Dominion Strategy Forum

Please login or register.

Login with username, password and session length
Pages: 1 ... 8 9 [10]  All

Author Topic: The Best Dominion Cards List 2012 Ed.1: $5 cards  (Read 78548 times)

0 Members and 1 Guest are viewing this topic.

timchen

  • Minion
  • *****
  • Offline Offline
  • Posts: 704
  • Shuffle iT Username: allfail
  • Respect: +233
    • View Profile
Re: THE Dominion Card List(s): $5 cards
« Reply #225 on: February 03, 2012, 09:42:37 pm »
0

Ok, I guess one way to do this is to list some key aspects of the cards and rank them, and the final rank can be some arbitrary combination of all the aspects.

1. % of boards you have to get the card (Goons>KC; probably something like 99% and 75% IMO)
2. Given the card in the deck, how important the card is in the strategy and how it shapes up the strategy (Goons~KC)
This is probably harder to quantify, but to me it is plausible: strategy often evolves around both cards when they are present.
3. The uniqueness/irreplacibility of the function of the card (KC>Goons; the VP token part of Goons is unique; other aspects are not. KC is just unique.)
Apparently the definition of "best" cannot be this alone; but it should still be weighted IMO.
4. The ability to combo with other cards (KC>>Goons; KC even creates new strategy scopes for some otherwise underwhelming cards, while Goons pretty much just take advantage of the extra +action +buy and not much else)

I do think KC is more game-wrapping than Goons. My definition for game wrapping would be something like (2+3+4)*1, from the criteria above. If you rank using this "game-wrapibility", I imagine KC will come out at top; while cards like possession takes a large hit from 1 and 4. In words, possession is game wrapping, but it only is when it is useful, which is not that often, and it does not change how other cards work that much.

Okay, now I can produce arbitrarily many definitions of my "best" by fine-tuning the coefficients of my formula which still produces KC at top. I guess that probably doesn't quite count in the 3-4 you want to hear, though. :P

I can provide another possible definition here: How good a card (given its cost) is can be measured by how tolerable it is to raising its cost. i.e., for simplicity, let's assume a uniform shift of 2 is an ok measure; how do the numbers change when Goons cost $8 and KC $9? I would say the criteria 1 and 2 above will shrink considerably for Goons and less so for KC. This is due to the fact that KC can generate money and make itself easier to get. Using the same criteria, some of the best cards are chapel (obviously..), ambassador, and cursers.
Logged

ackack

  • Explorer
  • *****
  • Offline Offline
  • Posts: 302
  • Respect: +19
    • View Profile
Re: THE Dominion Card List(s): $5 cards
« Reply #226 on: February 03, 2012, 10:08:32 pm »
0

It sounds to me like the way you worded it should strongly favor Goons. If you remove KC from a KC deck, you have a deck that can't get mega-turns, but still has all the pieces to do something. If you remove Goons from a Goons deck, you have absolutely nothing...

Goons as a slight patch on Big Money seems pretty common to me. If you then look at the kingdom minus Goons, it is most likely that if that's what was best before, what will be best now is basically a Big Money deck plus a terminal or two. Quite similar. By contrast, there are a lot of engines that become sufficiently strong to be best only in the presence of King's Court. If you remove King's Court from those boards, often you will move from an enginey setup back to a slightly modified Big Money deck being best.
Logged

Fabian

  • 2012 Swedish Champion
  • *
  • Offline Offline
  • Posts: 666
  • Respect: +541
    • View Profile
Re: THE Dominion Card List(s): $5 cards
« Reply #227 on: February 03, 2012, 10:26:39 pm »
0

It's not like I'm trying to, or want to, start a fight here, but honestly the entire post just seems like it's reaaaally stretching. Am I alone here, really? Why are we ranking a card's strength based on how good it is given it's already in the deck (hi Possession)? What does it matter how unique a card's function is when determining how strong it is (hi Saboteur, hi Counting House)? Why is 4. not simply some function of 1.? A card that "can" combo really well, but only 1% of the time, is still a card that combos 1% of the time. What matters is still the amount of time it's actually useful, no? If KC synergizes (combos) with other cards so much better than Goons, why is it a card you want to get on ~75% of the boards you want to get Goons? More importantly, why does it matter?

