Which Standard decks are polarizing? Answered with data

Multiplayer Discussion
1

First, how this was measured:

  1. Choose an archetype.
  2. Get matchup winrate data for matches against all other archetypes (plus vs itself, tautologically 50%).
  3. Take away 50% then take the absolute value. E.g. |46%-50%|=4%.
  4. Do a weighted average by the opposing deck’s general popularity.

Data was collected from HSReplay using the Past 7 days filter starting 3 days ago, until today. I was examining every archetype within reason (and some arguably without) so it took a while.

Second, link to Google Sheet:

 https://docs.google.com/spreadsheets/d/1oELtPBnbno-CwSC8CvaJf2BQsfOEvJ2uDLs3mQTt7Tc/edit?usp=drivesdk

Third, results!

Ten most polarizing, Top 1000 Legend

  1. N’Zoth Shaman
  2. Anacondra Druid
  3. Ping Mage
  4. Big Warrior
  5. Quest Paladin
  6. Quest Fatigue Warlock
  7. Quest Mage
  8. Poison Rogue
  9. Quest Priest
  10. Corrupt Priest

Edit: 10 most polarizing, Top 1k Legend, >45% winrate only

  1. Anacondra Druid
  2. Big Warrior
  3. Quest Fatigue Warlock
  4. Poison Rogue
  5. Fel Demon Hunter
  6. Quest Lifesteal DH
  7. Mozaki Mage
  8. Aggro Taunt Druid
  9. Quest Handlock
  10. Quest DH (non-Lifesteal)

Ten most polarizing, Legend but not top 1000

  1. N’Zoth Shaman
  2. Poison Rogue
  3. Anacondra Druid
  4. Quest Paladin
  5. Ping Mage
  6. Quest Priest
  7. Quest Mage
  8. Big Warrior
  9. Quest Lifesteal Demon Hunter
  10. Corrupt Priest

Edit: 10 most polarizing, Bottom Legend, >45% winrate only

  1. Poison Rogue
  2. Anacondra Druid
  3. Big Warrior
  4. Quest Lifesteal DH
  5. Mozaki Mage
  6. Fel Demon Hunter
  7. Garrote Rogue
  8. Quest Fatigue Warlock
  9. Quest Zoo Warlock
  10. Quest Handlock

Ten most polarizing, Diamond 4-1

  1. Quest Demon Hunter (non-Lifesteal)
  2. Ping Mage
  3. Quest Lifesteal Demon Hunter
  4. Anacondra Druid
  5. Quest Priest
  6. Mozaki Mage
  7. Fel Demon Hunter
  8. N’Zoth Shaman
  9. Poison Rogue
  10. Quest Mage

Edit: 10 most polarizing, D4-1, >45% winrate only

  1. Anacondra Druid
  2. Mozaki Mage
  3. Poison Rogue
  4. Big Warrior
  5. Big Priest
  6. Quest Handlock
  7. Quest Shaman
  8. Rally Priest
  9. Quest Hunter
  10. Quest Fatigue Warlock

Ten most polarizing, Diamond 10-5

  1. Quest Druid
  2. Quest Demon Hunter (non-Lifesteal)
  3. Quest Priest
  4. Quest Lifesteal Demon Hunter
  5. Poison Rogue
  6. Garrote Rogue
  7. Anacondra Druid
  8. Quest Paladin
  9. Quest Mage
  10. N’Zoth Shaman

Edit: 10 most polarizing, D10-5, >45% winrate only

  1. Poison Rogue
  2. Anacondra Druid
  3. Mozaki Mage
  4. Big Warrior
  5. Quest Fatigue Warlock
  6. Big Priest
  7. Quest Shaman
  8. Quest Hunter
  9. Quest Zoo Warlock
  10. Aggro Shadow Priest

Ten least polarizing, Top 1000 Legend

  1. Doomhammer Elemental Shaman
  2. Pirate Warrior
  3. Zoo Warlock (non-Quest)
  4. Face Hunter
  5. Quest Hunter
  6. Secret Libram Paladin
  7. Quest Rogue
  8. Secret Paladin
  9. Libram Paladin
  10. Handbuff Paladin

Edit: 10 least polarizing, Top 1k Legend, >45% winrate only

  1. Doomhammer Elemental Shaman
  2. Pirate Warrior
  3. Face Hunter
  4. Quest Hunter
  5. Quest Rogue
  6. Libram Paladin
  7. Rally Priest
  8. Quest Shaman
  9. Garrote Rogue
  10. Buff Paladin

