A few months ago, before Karazhan’s release, we published an article with data on the win rate differentials between going first vs. having the coin. Now that we’re well into the MSoG era, we’re revisiting the subject, in order to see the impact of turn order on the success of archetypes in the new Meta. For this analysis, we’ll be using the matchup data of the past three weeks, which comprises of 260,000 games.
First, we look at win rates for common archetypes in the Metagame. In the following table, we’ve calculated the “coin differential” in win rates against the field of every archetype. The number we are showing is the difference in win rates between going first vs. having the coin. A positive value indicates that going first has a better win rate, while a negative number indicates that this archetype wins more often when having the coin (in the Metagame of the sample).
Note that the calculated global advantage of going first is 3.3%, meaning that going first on average beats the coin 51.65% to 48.35%. In our first analysis of this subject, back in July, the global advantage was exactly the same at 3.3% (!).
Generally, tempo based decks hover around the top end of the table. Zoo Warlock is the archetype most dependent on going first, with an extremely high score. This perhaps hints towards the struggles of this archetype currently, since it doesn’t have a reliable way to win back the board when it’s behind. Several Shaman archetypes sit at the top end of the table, which isn’t a surprise since Shaman is blessed with an extremely powerful early game, and leverages a board advantage better than any other class.
At the bottom half, we see very different types of decks. Reno decks generally favor the coin, as having the extra card in the mulligan phase means they are more likely to get one of their swing cards in the opening hand. The same can be said for other control decks, like Control Shaman, which favors the coin more than any other deck. Reno Warlock has a stronger affinity to going first than the other two Reno archetypes, likely due to its powerful proactive turn 4 plays of Mountain Giant and Twilight Drake.
There are, of course, exceptions. Though Miracle Rogue did get an extremely powerful turn 1 play with Small-Time Buccaneer, statistically, it still favors the coin more than going first relative to the field. Miracle Rogue utilizes the coin better than any other deck as a tool to initiate combo’s and generate massive tempo swings, with Edwin Van Cleef being a particularly notorious beneficiary. There’s a reason Miracle Rogue runs 4 coins in its decklist, after all.
Pirate Warrior is the biggest exception. Even though it’s the most aggressive deck in the game, it performs significantly better with the coin, and this was also true back in July. It’s harder to pinpoint the exact reasons, but you can come up with a few. The increased chance to have Fiery War Axe in the opening hand is likely a big factor, since the tempo gained from playing it on turn 2 is bigger than the tempo gained from going first. The deck also has powerful turn 3 plays, while not having very impactful standalone turn 2 plays other than War Axe, which generally increases the strength of the coin. Finally, the extra card in a deck that doesn’t have any form of card draw, and relies on drawing an adequate amount of damage to win, might also be a factor.
Next we’ve prepared a special matchup table. In this table, we’re showing the “coin differential” of specific matchups. A blue color indicates that the deck in the row wins more often against the deck in the column when it goes first, while orange indicates it wins more often when it has the coin. The colors do not signal the actual win rates, just the “coin differential”! You can hover over the boxes for the win rate information.
Generally, when two decks with an above average coin differential face each other, there’s an advantage for going first. Similarly, when two decks that favor the coin face each other, there’s an advantage for having the coin. It’s important to note that some exceptions to this rule exist and can be observed.
When decks that have polarizing coin differentials clash, results may vary. For example, in the Aggro Shaman/Reno Warlock matchup, the difference between going first and having the coin is a wash. While Aggro Shaman generally benefits from going first, the extra chance that the Warlock draws a key card in the mulligan phase seems to cancel out this advantage.
However, against Reno Mage, Aggro Shaman has a greater chance of winning when going first, even though Reno Mage has a lower global coin differential than Reno Warlock. As we’ve said, this isn’t intuitive and depends on the matchup. Perhaps against Reno Mage, Aggro Shaman’s best chance is to go as hard as possible and try to win before turn 6, which is a line of play supported by going first. Alternatively, it might also be that Reno Mage isn’t as reliant as Reno Warlock on drawing a particular card to beat Aggro Shaman, so in this matchup, it prefers the mana advantage.
Our Data Reaper Project, including the Data Reaper Live (Beta) now has over 2,400 contributors. Without them, this project would not be possible, so we’d like to thank all of our contributors for their help. If you have not done so already, you can sign up with your Track-O-Bot information here.
This article was brought to you by:
Very nice content.
Thanks a lot.
Excellent article, thanks very much.
Very nice! While the weekly reports are extremely useful, my favorite posts are these occasional analysis about The Coin, specific cards etc.
I’d like to suggest that you do one on Patches. I’d like to know, out of games it is played, what % of times he is played for free and what % his ability fails. I’d also like to know the difference in win % in these scenarios.
To be clear, I want the games he is played divided in “Free Patches” when his ability triggers and “Fail Patches” when someone either plays him from hand or plays any pirate and the ability doesn’t trigger (to account for games where he is drawn but never played). This assumes, of course, that no one plays any Pirate without him in the deck (which I think is reasonable).
The analysis of both frequency and win rate of each scenario could be shown by Class, by Archetype and perhaps by Coin vs No Coin and/or Turn Played, depending on what you deem most interesting and availability of data
There’s an easy way to balance the advantage of going first: Give the second player an additional coin on a subsequent turn. Experimentation will determine the ideal way to do that, in order to even the score. For example, give the second player an additional coin on the fifth turn. If that proves too powerful, then try the sixth turn in the next month. Etc.
– toby robison, princeton, nj usa
The 3,3% doesn’t sound like much, but given the 51,65% win-rate you actually are 7% more likely to win than your opponent, when going first. Now that’s a number that seems slightly unfair.
This makes no sense. 51.65% – 48.35 = 3.3 %. Where do you get your 7%?
This is less than chess have for going first. I don’t think it’s really a big problem in Hearthstone.
They get it from percent change. You can get that by doing (51.65-48.35)/(48.35). 51.65 is 7% larger than 48.35.
the columm for control shaman in the chart is missing
Big thanks,i waited for this article so long…