I think "game warping" is a reasonable (part of) measure of a card's strength, and (unfortunately) not entirely easy to properly define. Cards that are really game warping (Ambassador ~90%, Chapel ~90%, Sea Hag ~80+%) tend to be pretty damn strong. I would suggest Goons is game warping maybe.. 60%? And KC is game warping.. 40%? 50%? Bridge type cards, cursers (given that you get to KC before the curses start to run out).. that kind of thing. To me it's just silly to say KC is game warping in a lot of games where you're "just" racing to Provinces. Unless you routinely play with Colony as a required card, that includes A LOT of games.

As far as the last paragraph goes, it doesn't sound like an attempt at a real argument to me. Let's assume all Dominion kingdom cards suddenly costs $2 more, what happens? Well for starters, BM just got a lot stronger. Apart from other obvious stuff, like cheaper cards being hurt more (a 67% increase for Ambassador hurts a lot more than a 29% increase for KC), I don't know what conclusions we're supposed to draw from this thought experiment while still being useful in real life. Why aren't we evaluating the actual cards rather than hypothetical similarish cards, anyway?
« Last Edit: February 03, 2012, 10:33:07 pm by Fabian »
Logged

ackack

  • Explorer
  • *****
  • Offline Offline
  • Posts: 302
  • Respect: +19
    • View Profile
Re: THE Dominion Card List(s): $5 cards
« Reply #228 on: February 03, 2012, 10:42:40 pm »
0

What matters is still the amount of time it's actually useful, no?

Obviously that's not all that matters, otherwise we'd be talking about Caravan and Lab as being among the very elite cards in the game. added: I just find "sort by % gained" to be a very incomplete metric. Sure, it's something to look at, but there's other stuff going on in my opinion.
Logged

Fabian

  • 2012 Swedish Champion
  • *
  • Offline Offline
  • Posts: 666
  • Respect: +541
    • View Profile
Re: THE Dominion Card List(s): $5 cards
« Reply #229 on: February 03, 2012, 10:50:18 pm »
0

I don't think that's fair ackack, I was saying what matters isn't how well a card combos with [something], it's how often it's actually useful, in response to allfail's point #4. To give a counter example, Native Village is awesome because it combos really well with Bridge, or Chancellor is awesome because it's sweet with Stash, or something. That still only happens in some very small % of games, which is why in reality Native Village and Chancellor are underwhelming cards. Sure, they're good/useful (#1) more often because of those combos, but they're not good because they combo well with other cards. #4 is simply a part of #1.
Logged

RisingJaguar

  • Minion
  • *****
  • Offline Offline
  • Posts: 527
  • Respect: +184
    • View Profile
Re: THE Dominion Card List(s): $5 cards
« Reply #230 on: February 03, 2012, 11:34:03 pm »
0

Fabian vs. the world, Goons vs. KC

1. I think its fair to use what % of games it is useable as a main measurement.  It probably answers a great chunk of the question as we have to consider the effect of the card on ALL games to determine the best, more useful card.  I do think there's more to consider though, as silver that is something that is almost useful (as I learn from WW) but its not overwhelming.  That's why a card like swindler, is probably better overall yet is taken in less games.

2. I think its fair to consider the overwhelming power of a card, the way it 'changes' a set or whatever people are calling it these days.  However, this can only depend on how often it is actually needed.  Lets take coppersmith, in my opinion underrated (because of its ability to be a huge source of money in engine games).  It is usually pretty bad in most situations, but when it is good, it has overwhelming power over a lot of $4. 

I know these are exaggerations, but they both have to be considered.