Ten least polarizing, Legend but not top 1000

  1. Doomhammer Elemental Shaman
  2. Pirate Warrior
  3. Zoo Warlock (non-Quest)
  4. Quest Rogue
  5. Secret Libram Paladin
  6. Face Hunter
  7. Secret Paladin (<45% winrate)
  8. Handbuff Paladin
  9. Libram Paladin
  10. Aggro Taunt Druid
  11. Clown Druid
    Edit: All in the above list are >45% winrate unless noted

Ten least polarizing, Diamond 4-1

  1. Pirate Warrior
  2. Zoo Warlock (non-Quest)
  3. Face Hunter
  4. Secret Libram Paladin
  5. Libram Paladin
  6. Quest Rogue
  7. Deathrattle Demon Hunter
  8. Handbuff Paladin
  9. Aggro Taunt Druid
  10. Elemental Shaman (without Doomhammer)
    Edit: All of the above have >45% winrate

Ten least polarizing, Diamond 10-5

  1. Face Hunter
  2. Libram Paladin
  3. Quest Rogue
  4. Doomhammer Elemental Shaman
  5. Elemental Shaman (without Doomhammer)
  6. Evolve Shaman (<45% winrate)
  7. Buff Paladin
  8. Zoo Warlock (non-Quest)
  9. Secret Libram Paladin
  10. Handbuff Paladin
  11. Aggro Taunt Druid
    Edit: Decks in the list above are >45% winrate unless noted.
5 Likes
2

i applaude the effort…but is this an average of all matchups?

because i don´t think that´s an overly productive stat.
I think polarization is most problematic if specific matchups strife far of the 55-45, not so much if a deck is overall often 50/50 but also a lot 60-40 ( i mean that can then be a balancing issue, but for the matches itself it should feel “fair” more often than not, opposite to a matchup were you´re opponent overall has a 80-20 to win it, despite the deck averaging out on a low polarity in most other matchups))

1 Like
3

No, that’s too simple a description. Let’s say Archetype P has a 64% winrate against Archetype R and a 43% winrate against Archetype S, and that S is twice as common as R which is in turn twice as common as P. P’s average winrate would be exactly 50%, but we’re not measuring average winrate, we’re measuring average deviation from 50% winrate. So in this example 1/7 of the games would be PvP (0 deviation from 50%), 2/7 would be PvR (14% deviation each) and 4/7 would be PvS (7% deviation each). (0+2×14+4×7)/7=8, so P would have a polarization of 8%.

Edit: The highest average polarizations approached, but never quite reached, 20% — that means 67-33 or 2:1 type matchups were typical for those decks. The lowest polarizations were typically around or even slightly below 7% in Diamond or 8% in Legend, so 57-43 or roughly 4:3 would be a typical experience for those.

4

What is Quest Paladin crushing so much to pair with its bad matchups?

5

Generally, Quest Paladin is strong against decks that go straight for combo and minimize defense. Specifically, Anacondra Druid — 69.9% winrate in Legend.

6

So if a deck saw a 0% playrate like Control warrior vs Pirate warrior in the Jade druid era, jade druid may or may not be seen as a polarizing deck because it’s so unpolarizing the polarized matchup might not even see play or not?

Lets just skip the head gymnastics and skip to the part everyone comes here for. Mage is great, mage is fine. We should really be trying to ban whatever convoluted thing this convoluted mess is, ban nzoth shaman from 1k legend and ping mage and the game will be fixed! Quest lock and Quest mage are only #6 and #7! Extremely useful conclusions as always.

We have to use these incredible random conclusions to immediately ban nzoth shaman from 1k legend next to ping mage and big warrior. That’ll sure show em. A archetype being so good, it has 90% winrate matchups it doesn’t even know about on a deck that could have previously countered aggro decks in the past like Control warrior (having a positive winrate against pirate warrior and midrange shaman in the 80% pirate warrior and shaman, 10-20% jade meta) not seeing play since it’s winrate vs 80% of the meta was like 55%, but the other 10-20% against jade Druid was like 10% is always a thing.

As always theres no magic bullet to every problem though and it’s at least labeled. Anaconda and fatigue lock and a post nerfed quest mage falling lower sound right. But certainly thats the reason balance patches were made, to stop the auto 70-90% lose rate matchups and put a bandage on it like Aggro vs quest shaman (60-70%) and Handlock vs Quest shaman / Slower deck (65-75% wr for questlock/mage at time).

Turns out decks that just win after a certain turn just always tend to win if they were a turn faster. I think it went something like Quest shaman was one of the second slowest quests (only 1-3 turns faster than quest priest. Which the quest shaman could occasionally get the 5 mana 7/7 or occasionally sharded. But usually won at 1/2 on the 7/8 mana card stage.

