Dominion Strategy Forum

Please login or register.

Login with username, password and session length
Pages: [1] 2 3  All

Author Topic: Which Cards Are Most Luck-Dependent?  (Read 18607 times)

0 Members and 1 Guest are viewing this topic.

Matt_Arnold

  • Thief
  • ****
  • Offline Offline
  • Posts: 90
  • Designer of "Overworld" by Magic Meeple Games.
  • Respect: +47
    • View Profile
Which Cards Are Most Luck-Dependent?
« on: June 04, 2012, 12:35:05 am »
0

Of course, it's a card game. There will be shuffle luck. Granted. But let's say you're allowed to choose your 2-5 or 3-4 opening split. For purposes of a tournament, which cards cause the outcome to depend on luck the most?
Logged

Robz888

  • Margrave
  • *****
  • Offline Offline
  • Posts: 2644
  • Shuffle iT Username: Robz888
  • Respect: +3388
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #1 on: June 04, 2012, 12:37:36 am »
0

Well, regarding the opening... it's Familiar, I think. That's where you are really crushed for bad luck.

There are also 2 cards I mostly like but they can be ragefully luck-reliant on Turns 3/4.... Baron and Salvager. Drawing/not drawing them with Estate is so frustratingly important. And often times they are strong openings on the board, so it's not like it's usually avoidable.
Logged
I have been forced to accept that lackluster play is a town tell for you.

michaeljb

  • Saboteur
  • *****
  • Offline Offline
  • Posts: 1422
  • Shuffle iT Username: michaeljb
  • Respect: +2113
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #2 on: June 04, 2012, 12:41:07 am »
0

Swindler--I've had games where I hit 3 Estates in a row, and I had one game where I hit 5 of my opponent's Coppers. That's real bad cuz Swindler's cursing does more to damage economy than even Sea Hag's--those Coppers are important early on.

King's Court, race to be the first to draw $7, then race to be the first one to connect it with something.
Logged
🚂 Give 18xx games a chance 🚂

cayvie

  • Explorer
  • *****
  • Offline Offline
  • Posts: 317
  • old
  • Respect: +235
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #3 on: June 04, 2012, 01:02:53 am »
0

I think Tournament is very luck-dependent.

I think this even though:

it is my highest % buy (96.4) (obviously too high)

it is my worst effect without (-39.86)

and i have a perfectly solid effect with (1.25)
Logged
18:28 MEASURE YOUR LIFE IN LOVE: you shouldve done the decent thing and resign rather than go on being that lucky all the time

she/her

Dsell

  • Saboteur
  • *****
  • Offline Offline
  • Posts: 1290
  • He/Him
  • Respect: +932
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #4 on: June 04, 2012, 01:52:33 am »
0

There are also 2 cards I mostly like but they can be ragefully luck-reliant on Turns 3/4.... Baron and Salvager. Drawing/not drawing them with Estate is so frustratingly important. And often times they are strong openings on the board, so it's not like it's usually avoidable.

Didn't you say before that you hate the luck factor in tournament? It's suspiciously absent from this list!!

Totally kidding. ;D (Hope that's not crossing any rules lines...)

Actually I totally agree and would add to that list Remake. A hand of Remake-Copper-Copper-Copper-Copper can really really suck.

Also I would add by far one of the most luck-dependent cards: Black Market. Oh you got the Mountebank and the Chapel? Well I got the CARAVAN!! WOO!

Honorable mention goes to Possession. It's hard to hit 6P but it's not so bad in an engine, where possession can benefit you the most.
Logged
"Quiet you, you'll lynch Dsell when I'm good and ready" - Insomniac


Winner of Forum Survivor Season 2!

michaeljb

  • Saboteur
  • *****
  • Offline Offline
  • Posts: 1422
  • Shuffle iT Username: michaeljb
  • Respect: +2113
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #5 on: June 04, 2012, 02:25:22 am »
0

Actually I totally agree and would add to that list Remake. A hand of Remake-Copper-Copper-Copper-Copper can really really suck.

But I usually like getting rid of 2 Coppers...but I suppose it's true that if you drew that, your Remake is missing the Estates, which is where you really want it, so that makes sense.

Also I would add by far one of the most luck-dependent cards: Black Market. Oh you got the Mountebank and the Chapel? Well I got the CARAVAN!! WOO!