3.  Lastly, I think timchen was essentially alluding to something similar to expected return of a card in ALL games.  Amount of times its useful*power+amounts of times not useful*weakness, I'm a little confused by dominion math.  Unfortunately, there's a lot of subjectivity in rating "power/weakness" and to a lesser degree "times its useful/not useful".  People's arguments should consider both sides is all. :)

PS. I choose KC over Goons, but this is coming from an engine maker, so i'm quite biased. 
Logged

chwhite

  • Saboteur
  • *****
  • Offline Offline
  • Posts: 1065
  • Respect: +437
    • View Profile
Re: THE Dominion Card List(s): $5 cards
« Reply #231 on: February 03, 2012, 11:39:07 pm »
0

I think "game warping" is a reasonable (part of) measure of a card's strength, and (unfortunately) not entirely easy to properly define. Cards that are really game warping (Ambassador ~90%, Chapel ~90%, Sea Hag ~80+%) tend to be pretty damn strong. I would suggest Goons is game warping maybe.. 60%? And KC is game warping.. 40%? 50%? Bridge type cards, cursers (given that you get to KC before the curses start to run out).. that kind of thing. To me it's just silly to say KC is game warping in a lot of games where you're "just" racing to Provinces. Unless you routinely play with Colony as a required card, that includes A LOT of games.

I more or less agree with your general reasoning, but not necessarily your numbers.  Specifically, I'd say that both Goons and KC are "game-warping" more often than you give them credit for, and Sea Hag is game-warping less often.  With KC, keep in mind that Colonies are in fact more likely to be around on KC boards than non-KC boards. Also, I do think that it's quite often that the presence of KC makes engines viable, not just in a strategic sense (with KC you go for the engine, without you go Big Money-esque), but in an absolute sense too (only source of +Actions is KCing a cantrip, or KC-KC-Smithy etc. without even cantrips).  How often KC makes the difference in these scenarios is obviously a judgment call; I happen to think it's pretty often.

Obviously there are also boards where KC and/or Goons are good buys without being "game-warping"; if you're using Goons as your terminal in a mostly BM deck, or KC is gravy on top of an engine that is decent but can't mega-turn no matter what.  (For example, Minion/Loan with no +Buy.) 
Logged
To discard or not to discard?  That is the question.

timchen

  • Minion
  • *****
  • Offline Offline
  • Posts: 704
  • Shuffle iT Username: allfail
  • Respect: +233
    • View Profile
Re: THE Dominion Card List(s): $5 cards
« Reply #232 on: February 04, 2012, 02:13:04 am »
0

Quote
It's not like I'm trying to, or want to, start a fight here, but honestly the entire post just seems like it's reaaaally stretching. Am I alone here, really? Why are we ranking a card's strength based on how good it is given it's already in the deck (hi Possession)? What does it matter how unique a card's function is when determining how strong it is (hi Saboteur, hi Counting House)? Why is 4. not simply some function of 1.? A card that "can" combo really well, but only 1% of the time, is still a card that combos 1% of the time. What matters is still the amount of time it's actually useful, no? If KC synergizes (combos) with other cards so much better than Goons, why is it a card you want to get on ~75% of the boards you want to get Goons? More importantly, why does it matter?
I have to say I am not stretching my argument at all. It is just that we are having quite different opinions I guess.

As I said, I am NOT rating a card given it is in the deck; the part how often it is in the deck is factored out in factor 1 (and multiplied back in my formula.) How unique may or may not matter depends on personal taste; this is precisely because this ranking does not matter in any practical sense, which is what I am saying in the first place. I would rather say your argument saying Goons is better given possible definition of best is more subjective. And now I gave you a fairly quantitative definition then you start to jump back...

Most importantly, 4 is certainly not a monotonic function of 1. Goons itself is a perfect counter example. The reason why Goons is often necessary in the deck is not because it gets along very well or has lots of synergies with other cards; it's just that its ability itself is pretty strong. The power to lengthen the game and at the same time increase one's own score is just that good. Cursers are good counter example too; the fact they are good has nothing to do with how they combo with other cards.

Again, how a card combo with other cards is not necessarily a factor to be included in the definition of a good card; however, precisely because this ranking has no objective meaning, the definition depends on one's taste. And here I am giving you the definition that reflects my taste on what I think to be a good card, as you required; if you think I am stretching anything it is probably yourself who is forcing some sort of imaginary objective measure on things, as you cannot state clearly what your definition is...