So quest shaman always beat quest priest by a turn.

Quest mage always beat shaman and Warlock by a turn, and won 70% of the time. But the other classes had better matchups vs the rest of meta.

And quest shaman won against aggro 70% of the time and was immediately yeeted for the sin until Aggro taunt druid got the same and was fine since nobody noticed it past 30 repeating looping greyboughs and always being a hidden sleeping top performer in many meta nobody ever noticed that took a couple weeks to get picked up.

Still i just find it kinda ironically funny a deck that used to be up there, like Poison rogue never made it to the top 5. That deck was like almost EXCLUSIVELY like a deck made to ONLY have a 70-90% winrate against mage and 10%-20% wr against everything else during the 10-22% mage + like half the meta being legend questlocks (handlock + runed rod fatiguelocks + mixes) at the time.

I mean it does spot a few things, it’s not completely bad, it just feels completely weird that poison rogue might not even show up on there (but probably lacks match data, even at height), but had one of the most ludicris spreads of nonstop 10%-90% matchups i’ve seen lmao and was pretty much a one trick pony just made to win vs quest mage and lose vs anything else that played a single taunt or minion lmao.

7

Omg this is stats for people who have no grasp of stats.
A bunch of decks that are so bad they don’t register on tier anywhere…

1 Like
8

No. The most important conclusion to draw is that polarization is not a valid metric to nerf anything. There are some decks that are polarizing that would be absurd to nerf, and some decks that are among the least polarizing that are arguably worthy of nerf. Polarization, if you’re not biased in your analysis of it, is meaningless to the balance conversation.

A (strictly!) secondary conclusion would be that Mage decks that are actually competitively viable get a lot of criticism for being polarizing when many other decks, even other Mage decks, are more polarizing and don’t receive anywhere near the same level of complaints because they’re either not Mage or not viable, but, again, strictly secondary.

9

Yes.
Why?

Get big or low winrate against your special snowflake doesn’t count.
Why?

Only you play It.

10

First off, I have Poison Rogue as polarizing. All sub-metas. So it made the list. Lists, really. Idk what you mean by “never made it there.”

Second, you’re exaggerating. Poison Rogue has almost exactly a 50% even winrate against Quest Shaman, for instance. On average, its matchups in Legend are roughly 65%-35% for one side or the other. No deck at all is worse than 70-30 on average. I’ve noticed it’s very common for players who complain about polarization to greatly exaggerate the degree to which it exists.

1 Like
11

I feel like that’s the Druid’s fault more than the Paladin’s. They’re notorious for folding to decks that play Hearthstone.

Guess it’s hard to assign “blame” for this sort of thing, though.

12

Eh. Q Pally is also 62.4% against Quest Mage and 65.2% against Big Priest. And (like Rukanth’s alternate reality version of Poison Rogue) is seemingly allergic to anything vaguely resembling 50-50 matchups.

It’s still deep into Tier 4 though.

13

Oh yeah. I was half expecting the explanation to be “it’s polarized by losing too much”.

I have noticed Quest Mage struggled with dudes (regardless of questline) but I’m baffled Big Priest can’t keep it down. I guess that deck just can’t beat Jaraxxus.

14

“Which Standard decks are polarizing? Answered with flawed Science and fuzzy math.”

2 Likes
15

Care to lay out what’s so flawed about this post that you’d make a reply like this? I don’t exactly expect perfection or bulletproof logic on forum topics, and I have to at least respect people putting in the work to try providing some consumable statistics on a rather relevant topic.

1 Like
16

The entire thing reads like a wishlist of what the Op wants nerfed.

17

I think the biggest thing this spots is relatively big picture: the polarized list is mostly combos, solitaire, quests, OTKs, etc. A whose who of decks people hate. The least polarized decks are all trying to win by landing minion damage and playing recognizable Hearthstone.

18

I can see no immediate flaw on his method, tho

1 Like
19

I see nothing in the method to indicate that he has proven anything.
The word “polarizing” is as subjective a term as has ever existed.
Therefore his math is meaningless.
You cannot quantify feelings.

20

Hahaha are you serious? First of all, what does the results being what OP wants have to do with their veracity?

Secondly, but just as importantly, are you saying with a straight face you think they want the likes of N’Zoth Shaman, Ping Mage, or Quest Paladin? Many of the polarized decks are straight garbage, and it’s actually unsurprising, because very low tier decks are known for tending towards less balanced matchups. If anything their presence on the list lends it credence as it’s showing an expected pattern.

I think you’re just mad ping mage showed up. Trust me, no one thinks this means the deck is a problem.