I would love to see someone play Black Market and draw Fool's Gold, Treasure Map, and Peddler all at once.
Logged
🚂 Give 18xx games a chance 🚂

kn1tt3r

  • Minion
  • *****
  • Offline Offline
  • Posts: 585
  • Respect: +278
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #6 on: June 04, 2012, 03:18:01 am »
0

Swindler and Black Market feel very luck dependend, but it might actually be more severe for real power cards that need to be bought soon, activated, drawn together etc., like Familiar, King's Court and such. I'd also like to add Jester to the list - gaining power cards vs. hitting Silver or even Copper is just a huge swing.
Logged

Matt_Arnold

  • Thief
  • ****
  • Offline Offline
  • Posts: 90
  • Designer of "Overworld" by Magic Meeple Games.
  • Respect: +47
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #7 on: June 04, 2012, 07:39:48 am »
+1

Can we define this quality we're identifying, in such a way that it could be measured in simulations?

I mean, sure, Swindler has an element of luck to it. But is it really more so than any other topdeck-affecting attack? Isn't it just a stronger attack than, say, Thief? Not more luck-based-- just stronger. The question isn't about strength. Thief is still able to steal a Platinum. It happens if you're lucky, and there's that word again: luck. It's a card game. It's guaranteed that there will be more than zero luck.

Many cards reward you for having them in hand with a specific other card. Baron and Estate. Moneylender and Copper. Is this really what we want to call swinginess? I mean, that's what the game is about. That's why there exist Warehouse, Tactitian, name a support card and that's pretty much why it exists. For that matter, trashers support that type of card too. Maybe what we mean by saying Baron and Salvager are swingy, is that they affect the outcome of the game too early in the game? (I'm not sure that's true, but I'm looking for a definition of swinginess.)

So. What do you think is a good precise definition of swinginess? The type of definition that a computer could score as a percentage by analyzing statistics? Can an equation, algorithm, or what have you, identify the level of influence of skill over games containing a particular card?
Logged

BaruMonkey

  • Bishop
  • ****
  • Offline Offline
  • Posts: 110
  • Respect: +191
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #8 on: June 04, 2012, 08:22:24 am »
+1

I'm surprised nobody's mentioned Treasure Map as an option yet!
Logged

cherdano

  • Moneylender
  • ****
  • Offline Offline
  • Posts: 166
  • Respect: +54
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #9 on: June 04, 2012, 08:36:58 am »
0

I am afraid of posting in this thread since rrenaud could prove anything wrong with half an hour of coding...  8) ;)
Logged

greatexpectations

  • Saboteur
  • *****
  • Offline Offline
  • Posts: 1097
  • Respect: +1067
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #10 on: June 04, 2012, 08:40:58 am »
0

I am afraid of posting in this thread since rrenaud could prove anything wrong with half an hour of coding...  8) ;)

haha good call.  i wonder if our gracious hosts could give us some information about tournament.  % of the time P1 hits their tournament first, win rate of player who hits their tournament first, stuff along that line.
Logged
momomoto: ...I looked at the tableau and went "Mountebank? That's for jerks."
rrenaud: Jerks win.

popsofctown

  • Adventurer
  • ******
  • Offline Offline
  • Posts: 5477
  • Respect: +2860
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #11 on: June 04, 2012, 09:29:33 am »
0

Familiar is the only card I think is really clear cut, obvious, not helpful card for a tournament.  That 2p.

I would argue Treasure Map is a bad idea for tournaments, and actually during the last isodom I played I asked to forbid it (I'm older and wiser and would add Familiar).

Tournament is pretty controversial right now.
Logged

Captain_Frisk

  • Saboteur
  • *****
  • Offline Offline
  • Posts: 1257
  • Respect: +1263
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #12 on: June 04, 2012, 10:32:56 am »
+1

So - I think we're going to be unlikely to put forth a formula that lets us explicitly indicate which ones are more or less luck dependent.

In my mind we have several different cases of luck:

1.  Luck in ability to purchase the card because of its price

Prizes - The ultimate here - you can only gain them in very specific circumstances - and if your opponent does first, you can't gain it at all.
Possession - 6p is expensive.  It's hard to reliably get early, and if you get it late - the value of your possession is reduced.  As others have pointed out - this is mitigated if you go for possession in an engine deck - which is nice because your engine is more likely to play possession often.  It is also more likely to have your opponent want to quit dominion.
Golem - 4p is also expensive.  You want to wait a reshuffle before you go for the potion and the golem, but then you run the risk of overpaying for both of them.  Draw 6p?  Do you want the gold or the golem?
Familiar - I'm sure that Geronimo0 has simmed it, but I'm pretty sure that opening potion + $2 terminal/silver is the way to go here, but if you fail to buy the familiar on turn 3/4, then bad things happen.  Do you want to have some real rage?  You are 2nd player, your potion falls to turn 5 AND you draw 2p, and your opponent has already bought 2 familiars.
Alchemist - Not as bad as familiar - although just as hard to buy.  Since they are more of a long term play - its not as crippling to fail to buy them early.  Still not great though.
Forge - The ability to buy buy forge on turn 3 is obscenity.
King's Court - This will show up again - but as others have pointed out, getting 7 to buy KC is hard - but the value in pulling it off is crazy.
Goons - First one to goons has a decent advantage.
$5 Cards - There are several cards that are significantly more valuable if you can buy them early.  Sometimes you get lucky and can buy them on turn 1!  Trading Post, Mountebank, Witch, Torturer....
Mint - Did you happen to draw 5+ coppers?  Sweet!  Did you have silver + 3 coppers instead?  Sorry - you lose.  This is low on my list because it is rarely pivotal.  There is a special case of Mint / Fools Gold + 5/2 split, but we'll save fool's gold for later.

2.  Luck in ability to use the card to its fullest effect due to combinations
Tournament - I think this is self explanatory.  First you had to buy an early province.  Then you have to line it up with yours.  If you go heavy on tournaments, you can get lucky and prize sweep.  If you go heavy on tournaments and your opponent blocks it - you're f'd.  Unlike other lucky cards (treasure map is going to get its time) - ignoring tournament is generally a bad idea.
Remake - Lets look at two players.  They both open Remake / Silver.  On turn 3, player 1 draws Remake, 2 estates and 1 copper, and 1 silver.  He remakes his 2 estates into silvers, buys another silver, and draws his next hand: 5 coppers.  Player 2 draws Remake + 4 Coppers.  He trashes 2 coppers, and ends his turn.  He draws his next Hand - Estate x3, Copper x2.  Turn 4 - Player 1 Buys a Wharf.  Want to guess how this game is end?
Baron  Same as above.  One person uses baron to buy an early game forge.  The other person draws it with 4 coppers.
Steward - see above. 
Salvager - see above.
Chapel - see above.  The difference between "Ultimate Chapel" (Trash 3 estates and 1 copper, buy a Gold on Turn 4) and "Nightmare Chapel" - Trash 4 coppers, do nothing.

Young Witch - Best case - you draw the YW with your silver, buy a nice $5 card, and your opponent takes a curse.  Worse case - your YW doesn't get you to 5, your opponent blocks you with the bane.

Treasure Map - A turn 5 treasure map hit is obscene.  Turn 10?  Not so much.  This doesn't bother me much - because on average buying tmap is a bad idea.
King's Court -  Did you draw your KC with an action?  Awesome sauce.  A cursing action?  Good game sir.  Another king's court and a card drawer?  WOOOOOOOHOOOOO.  Did you draw it with treasures and estates?  Not so much.  Added bonus rage if you have multiple KC's, and you draw them dead together.
Throne Room - As above, but less exciting
$Village - Did you draw it with 2 actions?  Cool.  Did you draw all your villages in one hand, and your actions in another?  Boo.
Counting House  Did you draw it after a reshuffle?
Chancellor  Did you use it to buy a gold on T3 and reshuffle it?  AWESOME.  Did it show up on T3 with 3 estates? 


3.  Cards where the effect of playing them results in randomness.
Swindler - I swindle your estate into an estate.  You swindle my chapel into an estate.  I swindle your province into a province.  You swindle my duchy into a counting house.
Jester - I cloned your king's court.  You hit my silver. 

Finally - some bad cards that are random, but generally bad so its hard to have rage.
Saboteur
Tribute
Thief
Pirate Ship

Probably some more here, but I've posted enough for today.
Logged
I support funsockets.... taking as much time as they need to get it right.

WanderingWinder

  • Adventurer
  • ******
  • Offline Offline
  • Posts: 5275
  • ...doesn't really matter to me
  • Respect: +4381
    • View Profile
    • WanderingWinder YouTube Page
Re: Which Cards Are Most Luck-Dependent?
« Reply #13 on: June 04, 2012, 10:38:12 am »
0

I am not so sure that tribute is bad...

kn1tt3r

  • Minion
  • *****
  • Offline Offline
  • Posts: 585
  • Respect: +278
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #14 on: June 04, 2012, 10:49:50 am »
0

I am not so sure that tribute is bad...
In general Tribute is good if your opponent plays a similar strategy. It's ok in engines if you need actions, but that only works if your opponent also has a high action density, and it's quite great in Big Money, but only if your opponent doesn't play a competing engine deck.