Quote
As far as the last paragraph goes, it doesn't sound like an attempt at a real argument to me. Let's assume all Dominion kingdom cards suddenly costs $2 more, what happens? Well for starters, BM just got a lot stronger. Apart from other obvious stuff, like cheaper cards being hurt more (a 67% increase for Ambassador hurts a lot more than a 29% increase for KC), I don't know what conclusions we're supposed to draw from this thought experiment while still being useful in real life. Why aren't we evaluating the actual cards rather than hypothetical similarish cards, anyway?

You didn't seem to get to the point of the argument. The point certainly has nothing to do with increasing the cost of all cards. All I am saying is that there are indeed some cards, which are more resilient to a cost increase. And they are not necessarily the more costly cards either. When this kind of thing happens, it is a pretty good measure that a card is quite good relative to its cost. And if you buy this measure, I think KC is more resilient to a cost increase and therefore the stronger card. BTW, $2 is chosen due to the small difference of the cards sitting at $2-$4. Again, it is just a measure, so I am not saying it measures anything that is directly applicable in a game. But so is the nature of this list in the first place.
Logged

toaster

  • Young Witch
  • ****
  • Offline Offline
  • Posts: 133
  • Respect: +46
    • View Profile
Re: THE Dominion Card List(s): $5 cards
« Reply #233 on: February 04, 2012, 03:43:05 am »
0

I would suggest Goons is game warping maybe.. 60%?

I think that figure is far too high.  For Goons to be game warping imo, you really need to be able to play at least 3 in a single turn....which I'd guess is a good idea in a decent bit under 50% of Goons boards.
« Last Edit: February 04, 2012, 03:45:23 am by toaster »
Logged

Fabian

  • 2012 Swedish Champion
  • *
  • Offline Offline
  • Posts: 666
  • Respect: +541
    • View Profile
Re: THE Dominion Card List(s): $5 cards
« Reply #234 on: February 04, 2012, 07:54:21 am »
0

RJ (and others),

In the most general sense, a card's strength (expected value) is something like [the amount this helps you win the game] * [the amount of time it's useful] for each kingdom, or more accurately, the sum of its expected values across all possible kingdoms. I agree allfail is setting up a similar sort of calculation, but I don't think the factors he's using are meaningful. I'll get back to that.

Now, the [the amount of time it's useful] part, we can estimate by looking at a kingdom. KC, Expand, Monument, Laboratory, Remake, Hamlet, 4 other cards, no Colony. How often do you estimate you'll want KC in this kingdom? Maybe it's 10%, maybe it's 90%, maybe it's some other number. Let's call it 0.9. Now look at all the other possible kingdoms and do the same, then weigh these estimates together. My best estimate so far for KC seems to be ~0.68, WW's is ~0.55, chwhite's is ~0.85, etc. The best estimate if we look at the sum of all players on isotropic is ~0.84.

The [the amount this helps you win the game] is probably more difficult for us to properly estimate, but in theory we can compare our winrate when using a card to our winrate when not using a card, across all kingdoms. We have councilroom data for this too, but intuitively it seems to me that winrate estimates need a lot more time to converge than buy% estimates, and I would guess numbers from individuals probably are very uncertain here? Someone who's better at statistics than I am could probably calculate that, I'm sure. Whatever, it's mostly a thought experiment at this point.

I reject allfail's notion that these lists are subjective in nature. These things are quantifiable in principle, even if we might not be able to do the calculations properly/accurately. I hope I've given a better answer to what (I think) should go into a card strength evaluation now, at least.

The problem I have with a lot of the arguments brought up is that they're, at best, mere indicators of card strength, and at worst, not even that. KC isn't strong because it has a unique effect. KC might be strong AND have a unique effect, but a unique effect does not a good card make. Correlation does not imply causation; KC just happens to be both strong and unique (ish). Counter examples, cards that are bad and unique, have already been brought up (Saboteur, Counting House). KC is good because KC is good, not because of its uniqueness (seems I'm repeating myself from the old Cellar vs Warehouse thread now, heh).

Anyway as mentioned earlier, I guess one reason we disagree so much is because you look at this as a subjective list where you can more or less make things up because they're your opinion. I would suggest this works a lot better on a "favorite cards" list than a "best card" list. Taken to its extreme, I can claim Herbalist is the best card in Dominion because.. and then make something ridiculous up. It's all subjective anyway, right? I imagine people would respond with some reasonable things, and then suggest to me that maybe I just really like Herbalist?

