long decisions, and there is a convenient place to anchor discussion (turn 2 discussion occurs right on turn 2).
On the other hand, the logs from iso are limited, you don't get to see as much in the logs as you do during the game (maybe the official version will fix this? fingers crossed). We certainly wouldn't get the stream of consciousness thing going on, which could be a loss. And it's more work, I would still need to write the software, and transcribing the videos would take some work.
48
Dominion General Discussion / Ill-Gotten Gains - bad for the game?
« on: January 03, 2012, 04:57:45 pm »
Games where Ill-Gotten Gains is the dominant strategy are uninteresting and not fun. This is not complaining that I lose to IGG, or that I don't know how to play it. It's my 5th strongest win given avail. I win more often than I usually do both when I gain it and when I don't. I am not complaining that it's too good. It's somewhat frequently ignorable (maybe 30% of games) and when it's a good but not dominating option (maybe another 30%), it can be fun. I am not complaining that it's too strong.
But when it's clearly the card to go for in the set, the game is just kind of crappy. I feel sorry for my poor opponents when I IGG spam them, and it disrupts what would otherwise be a nice engine. It feels different than a gardens rush say, where my opponent has the choice to go for an engine, or go for a gardens spammed deck. When I go for an IGG, my opponent has no choice but to get spammed, and I don't have a lot of interesting decisions to make. It leaves me bored and my opponent frustrated.
IGG, it's the new Minion, only worse.
But when it's clearly the card to go for in the set, the game is just kind of crappy. I feel sorry for my poor opponents when I IGG spam them, and it disrupts what would otherwise be a nice engine. It feels different than a gardens rush say, where my opponent has the choice to go for an engine, or go for a gardens spammed deck. When I go for an IGG, my opponent has no choice but to get spammed, and I don't have a lot of interesting decisions to make. It leaves me bored and my opponent frustrated.
IGG, it's the new Minion, only worse.
49
Simulation / Simulations I'd like to see
« on: January 03, 2012, 02:19:31 pm »
(It's so nice to be the one requesting, rather than providing data 
).
Following WW's Silk Roads article. How do Silk Road combos compare to Gardens combos? Not Gardens vs Silk Roads, but (Gardens/X vs Y) vs (Silk Road/X vs Y). Maybe Y is Smithy/money, or maybe witch/money. Maybe X is iron works, baron, workshop, etc.
Tunnel/Warehouse? How do you do it? When I play this, I tend to open warehouse/warehouse, figuring I am somewhat likely to be able to get a 3rd turn tunnel, and then I've got plenty of warehouses to discard the tunnel with. But most players open Warehouse/tunnel.
).Following WW's Silk Roads article. How do Silk Road combos compare to Gardens combos? Not Gardens vs Silk Roads, but (Gardens/X vs Y) vs (Silk Road/X vs Y). Maybe Y is Smithy/money, or maybe witch/money. Maybe X is iron works, baron, workshop, etc.
Tunnel/Warehouse? How do you do it? When I play this, I tend to open warehouse/warehouse, figuring I am somewhat likely to be able to get a 3rd turn tunnel, and then I've got plenty of warehouses to discard the tunnel with. But most players open Warehouse/tunnel.
50
Game Reports / One Vault, Two Tacticians, and Three Cellars: Pure, simple, and consistent
« on: December 31, 2011, 11:55:55 am »
We all spend a lot of time thinking about Dominion strategies. Vault and Tactician would be great together, if only we could consistently draw them together. And Cellar is usually a lesser Warehouse, unless we have a big hand of mostly crap, then it's much, much better than a Warehouse. Hey, if we just Tacticianed, and there is no trashing, then our hand is likely full of lots of crap.
But only very rarely do we get to execute such a theoretically pure and consistent strategy. Any particular 3 card combo is very unlikely to show up in a particular random set. We are usually distracted by building up buying power and support for our intended combos, if there is even a combo to be played. We have the late game slogging through Duchy buys. Occasionally even having to react to our opponent. We make $6 and then $4, when we really wanted $5 and $5.
Vault, Tactician, Cellar. An uber consistent $8 per turn. Start with a Tactician, Cellar off many, find a Vault and a Tactician, Vault to one card left, play a Tactician, repeat. Turn 10 to Turn 17, one Province per turn.
http://councilroom.com/game?game_id=game-20111230-121121-093b1476.html
But only very rarely do we get to execute such a theoretically pure and consistent strategy. Any particular 3 card combo is very unlikely to show up in a particular random set. We are usually distracted by building up buying power and support for our intended combos, if there is even a combo to be played. We have the late game slogging through Duchy buys. Occasionally even having to react to our opponent. We make $6 and then $4, when we really wanted $5 and $5.
Vault, Tactician, Cellar. An uber consistent $8 per turn. Start with a Tactician, Cellar off many, find a Vault and a Tactician, Vault to one card left, play a Tactician, repeat. Turn 10 to Turn 17, one Province per turn.
http://councilroom.com/game?game_id=game-20111230-121121-093b1476.html