Tributes big problem is of course its unreliabilty, which certainly is much less of an issue in BM strategies.
Logged

Captain_Frisk

  • Saboteur
  • *****
  • Offline Offline
  • Posts: 1257
  • Respect: +1263
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #15 on: June 04, 2012, 12:44:00 pm »
0

I am not so sure that tribute is bad...

Sorry - that was probably strong, my posting got sloppier and sloppier - AND I forgot Minion... which can either destroy or save a hand.

As for Tribute:

Overall is is purchased 43% of the time, for a win rate with of 0.92 and a win rate without of 1.06.

You purchase it even less - 34% of the time.  You win with it @ 1.07 vs. 1.18 without - both lower than your average win rate of 1.19 (win rate given provinces available)

Maybe calling it a bad card isn't so strong after all....

Logged
I support funsockets.... taking as much time as they need to get it right.

jonts26

  • Margrave
  • *****
  • Offline Offline
  • Posts: 2746
  • Shuffle iT Username: jonts
  • Respect: +3668
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #16 on: June 04, 2012, 12:55:09 pm »
0

Tribute isn't that bad. It's main strength is against big money. And you can use it either in an engine or big money yourself.

And I still have to say you have to use council room stats with a grain of salt. The community stats as a whole are skewed by lower level players not playing cards correctly and individual stats are skewed by the fact that it includes ALL of a players games. WW was at one point not so good as he is now.
Logged

greatexpectations

  • Saboteur
  • *****
  • Offline Offline
  • Posts: 1097
  • Respect: +1067
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #17 on: June 04, 2012, 01:03:04 pm »
0

And I still have to say you have to use council room stats with a grain of salt. The community stats as a whole are skewed by lower level players not playing cards correctly and individual stats are skewed by the fact that it includes ALL of a players games. WW was at one point not so good as he is now.

regress regress regress
Logged
momomoto: ...I looked at the tableau and went "Mountebank? That's for jerks."
rrenaud: Jerks win.

Captain_Frisk

  • Saboteur
  • *****
  • Offline Offline
  • Posts: 1257
  • Respect: +1263
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #18 on: June 04, 2012, 01:13:28 pm »
+1

I would love it if rrenaud would add some filters to popular buys...

Namely - filter to only games between players of winrate 1 or greater.

Filter to a players last N games.
Logged
I support funsockets.... taking as much time as they need to get it right.

WanderingWinder

  • Adventurer
  • ******
  • Offline Offline
  • Posts: 5275
  • ...doesn't really matter to me
  • Respect: +4381
    • View Profile
    • WanderingWinder YouTube Page
Re: Which Cards Are Most Luck-Dependent?
« Reply #19 on: June 04, 2012, 01:16:20 pm »
0

Tribute isn't that bad. It's main strength is against big money. And you can use it either in an engine or big money yourself.

And I still have to say you have to use council room stats with a grain of salt. The community stats as a whole are skewed by lower level players not playing cards correctly and individual stats are skewed by the fact that it includes ALL of a players games. WW was at one point not so good as he is now.
Ironically, my recent winrate is much worse than my long-term one, because the people I play now are MUCH better.
CR stats in terms of overall card effectiveness are pretty meaningless.

Captain_Frisk

  • Saboteur
  • *****
  • Offline Offline
  • Posts: 1257
  • Respect: +1263
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #20 on: June 04, 2012, 01:24:47 pm »
+1

Tribute isn't that bad. It's main strength is against big money. And you can use it either in an engine or big money yourself.

And I still have to say you have to use council room stats with a grain of salt. The community stats as a whole are skewed by lower level players not playing cards correctly and individual stats are skewed by the fact that it includes ALL of a players games. WW was at one point not so good as he is now.
Ironically, my recent winrate is much worse than my long-term one, because the people I play now are MUCH better.
CR stats in terms of overall card effectiveness are pretty meaningless.

Agreed that they aren't perfect - but i think that its fair to say that that given a good player - the relative frequency with which they buy a particular card is probably a good indicator of card strength.

For example, your top 5 $5 cards in terms of % purchased are:
Mountebank, Ill Gotten Gains, Wharf, Witch, and Hunting Party

Your lowest 5 are: Outpost, Contraband, Counting House, Royal Seal, and Harvest.