Like, if you think "a card's ability to combo well with other cards" is a good critera to use for evaluating a card's strength, then that's fine and all that, but maybe you just really like building combo decks?

Anyway it's all good I hope, I can live with my imaginary objectiveness :)

Finally, as for the random numbers I threw out about game warping and stuff, you all very well might be right. I certainly didn't think very hard about it. Of course, in the more general sense, saying a card is game warping is really just a roundabout way of saying something like you probably need to buy it more often, and it'll probably be really good on average in those cases (as per the expected value calculations above); a card being "game warping" has no quantifiable meaning in itself. I think I might be thinking of "game warping" slightly differently, but then again maybe not, and I don't think it's a very important point.
« Last Edit: February 04, 2012, 08:00:26 am by Fabian »
Logged

timchen

  • Minion
  • *****
  • Offline Offline
  • Posts: 704
  • Shuffle iT Username: allfail
  • Respect: +233
    • View Profile
Re: THE Dominion Card List(s): $5 cards
« Reply #235 on: February 04, 2012, 09:07:07 am »
0

Fabian:

Good that you try to make yourself clear. I am going to try to persuade you that your objective standard does not really exist. :P

So your measure contains two parts. The first part I understand: it is just the % of the setups with the card in question that the card should be gained/bought. It's not too different to #1 in my criteria, I hope you agree...

Now the second part. You are proposing something that measures the effectiveness/importance of the card when it is bought. Now this quantity (obvious for me, but I guess depends on how you see it) I don't think can be measured by a single scalar. For example, in our original argument, suppose KC and Goons are both in a setup and they are categorized as a must-buy for this setup. Which one is more important/effective? I don't think this question has an answer, as what one should probably do is to get both.

I think the conceptual problem is more severe. What is the difference between the first and the second criteria? If a card should be in the deck but is not so important, then what does it really mean for us to say that it should be in the deck? I imagine you are thinking about some percentages that a card add to your win rate when you put it in the deck, but in this case you don't need the first number at all-- you can just rank by this number. It can be negative and we don't need to throw away that part.

And there is still a real problem. How do you measure this number? Suppose we want to rank about some card which cost $3 for example, so people can always get it if they want. Now the win rate with and win rate without for some given sample of people actually do not measure the effectiveness of the card; instead it measure the skill of the sample of the people to use the card. Consider one extreme example where everyone uses the card perfectly. Both win rate with and without will be 50% in this case, as everyone just takes the card when it is good and skip it when it's bad. So this number provides no information when the players have perfect knowledge. If we consider people of random skills, such as on council room, the average win rate with and without just tells how this ensemble tends to err on the two sides. A high win rate with does not mean that this card is super important or effective when it is bought or gained; it just means in our ensemble people tend to err on the side not getting the card.

If we think about a $6 or $7 cost card, then the situation becomes a little bit more complicated. Let's think about the perfectly knowledgeable players again. In this case, the win rate with will be higher for an important card, as it is possible for one player not being able to get the card even if he wants to, given his bad luck. However, in this case the number still does not measure the importance of the card; it just measures how hard it is to get the card. Throw in random players then we get a mix of the measure of the skills of the players and the hardness for the card to get.

So the number basically cannot be defined as win rate with/without. One can theoretically define it as a winrate differential against the optimal deck however, I guess. Given a card, for a random kingdom, one can calculate the win rate differential against the optimal deck when the card in question is added/removed from the optimal deck, then average over all kingdoms. Yeah, I guess this number is probably what you are looking for. I don't know how to approximate this number from the data though.

Even such number exists, the problem of this number is that while it tells you something, it is not really useful. And since it is not useful, it is just a way to rank the cards. It is only objective in the sense you define it in the first place. There is no inherent reason why ranking based on this number or something else is more objective than other measure. I think the previous discussion of mountebank vs. witch is a very good example. From the council room data we can in some sense say that the mountebank is in average the better card. But this statement has no practical use at all. And more ironically, in this case when they both appear in a kingdom, it is actually the witch which is the better card to get.

