Print

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

[traces Around]

You nitwit. There are 299 cards which you rank 1 through 299, so the average card has a ranking of 150.

No no, that's not what I mean.

So, I was doing my card rankings, right, and I am down to needing to rank 33 cards which fall somewhat in the middle - spots 128 through 165. So I'm looking over this list of 33 cards - included are cards such as Warehouse, Village, Catacombs, and Treasure Trove - and I am going, "Gee, these are all pretty good cards." Meanwhile, the cards which I might more often consider average ones - Market, Squire, Candlestick Maker - fall closer to 200 than 150.

From this observation, it seems to be the case that there are a lot more good cards than bad ones, which I suppose is a testament to how well designed the game is. On the other hand, perhaps those bad ones are so bad (and gone in 12 cases) that the average is dragged down by them, or perhaps I need to revisit what I consider an average card.

So here is the question that this brings up and that I am curious as to different perspectives on: what is a Dominion card that fits the term "average", and where would you place it or have you placed it relative to other cards?

[Beyond Awesome]

I don't believe the Bell Curve is the best way to look at things in Dominion.

As for an average card, I would consider something like conspirator the definition of average.

[navical]

Partly this is because "average", as well as meaning "better than half but worse than half", has also come to mean "substandard". (See, for example, the thought process that leads politicians to say things like "every school in the country should be above average"). Even if you are thinking of the first meaning, the second will be subconsciously affecting you.

[luser]

Explanation here is that you encountered ranking bias. Humans couldn't well rate equally on numeric scale, they give mostly higher scores. Good example are movies, if movie gets 7/10 rating  its pretty average one.

[Qvist]

Explanation here is that you encountered ranking bias. Humans couldn't well rate equally on numeric scale, they give mostly higher scores. Good example are movies, if movie gets 7/10 rating you its pretty average one.

Yeah. I totally agree with that as I wrote exactly the same a couple of months ago. That's why I prefer a ranking than a rating and that's why I transfer the rating list into a ranking list.

[werothegreat]

Explanation here is that you encountered ranking bias. Humans couldn't well rate equally on numeric scale, they give mostly higher scores. Good example are movies, if movie gets 7/10 rating  its pretty average one.

I mean, grades are the same way.  I guess it's because we expect, on average, to do better than 50%.  Only doing half of something correctly is considered a failure.

[sorawotobu]

I think OP may have conflated "average" with "mediocre". If an averagely powerful card is pretty good pretty often, that's actually just a sign of good card design. It's evidence that the best strategy will often involve a bunch of cards on the board and not just two or three.

On the point of movie ratings etc.: People learn how to rate something based on the rating more experienced people give. So people are bad at grading because other people are also bad at grading, I don't think humans have a disposition to give better ratings than they should. Also your perception might be influenced by the fact that you look up reviews of above average films more often than you look up reviews of below average films since you're less likely to have heard of the latter.

[gloures]

I feel this might have a lot to do with how we're graded in school/uni, in general 60 is passing grade in most places, so people tend to associate that to something that's borderline acceptable, and rate accordingly to that notion

[LibraryAdventurer]

Maybe we should have a bracket and vote for what is the most average card.





or not

[AJD]

Explanation here is that you encountered ranking bias. Humans couldn't well rate equally on numeric scale, they give mostly higher scores. Good example are movies, if movie gets 7/10 rating  its pretty average one.

I mean, grades are the same way.  I guess it's because we expect, on average, to do better than 50%.  Only doing half of something correctly is considered a failure.

I think the intuition here comes from something like: on a test consisting of true-or-false questions, someone who's learned absolutely nothing should get 50% right by guessing. So in order to "pass" you have to do significantly better than 50%. Obviously in most situations other than true-or-false questions we wouldn't usually expect someone to hit 50% correct just by chance, but I kind of suspect that that's where the idea comes from.

[traces Around]

Maybe we should have a bracket and vote for what is the most average card.

This thread happens to have originated out of a similar joke.

I am quite aware of these biases, and after thinking more about it, I do not believe that they are what is resulting in the discrepancy.
Rather, the ideas of cards being "good", "average", "bad", or above and below those evaluations are doing something different than the rankings: the categories are something which the effect of the card is, whereas the ranking is going relative to other cards.

When I say a card is "average", I am looking at a ranking of effects - good effects have high amounts of redundancy, bad effects not so much, so although terminal draw takes up some relatively large amount of spots in the card rankings, it only takes up one in the effects evaluations, and so those effects, and with them cards, that are average show up below the middle. Perhaps in, say, April or May, an effects ranking could be attempted.

[thespaceinvader]

I don't believe the Bell Curve is the best way to look at things in Dominion.

As for an average card, I would consider something like conspirator the definition of average.

You... know that it's called a bell curve because it's shaped like a bell, right?  It's not named after anyone...

For my money, these are difficult cards to rate because their utility really depends on the kingdom.  Market Square, for instance, is mostly not that amazing, except when it has a bunch of synergy (Hermit, maybe Bishop, Apprentice, other TFB generally) at which point it's absolutely dominant.  How do you rate that?

[Jack Rudd]

I once worked out what the most average card was by using the metric of "which card is gained on a proportion of boards closest to 50%".

To my great amusement, said card turned out to be Pearl Diver.

[Watno]

I guess the average card is alot better if you remove the cards not in the second edition.

[drsteelhammer]

How come that nobody has pointed out the difference between the median card and the average card? The 150th card is the median card and it's highly unlikely that the average card is equal to the average card. If you think the average card is below, it just means that the negative outliers are more extreme than the positive ones. Or your feeling about average cards is flawed, which is usually the right explanation when feelings get into conflict with mathematical concepts.

I don't believe the Bell Curve is the best way to look at things in Dominion.

As for an average card, I would consider something like conspirator the definition of average.

You... know that it's called a bell curve because it's shaped like a bell, right?  It's not named after anyone...

For my money, these are difficult cards to rate because their utility really depends on the kingdom.  Market Square, for instance, is mostly not that amazing, except when it has a bunch of synergy (Hermit, maybe Bishop, Apprentice, other TFB generally) at which point it's absolutely dominant.  How do you rate that?

You rate the card individually for all the 10^600 kingdoms it appears in (or whatever nubmer it is), add the scores and divide them by 10^600. That's how we all do it, right?

[Awaclus]

highly unlikely that the average card is equal to the average card.

Uh, that sounds pretty damn likely to me actually.

[Chris is me]

highly unlikely that the average card is equal to the average card.

Uh, that sounds pretty damn likely to me actually.

I think he is trying to highlight how both median and mean can be referred to as "average".

[Awaclus]

highly unlikely that the average card is equal to the average card.

Uh, that sounds pretty damn likely to me actually.

I think he is trying to highlight how both median and mean can be referred to as "average".

I was just making a joke, why do you have to be so mean?