vS Data Reaper Investigates 4 Mana 7/7 – How Good is it?

An analysis of Flamewreathed Faceless and its effect on winrates in Shaman decks.

flamewreathed faceless cardFlamewreathed Faceless is one of the most hotly debated cards in the game, not only widely hated by many players in the community, but has evolved into its own meme as a symbol of a ‘blatantly overpowered card’. It is a staple card in Aggro Shaman and finds its way into other archetypes of the class as well. Dropping a 7/7 body on turn 4, with just 2 overload, seems very undercosted, as you’re paying 6 mana for a better Boulderfist Ogre, while also being able to drop it earlier. It can be a very frustrating card to play against, giving one of the most aggressive decks in the game a massive threat to complement its already powerful early curve of minions.

So how good is the card and what is its effect on the win rate of Aggro Shaman against popular archetypes in the game? We will try to provide some insight to this question. The analysis provided is based on data from the past seven weeks, compiling of 40,000 games.

First, let’s look at the drop rates of Flamewreathed Faceless on turn 4 (or turn 3 with the coin) against the common archetypes in the field. Note that a card with two copies in the deck that isn’t being searched for in the Mulligan phase will be in the player’s hand on turn 4 about 40% of the time. So how often does it get played on the turn where it is arguably most powerful?

We can clearly see a pattern here. Flamewreathed Faceless is most often played on turn 4 against slow decks, that don’t possess a lot of early game minions that can contest it. This makes sense, as dropping a large body against an empty board, or one that cannot challenge it effectively, is often an advantageous situation. Against faster decks though, the card gets played much less often. Its drop rate is less than half the times in which it is theoretically in a player’s hand on turn 4. Of course, there are situations where you cannot play the card because you are overloaded on the turn, but that still means you have a card in your hand that you cannot utilize at its most effective turn (we’ll discuss that later). In other cases, the numbers imply that the player decided that dropping Flamewreathed Faceless is not the most optimal play for other reasons.

Now that we’ve seen the drop rates for Flamewreathed Faceless, what is its effect on win rates against the most common archetypes in the Metagame, once it gets played on turn 4 compared to when it doesn’t? Before we look into it though, we’d like to offer some notes of caution about selection bias.

Selection Bias

It is important to understand exactly what can (and cannot) be learned from win rates based on a card being played. Generally, we have to be careful drawing conclusions based on a card being dropped. The reason is “selection bias”. Since a card does not get dropped “randomly”, but gets picked to be played, one cannot look at win rates and immediately draw conclusions on the influence of the card. Several factors need to be considered before a firmer conclusion can be drawn:
First, the fact that a card is dropped, suggests that the player concluded that this was their best play available. So naturally, a card being dropped optimally is expected to positively affect the win rate. Thus, when examining win rates, one needs to take this into account. The more options a player has for the particular play (e.g. later in the game) the more potential bias this issue can introduce.

  1. Sometimes, the likelihood of a card being dropped is related to the state of the game. There are cards that are more likely to be dropped when a player is ahead and likely to win (e.g. Savage Roar). Thus, one cannot conclude from observing the win rates of Savage Roar that it is a great card. There are certainly scenarios where the card is kept in the hand because there are no good uses for it and the player is in a losing position.
  2. There are other times when a card is played when the player is behind, trying to salvage the game. If the player ends up losing, it doesn’t necessarily mean that the card that was played is bad, since it serves as a comeback mechanic and will see more play when the player is likely to lose (e.g. Yogg-Saron).
  3. To minimize the influence of these selection issues, the best scenario that allows us to analyze a card performance, conditional on it being dropped, is to know the board state at that time. Because we do not know the board states based on our data, the next best thing is to analyze situations when the board state is fairly neutral. This means that analyzing early drops is safer from that perspective. Also, it is important to be aware of these “selection issues” when interpreting the data.

Now, let’s look at the results:

Flamewreathed Faceless appears to be most effective against several archetypes of Warrior as well as Druid and Reno Warlock. Even so, most of these numbers are not too dramatic (against Yogg Druid and Dragon Warrior, the difference is borderline statistically significant, though it might be with a larger sample). Against most of the Metagame, it appears to be fairly neutral, not displaying any particular advantage, or displaying one that is marginal at best.

Against Zoo, Miracle Rogue and Paladin, it performs poorly in particular. Miracle has a strong answer to the card (Sap), and preventing a turn 5 Doomhammer line of play might be another reason for the card’s detrimental effect on the matchup. Paladin also has effective tools to deal with a large threat, and Zoo’s effective trading, board flooding, as well as the ability to generate multiple taunts, seems to shine against it.

In addition, we can confidently state that based on our observation and further analysis, Flamewreathed Faceless’ performance becomes less effective across the board the later the game goes on against any deck. Turn 4 displays its biggest positive impact and its arguable effectiveness in a top decking war in the late game does not show up in the data.

So how good is Flamewreathed Faceless?