I would rather say this kind of ranking is probably useful only when the cards in question are relatively far away-- in this case, then we can say with fair certainty that one card is likely better than the other in average; and the statement is useful as one can use it as sort of a first step analysis on the board. But surely enough, I would say most of our different ranking methods will give the same result for the cards far enough away. So each of our subjective ranking can be equally good in this aspect.
« Last Edit: February 04, 2012, 09:12:29 am by timchen »
Logged

Fabian

  • 2012 Swedish Champion
  • *
  • Offline Offline
  • Posts: 666
  • Respect: +541
    • View Profile
Re: THE Dominion Card List(s): $5 cards
« Reply #236 on: February 04, 2012, 09:55:10 am »
+2

Oh we definitely shouldn't just be looking at win without and win with on councilroom to figure this out, I hope I didn't make it sound like that either. Obviously there's a whole bunch of problems with the accuracy of that, which WW pointed out a few pages ago.

I'm not saying actually calculating an expected value for a card is easy. It's probably incredibly complex, since, as you point out, it depends on what other cards are available. I even said as much:

"These things are quantifiable in principle, even if we might not be able to do the calculations properly/accurately."

I will say again though that there definitely IS such a thing as an expected value for a card on a given board (and thus, an expected value for a card overall once you weigh all expected values of each kingdom together). I'm not too concerned with being able to measure it personally; I know that's probably close to impossible and probably requires us to solve the game of Dominion before we can attempt it. I feel extremely confident it exists in principle though, and so I reject the notion that how good a card is is determined solely by peoples' opinions (extreme example being me saying Herbalist is the best card because random crap reason doesn't make Herbalist the best card).

"Now the second part. You are proposing something that measures the effectiveness/importance of the card when it is bought. Now this quantity (obvious for me, but I guess depends on how you see it) I don't think can be measured by a single scalar. For example, in our original argument, suppose KC and Goons are both in a setup and they are categorized as a must-buy for this setup. Which one is more important/effective? I don't think this question has an answer, as what one should probably do is to get both. "

I can't quantify that accurately, no. That doesn't mean there isn't an answer, single-scale or otherwise.

"One can theoretically define it as a winrate differential against the optimal deck however, I guess."

I agree! More or less, anyway, it's probably more complicated still. I agree we can't get a very good approximation of this (real) number based on councilroom data.

I strongly disagree that this theoretical number is "not really useful" for making a "best cards" list if we (in theory) had it. A card with the highest EV is the best card, how is it even possible to argue anything else? This goes for any game, not just Dominion (AA is the best hand in Hold'em because its EV is the highest of all hands (which I guess hasn't been rigorously shown, but there's very strong arguments suggesting this is the case, both logical and quantitative, and I doubt many people would disagree. If you like 72 off-suit better, then cool story bro and maybe you just like playing with bad hands?), etc). Of course, just like with Dominion, EV of cards/hands/similar probably haven't been properly quantified in most games, but they still exist in theory just like they do in Dominion.

With the Witch/Mountebank example, you again go to councilroom data to tell me my thinking is wrong. Councilroom data doesn't answer these questions, as we've both addressed several times by now.
« Last Edit: February 04, 2012, 10:00:41 am by Fabian »
Logged

Fabian

  • 2012 Swedish Champion
  • *
  • Offline Offline
  • Posts: 666
  • Respect: +541
    • View Profile
Re: THE Dominion Card List(s): $5 cards
« Reply #237 on: February 04, 2012, 10:10:42 am »
+1

Aanyway I think I've contributed to this derail into weird philosophical debates enough. Let's agree to disagree and patiently await the $6+ list :)
Logged

rrenaud

  • Administrator
  • *****
  • Offline Offline
  • Posts: 987
  • Uncivilized Barbarian of Statistics
  • Respect: +1177
    • View Profile
    • CouncilRoom
Re: THE Dominion Card List(s): $5 cards
« Reply #238 on: February 04, 2012, 11:10:40 am »
+3

I don't think you two really disagree that much.  If you had a Dominion oracle (which tells you the perfect play for any move), then you can get objective numbers.  Fabian and Tim both realize this, but Fabian cares more about it (because of principle), while they both know it's impossible to actually achieve, and so Tim disregards it.

