"Ultrarare" part II

[Ringel]

The problem is your data is incomplete. You haven't run any statistical analysis on your data and you are selective in which rares you consider. Interesting that all the under 30 count rares are Black Assassin rares which backs up the increasingly available data.

Notice that if I am correct, the numbers show that each Assassin black rare will show up approximately about 2/3s as often as each other Assassin rare. But that would be on average.

[Ringel]

I'm not claiming anything is actually ultrarare, but that the rare cards aren't evenly distributed-- black assassin rares and blue templar rares appear to be rarer than other rares.

I didn't think this was true before examining the actual data.

[thedude808]

Mar 15, 2012 12:43:32 GMT -5 Ringel said:

I'm not claiming anything is actually ultrarare, but that the rare cards aren't evenly distributed-- black assassin rares and blue templar rares appear to be rarer than other rares.

I didn't think this was true before examining the actual data.

This.

I noticed this with my cards a couple of weeks ago, but I assumed it was because of a relatively small sample size or a statistical anomaly. After I saw other people talking about the same thing with their cards, though, I thought their might be something to it.

I still don't know if this is true, but it sure looks more likely than it did when it was just me looking at my own cards.

Also, I'm not saying this is necessarily a bad thing. It's an observation.

[redpipe67]

Templar
Gold 46 from 7 different rares
Black 22 from 3
Red 50 from 7
Purple 43 from 7
Blue 47 from 11
Green 52 from 7

Total 260

Assassin
Gold 39 from 7
Black 27 from 11
Red 36 from 7
Purple 31 from 7
Blue 15 from 3
Green 33 from 7

Total 181

[jeremyat]

For my part, I have the following:

Templar Rares:

Gold: 17
Black: 7
Red: 21
Purple: 19
Blue: 24
Green: 19

Assassin Rares:

Gold: 10
Black: 7
Red: 11
Purple: 8
Blue: 2 (really?)
Green: 8

What's really annoying is the rarity of some of the uncommons. For example, I have two Borgia Towers (out of 107 templar packs) but six copies of Uberto Alberti. I will trade two-for-one Ubertos for Towers. I've got thirteen Zealous Bankers though!

[Ringel]

I think the distribution of uncommons might also not be even (blue templar and black assassin being more uncommon than the rest), but I don't want to gather the stats for that as well.

[eev]

dont know about all rares but more annyoing are uncommons, especially because after buiyng i think more than hundred of templar boosters i still dont have a single copy of blue uncommon counter and this sucks.

[jefmajor]

I don't know if you wanted these results filtered by the rares possible or whatnot, but, this is just a straight count of my rares in each category/

I have one of every rare except: 3 Assassin rares. All in black.

Templar-

Gold: 46
Black: 27 (Makes sense to be low, only 3 possible rares)
Red: 43
Purple: 61
Blue: 40 (The distribution is terrible, 3 of those rares I have only 1 copy of)
Green: 43

Total: 260

Assassin -

Gold: 23
Black: 16 (Missing 3 and have bad distribution as well)
Red: 27
Purple: 26
Blue: 14 (Makes sense to be low, only 3 possible rares)
Green: 20

Total: 126

[jefmajor]

I know this math is unrelated to the current study AND it is a false result because it is based on only my own data and other data refutes it, making it at best irrational, but, I'm going to post it anyway because it makes me angry.
For me:

Cesare = 0.38% chance per templar pack.

Grrr.

[lefthighkick]

I had 5xCesare before I had 5 of any other rare. This is pointless you guys.

[Tuism]

It is human nature to find patterns in random and claim that it has "figured it out". And everyone wants to think that they were screwed by someone as opposed to "luck". It's better to blame someone than no one

I remember in the beginning half of us thought there were some kind of matchmaking that matched players against the same decks. But it turned out to be pure paranoia and pure chance.

[jefmajor]

I didn't mean to derail the thread: The OP had nothing to do with Cesare. My post was prefaced by saying exactly what you just said LeftHigh. Disregard it. It was an emotional statement in a thread about statistics and has no place.

Apparently my disclaimer wasn't big enough.

[Ringel]

That's why I'm using statistics rather than relying on human error. Both jefmajor and lefthighkick's observations about Cesare are examples of human error.

The stats are definitely showing that the distribution is not random, however. This surprised me when I discovered it, because I am well aware of our human ability to make mistakes about randomness.

The stats are much more solid evidence than anything to the contrary-- much more so than any human observation of pattern.

[eev]

in real games like mtg you can buy a display and there will be some kind of a system. so you can be sure you won't get dozens of copies of a crap rare but of course it works opposite too. in case of online tcg like this one the pure randomness of boosters and no displays to buy makes getting all cards more or less equally complicated and random too. so actually it might be a good idea to make some system for displays and sell them in store too. but of course the auction stuff gonna make it work too.

[Hurdler]

I can guaranteed you that it's truly random, as far as a computer program can be truly random (that's a complexe subject). We use the Python random algorithm.