I see a general "good vs. bad" discrepency there.  Perhaps "bad" is probably the wrong word... Situational is probably more reasonable... they just have fewer situations in which they are useful....
Logged
I support funsockets.... taking as much time as they need to get it right.

popsofctown

  • Adventurer
  • ******
  • Offline Offline
  • Posts: 5477
  • Respect: +2860
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #21 on: June 04, 2012, 01:40:51 pm »
+1

Village?  At that point you're getting close to "Silver is a variant card because sometimes it misses reshuffles and sometimes it doesn't".

Village is at the point where I feel like we're playing dominion, managing probability, and if we play best of 7 we're going to have a pretty fair result.  If you play a BM/X and Y game instead of a village game, both players are more likely to mirror eachother's strategies more closely, and that produces a sort of variance in itself.

Variance is how frequently a card causes the player with the mathematically superior strategy for the entire game to lose, that is how I define it.  It's hard to simulate that because, although you can easily make a bot with buy rules that are smarter than another bot's, it's arbitrary how superior the superior strategy is.  It depends on the strength of the players.   You'd have to get players ten ranks apart to create buy rules for specific kingdoms and then simulate and get that rank 30 player's bot buys Familiar and goes on to build a wisely proportioned Hamlet/Menagerie/Steward deck to win 55% of simulated games, but wins 70% of the games if Witch is the curser instead.  Something like that.  But most Dominion players don't know how to build a good bot.  And you'd still have a variance specific to those two player strengths, the variance always changes (though some permeate all levels of play and all types of skill gaps so much we collectively whine about them)
Logged

Captain_Frisk

  • Saboteur
  • *****
  • Offline Offline
  • Posts: 1257
  • Respect: +1263
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #22 on: June 04, 2012, 03:40:10 pm »
0

Village?  At that point you're getting close to "Silver is a variant card because sometimes it misses reshuffles and sometimes it doesn't".
I did have it lower down on my list. 

To me, it's more a question of "when does your engine take off".... we can both build identical engines - but the first person who can successfully get their villages lined up in the right place can make a huge difference - either in picking up the rest of the pieces of the engine - or in the special case of Torturer - driving the opponent from the hobby.

I also forgot to mention a 4th case of luck dependency - and those are cards that favor the person who wins the split.  Duke / Duchy... Fools Gold... Minion etc... Cursers etc.  You can go identical strategies, but if you end up on the losing side of the split, you may have just lost on account of luck.
Logged
I support funsockets.... taking as much time as they need to get it right.

chwhite

  • Saboteur
  • *****
  • Offline Offline
  • Posts: 1065
  • Respect: +442
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #23 on: June 04, 2012, 03:50:03 pm »
0

CR stats in terms of overall card effectiveness are pretty meaningless.

I couldn't disagree more.  While CR stats are not perfect, it's not like any other evidence we have at our disposal (primarily simulation and individual games) is perfect either.  CR stats are usually in the ballpark of "correct" and if you're careful with them (i.e. understanding WHY Silver has such a high "Win Rate Without") they're basically always useful.
Logged
To discard or not to discard?  That is the question.

chwhite

  • Saboteur
  • *****
  • Offline Offline
  • Posts: 1065
  • Respect: +442
    • View Profile
Re: Which Cards Are Most Luck-Dependent?
« Reply #24 on: June 04, 2012, 04:01:27 pm »
0

As for which cards are most luck-dependent, there are several categories I can think of.  You've got the $5 cards which are super-good to get on Turns 1/2, and thus magnify the luck of the opening split: Mountebank, Trading Post, Witch (and, conversely, cards which make 5/2 a death sentence, like Ambassador).  You've got cards which introduce heavy variance on Turns 3/4: Familiar, Swindler, Baron are all like this.  Deck inspection attacks tend to be high-variance too: Swindler and Jester are the worst offenders here since the other cards tend to not be good enough to matter, also Sea Hag because of the random discarding.  And then there are cards which reduce the game to one dimension, usually involving a pure rush (IGG, Fool's Gold- I don't count Minion here because the other Actions matter too much), to a lesser extent the strong BM+X cards might fit here but they can more often be a part of more diverse strategies, or avoided.

Once you get to Turn 5 and beyond the butterfly effect is less severe, so cards which appear to become swingy later (say, Tournament or Grand Market) are better than cards which swing right out of the gates.  You can usually mold your opening strategy to deal with them, so they're not nearly as bad.
Logged
To discard or not to discard?  That is the question.
Pages: [1] 2 3  All
 

Page created in 0.053 seconds with 20 queries.