Another way of thinking about it is this.  Imagine you play a bunch of Dominion, and every time you randomly ban a card for your opponent.  He cannot use the banned card, but you can.  Then you compute how much your win rate changed as a function of the banned opponent card.  Assuming perfect play, when you sort by that change in win rate, you have a nice objective measure of the strength of the cards.
Logged

HiveMindEmulator

  • Mountebank
  • *****
  • Offline Offline
  • Posts: 2222
  • Respect: +2096
    • View Profile
Re: THE Dominion Card List(s): $5 cards
« Reply #239 on: February 04, 2012, 05:39:04 pm »
+1

Another way of thinking about it is this.  Imagine you play a bunch of Dominion, and every time you randomly ban a card for your opponent.  He cannot use the banned card, but you can.  Then you compute how much your win rate changed as a function of the banned opponent card.  Assuming perfect play, when you sort by that change in win rate, you have a nice objective measure of the strength of the cards.
This is pretty-much exactly what I think the best criteria for "best" cards is. And I think Goons is better than KC. If you can't buy KC, you can try to out-race it to half the VPs. You might still lose 55-60% of the time or something, but you still have a fighting chance if you get lucky. It's much harder to outrace Goons for 2 reasons: (1) Goons is an attack that slows you down, and (2) Goons has access to an unlimited source of points in VP chips, so you can't just race to half the available point.

On a related note (and actually related to $5 cards) this is why I put Torturer ahead of HP, though I'm not that sure my estimates of numbers is correct. Even though HP is often part of the best strategy (80% of of the time?), a lot of times there will be an inferior HP-free strat that's a 43/57 underdog (or something). But with torturer, even though you may only want it 50-60% of the time, in those situations it's like an 80/20 favorite vs torturer-free strategies.
« Last Edit: February 04, 2012, 05:42:50 pm by HiveMindEmulator »
Logged

timchen

  • Minion
  • *****
  • Offline Offline
  • Posts: 704
  • Shuffle iT Username: allfail
  • Respect: +233
    • View Profile
Re: THE Dominion Card List(s): $5 cards
« Reply #240 on: February 04, 2012, 11:54:53 pm »
+2

I don't think you two really disagree that much.  If you had a Dominion oracle (which tells you the perfect play for any move), then you can get objective numbers.  Fabian and Tim both realize this, but Fabian cares more about it (because of principle), while they both know it's impossible to actually achieve, and so Tim disregards it.

Another way of thinking about it is this.  Imagine you play a bunch of Dominion, and every time you randomly ban a card for your opponent.  He cannot use the banned card, but you can.  Then you compute how much your win rate changed as a function of the banned opponent card.  Assuming perfect play, when you sort by that change in win rate, you have a nice objective measure of the strength of the cards.


As much as I'd like to agree, I cannot.

How hard it is to calculate the index is not the reason why I forgo this index. I would like to emphasize that I think while there are many plausible ways to gauge how good a card is, there is no one definite way which is more fundamental or better than others.

Take the above measure for example. I would imagine from the criteria, a Gold would very often be a better card than a Platinum, and a Silver is very likely to be better than a Gold. It is true in the sense that silver is probably a more necessary card than gold in many decks. However, is this what one has in mind regarding the best card? I would say maybe, maybe not.

But ok, let us accept this index for the sake of the argument. Now, why can we not instead consider an alternative, say not banning the entire stack, but just not allowing a strategy to buy the card in question, the first time it had the chance and would like to do so? (i.e., ban it only once instead of the entire game.) From any kind of higher level argument I can only imagine this index is as good as the original one. The ranking will be drastically different though; Silver is not as dominant any more.

Let me make an analogy to indices in stock market. There are various indices to indicate different things. But is there a single index which can guarantee a stock to go up next day? No. Is there a single index which fundamentally means more than any other? No. All indices carry information, and they correlate or anti-correlate with each other to some extent. Imagine you are going to rank some set of stocks based on their performance. There are many ways to gauge and the ending result would be correlated but not quite the same for the various ways. The question would then be, forgetting about how to setup the indices, is there any fundamental sense that one stock is just better than the other? My answer is no. It really depends on what you are looking for.

