Hearthpwn

My best guess is that the newly shuffled raptors, in addition to having +2/+1, also adopt all of the +2/+1s from the Raptor that created it. So you play your 1/3 raptor, it shuffles a 3/4 raptor, then THAT shuffles a 5/5 raptor, etc.

Probably my favorite Core Set card revealed up to this point.

Never underestimate the impact of "Start of Game: Draw this."

Even at 5 mana in pool 2, this thing can single-handedly win games. I can't imagine what it must have been like at 4 mana in pool 1.

I haven't been able to find much information about what factors go into what cards you're offered. One of the in-game loading screen tips says outright it depends on your deck, and Blizzard has confirmed that this is the whole idea behind the "Group Learning" bucket; however, I suspect there are similar under-the-hood factors that go into the other buckets as well.

I suspect (but can't prove) the following:

• Picking a (non-GL) bucket increases the odds of finding the same-name bucket in the future.
• Some buckets are coded to always offer three copies of the same card ("Bruisers", "Why? Why not?"), even though those buckets definitely contain more than one card.
• There is a bias in favor of offering three of the same card for certain cards, even in otherwise normal buckets.
• There is a small chance that the game will deliberately offer you a bucket containing the exact same three cards as the bucket you picked last.

But my personal suspicions aside, I want to learn more about what factors influence the buckets you're offered – not only to increase my winrate by picking more strategically, but also to enable me to feel more in control of my Duels runs. I wasn't able to find a website that offered any definitive stats, and I'm not sure how to investigate it myself; if anyone else has any ideas, I'm all ears.

Quote from D_Lord >>

I have played a lot of Duels recently and my tier list for classes goes like this:

Tier 0 - Druid (Tokens) - I went 12-1 twice and those runs were absolutely disgusting (spells cast twice, zombies spawning everywhere, spells cost less, 4+ Arbor ups, even the Rager treasure). I had only one 3-3 run, but 5 more runs with 7, 9, or 10 wins. This is definitely the best deck by miles. 7/8 runs 7 wins or better. Or in other words: 78% winrate.

Tier 1 - Warlock (Discard and also Soul) - I went 7-9 wins with Discard a couple times but the buckets are sometimes just not what you want. Can also go terribly wrong against Aggro Weapon Rogue or Bomb Warrior or really any good burn. Killmox nerf hurt this quite a bit but it was necessary. I would also never underestimate the Demon Treasure you can pick at the start that gives you several demons and reduces their cost. That thing becomes a 3 cost draw 5+ in later stages that easily wins games on its own. The soul hero power can screw over token druid with its tempo and warlock seems to be the best answer to token druid in general.

Tier 2 - Mage (Elementals, Control), Hunter (Deathrattle), Shaman (Overload, Burn), Priest (Control, "Suck Health") - These 4 can highroll like crazy, have good matchups, be quite flexible and can get you somewhere. Some of the treasures make or break these decks, though, so they become rather unreliable unless you get the absolute nuts buckets. Regarding highrolls, I met a deathrattle hunter yesterday who went turn 1 Egg into Deathstrider to summon 5 4/4s. So yeah, that happens when treasures are slightly unbalanced. You need an insane deck to get through later stages. 

Tier 3 - Warrior (Bomb, Enrage), Rogue (Weapon), Paladin (Big/Legendaries, Divine Shield/Stats exploits), Demon Hunter (Combo/Aggro). While bomb warrior can get some wins early, it feels a little too slow in the later stages because decks get too big. Rogue is very strong but it's worst matchup by far is the best deck in the meta (token druid), which even wins 99% of the time in round 1 against it (despite rogue having so much burst available). The deck just can't deal with wide boards. Paladin is pretty bad overall, but I have seen some interesting exploits of the base hero power and the weapon, so I might try it again. Demon hunter feels pretty weak right now overall.

 As someone who has likewise played a fair amount of Duels, I mostly agree with this; the only thing I'd change would be moving Frost Shards mage to tier 1. Admittedly, I haven't played it myself (don't want to spend 1600 dust to get it), but one cost pings do quite a number on aggressive strategies while control decks are blown out of the water by Grand Finale

Quote from Avaidaocu >>

 Thats more or less what I had in mind with my suggestion, but bear in mind yours is far stronger. In pure paladin, sometimes Alura will pull Consagretion, First Day at School, Libram of Justice etc, spells that are not great to play at any time. Making her pull only buffs could make her highroll potential more consistent, since "targets this" also includes Libram of Hope.

 If the power level of this design is too high, it could always be accompanied by a mana cost increase. I doubt this design would be unfair at 5 mana, but it could be set to 6 if Blizzard wants to play it safe. (I will admit I didn't think of Libram of Hope, argh!)

Quote from Screwjank >>

Also, a general question. Why do people say Patches the Pirate is a strong card? 

Patches the Pirate is a strong card because he comes with essentially no cost – on its face, there's no reason not to run him (in a deck with at least a handful of Pirates). He removes himself from your deck so you don't need to worry about drawing him when you wanted to draw something higher-impact, he doesn't cost mana to put into play, nor a card from your hand... the only cost of any kind to putting him in your deck is the off chance you draw him in your opening hand (or otherwise before you have a chance to play your first Pirate) and feel really sad.

Regarding High Abbess Alura: I seem to recall Blizzard saying at one point that the intent of Alura was for cheating out buff spells like Blessing of Kings and Blessing of Authority. That being the case, if Alura does indeed need a nerf, the nerf I would favor would be:

"Spellburst: Cast a random spell from your deck that can target this. (It does.)"

That quite cleanly preserves the design intent, eliminates the possibility of cheating out massive board summoning cards, and keeps the card text simple and straightforward.