It seems to be a moderately good card: powerful in some situations, weak in others. Of course, many variables come into play with the statistics that we present, and our philosophy is to never draw conclusions without firm evidence. We believe that the data does not support the notion that the card is overpowered in the current state of the Metagame. We emphasize that we cannot say what would happen if the card was dropped from the deck and replaced with something else, and we cannot estimate the impact on the opponent’s decision making accounting for the possibility of Flamewreathed Faceless being played.

nw_banner_static_2015_paladin_728x90_EN

As the writer of the article myself, I was genuinely surprised at these statistics. There have been many cases where the card is played against me and I am frustrated to not have an answer or a good line of play against it. This is a great example of how possible individual biases (confirmation bias in particular) can affect how individuals, and the community as a whole, might develop its perception on the strength of cards. Sometimes though, these perceptions do not line up with empirical evidence.

Finally, one of Hearthstone’s game designers, Iksar, recently tweeted about the subject, which we also found interesting:

“4 7/7 commands such enthusiastic opinions. Maybe most of it is how it feels to not have an answer for it that early in the game.”

“Things like Doomhammer or Trogg feel like they fly under the radar even though you might expect both to command equal enthusiasm.”

We are looking forward to the community’s thoughts and feedback on these findings and will continue to do our best in order to produce interesting analysis that will shed light on many hotly debated subjects in the game.

Our Data Reaper Project, including the Data Reaper Live (Beta) now has over 1,300 contributors. Without them, analysis such as this would not be possible, so we’d like to thank all of our contributors for helping us on this project. If you have not done so already, you can sign up with your Track-O-Bot information here:

Sign Up - Data Reaper Report

Invitation to Contribute

Our team is looking for people who are interested in getting involved in writing articles such as this one. We are happy to provide qualified individuals with data so that they can analyze it on their own, and then develop an article that we can publish together. Our goal is to expand the community’s knowledge about Hearthstone. We know that there are plenty of qualified individuals out there. If you are interested, please contact us here.

This article was put together by:

endofdays-2Tzachi

5 Comments

  1. I think in such a smorc format of course flame wreath isnt OP this shows in the stats. but if they nerfed some of the fastest cards in the game to slow standard down, I feel like something like flame wreath would therefore get better. leading to its eventual nerf. until then im rooting for either a nerf of all the top aggro decks or a printing of new cards to help police them, for all the classes like what warrior has (fiery war axe)

    • No offense, but did you even read the article you are commenting on? The two strongest decks against Flamewreathed are Control Pally and Miracle, both decks that struggle in faster metas, but thrive against slow decks. The same is true of Midrange Shaman and Freeze Mage, who also deal well with it. The only deck that can reasonably called aggro that deals well with it is Zoo. So based on the data, the opposite is much more likely: If the meta gets slower, flamewreathed will completely get cut and not used anymore. This makes sense since in a slower meta decks will generally run as much hard removal as possible.

  2. I think a huge reason for why people are so against the card is that if it is played on curve and you don’t draw an answer to it you basically lose on the spot. Most aggro win conditions in Hearthstone don’t get dropped until turn 5 (doomguard, leeroy, doomhammer, arcanite reaper, etc) so being able to drop this thing a full turn earlier (while also having arguably the 2nd strongest early game curve of minions in standard) can make the card a bit overwhelming for some decks.

    Some classes just don’t have efficient means of dealing with a 7/7 on 4 because they’re usually very far behind on board. Just look at Priest for example, yes they have a fantastic answer in Shadow Word: Death, and if they manage to draw it and play it against an on curve 7/7 they’re in decent shape, but if they whiff there is pretty much no way in the entire deck to deal with it at all so you’ll lose automatically. I personally would really like to see the winrate of aggro shaman when 7/7 gets to attack on turn 5, as that will give us an idea as to just how important having and playing an answer to it is.

  3. This supports what I have said to people many times about the card. It is my belief that the main thing that makes the card strong is that it comes down AFTER playing a very strong early game. Handlock commonly played an 8/8 on turn 4 without causing the frustration that people aim at Flamewhreathed. I believe they don’t understand where the real power in a shaman deck lies. I truly believe that if Flamewhreathed was replaced with Fireguard Destroyer or Venture Co. Mercenary, the win rates of the deck wouldn’t change that much.

    What I would be more interested in seeing is the effect Tunnel Trogg played on turn 1 or on turn 2 after coining out Totem Golem has on win rates.

  4. What about comparing having it in hand vs not having it in hand? That shouldn’t be as biased playing vs not playing, since aside from mulligan, you don’t have any control about what you draw.

    Anyway, this doesn’t surprise me. I’ve cut flamewhreathed from my midrange shaman after having used it for quite some time. In a deck that otherwise uses nothing big, it’s just too clunky and easily removed. In the mirror, I love my opponent playing it since it’s basically the only decent hex target he’ll have.

Comments are closed.