Same goes for ranking the cards here. Sure you can make some index (whether it is hard to calculate or not), but in the end, why should the card strength  be single-handedly measured by the index, I just don't see it. The ranking just means to measure what it is defined to measure, no more and no less.

EDIT: let me add a comment to the herbalist example. If you make an index for the best card which the herbalist jumps at top, then yes, I would say the index is probably way off. The problem is, the existence and uniqueness of some partial ordering does not imply a unique ordering for the entire set.

« Last Edit: February 05, 2012, 12:10:09 am by timchen »
Logged

DrHades

  • Conspirator
  • ****
  • Offline Offline
  • Posts: 206
  • Respect: +79
    • View Profile
Re: THE Dominion Card List(s): $5 cards
« Reply #241 on: February 05, 2012, 09:04:39 am »
0

I compare cards by two measures:

1. Percentage of boards, where the card is useful.
2. Winrate of player that buys the card versus a player that doesn't, BUT only on boards where the card is useful.

So basicly - If the card is a good buy on many boards, it is a good card. But if it doesn't make much of a difference against someone who didn't buy it, it is not all that great. Coppersmith is very bad at 1, but good at 2. Laboratory is very good at 1, but not very good at 2.

Beware! This isn't evaluating "traps". If the card is bad on the board, but you buy it anyway, it is your fault, not the cards weakness.

Goons vs. King's Court:

1. Goons is much better. Many boards are so quick that you must pass on KC (mainly because it cost 7$ and you get to 7$ later then to 6$). Goons on the other hand slows the game down, so even if you get them later, it hurts much less than with KC.

2. KC is definitely better. There are more boards where without KC you are dead. But there are also boards, where KC boosts you, but your opponent has a fair chance to take 4-5 Provinces and make your winning much more dificult. With Goons? Enjoy slowly taking Province after Province while once upon a time I do a turn where I buy 3 cards I wanted anyway and take 9 VPs...

If KC was 6$ or Goons were 7$, there will be no doubt. But when it is like this, I am 100% for Goons.

Now I looked at my list...I put GM first -_-;;; But it still wasn't the biggest mistake on my lists - that is definitely IGG at #31  ;D
Logged

dan11295

  • Navigator
  • ****
  • Offline Offline
  • Posts: 75
  • Respect: +5
    • View Profile
Re: THE Dominion Card List(s): $5 cards
« Reply #242 on: February 05, 2012, 12:29:55 pm »
0

King's Court tends to be much more deck dependant in regards to how useful it is as opposed to Goons. I can think of multiple cases where that KC may not be worth it: Games with no +buys is the biggie. In non-colony games the game may end too quick to make good use if it, especially true with cards like JoAT in play. Lastly, its purchase can be questionable in IGG rush games or other situations with no trashing and hands get filled with junk quickly, much more likely to have nothing to KC when you draw it in though situations.

Goona on the other hand can rarely been ignored completely, although with no +actions it is not as game changing due to not being able to rack up the VP with it as much. Meaning a Goons engine is not really an alternative to VP cards in those cases.

That said KC has a bigger game-warping ability that Goons in kingdoms where it is a factor. many cards if KC'd fundamentally change the game, including Bridge, many attacks, Possession, Grand Market, card drawers, etc.

For pure Goons engine to be effective, you need +Actions, trashing, and card drawing. ou need all these if you want to expect to play 3+ Goons in a turn.   

 
Logged

Qvist

  • Mountebank
  • *****
  • Offline Offline
  • Posts: 2400
  • Shuffle iT Username: Qvist
  • Respect: +4072
    • View Profile
Re: THE Dominion Card List(s): $5 cards
« Reply #243 on: February 06, 2012, 07:50:47 am »
0

One weekend not at home, and already a lot of discussions going on, wow. And not about this list, it's about the upcoming list. It's really interesting to read your comments. I haven't read all of them in detail and won't comment them on my own now. I just want to say that all of your comments seem reasonable and it's really hard to rank King's Court, Goons and even Grand Market as they are all very dominating cards on nearly each board. I won't keep you in suspense so long. I will post the $6+ list in a few minutes.
Pages: 1 ... 8 9 [10]  All
 

Page created in 0.097 seconds with 21 queries.