Win rate

A win rate refers to the percentage or proportion of games that a deck or player typically wins. The term is usually used as a measure of success, and effectiveness in the current meta, or in specific matchups.

Discussion
Win rates are typically used either as record of a specific player's performance, or as a record for predicting future success with a given deck. In Ranked ladder play, an estimated positive win rate in the current meta is often used to determine the speed at which a deck will allow a player to ascend, although this is a large oversimplification of the process involved. Decks with higher estimated win rates will in theory win a larger proportion of their matches, and thus reach higher ranks more quickly, which given the finite length of each Ranked season is often conflated with the likelihood of reaching Legend that month. However, due to the various decks and deck types seen in the meta, in practice even most successful decks have a win rate fairly close to 50%.

The actual observed win rate will also vary strongly depending on the player's current rank, as well as their ability with the deck, with highly skilled players seeing far higher win rates with a given deck than the average player. Ben Brode states that top players have been known to exceed average deck win rates by at least 30%, such as achieving a 70% win rate with a deck with an average win rate of 40%.

Matchup win rates
Win rates are used in specific deck matchups to predict each side's chances of success. Strong/poor matchups are often speculated to be as extreme as 90%:10%, while 45-55% matchups can be considered fairly even.

The proportionate combination of the specific matchup win rates for each deck seen in the current meta adds up to the deck's overall win rate; overall win rate is therefore highly dependent on the current meta. For example, a deck that fares poorly against currently dominant decks will likely fail to succeed, even if it has an extremely high win rate against most other decks.

Similarly, even the most successful deck in the current meta will have some decks against which it has poor chances of success. For example, in January 2017 one of the most successful decks in the meta was Aggro Shaman, with around a 53% win rate overall; but when faced against Control Warrior, this shrank to a 35% win rate, with the deck losing nearly 2 games out of every 3.

Overall win rates
Overall win rates also vary strongly depending on the ability of the player using the deck. While any player will learn to improve their win rate (at a given rank, or in a given mode) over time, more complex or simplistic decks can present a strong difference in effectiveness at different levels of play: decks that are easy to play well may see a lot of success at lower ranks, but far less at higher ranks, where more complex decks outperform them; while decks with a high skill-cap often perform very well at higher ranks, but see some of the worst win rates at lower ranks. This can produce a misleading impression of the effectiveness of a deck: a professional player may be able to get great results with a complex deck, but the average player's results may be far poorer than with 'weaker' decks which they are more easily able to grasp. For example, during its peak Patron Warrior was one of the most successful decks in the game at high levels, but one of the worst decks for a large percentage of the lower ranks. This variance is not limited to deck complexity, but also changes to some degree with the different cards and decks seen at different levels of play.

In contrast to the extreme win rates seen in certain match-ups, overall win rates for even the best decks tend to be very modest, with even highly successful decks usually having only slightly better than a 50% win rate. The difference in win rate between the top decks also tends to be small. For example, in January 2017 the best deck had a win rate of 53%, while the eleventh best deck had a 49% win rate. This often contradicts player experience, where certain decks (usually the opponent's) may be perceived to win "every game". Dean Ayala states that "very, very few decks" have ever reached a "dangerous" win rate, even over 55%. Pre-nerf Undertaker Hunter, the most dominant deck in the game's history, only had a win rate of 60%.

Despite this, decks that are perceived to have a high win rate frequently become highly popular, due to a wish to play an effective deck. Contrary to popular opinion, the main problem with 'overpowered' decks tends not to be an excessively high win rate but excessive population size, with the deck being seen in too large a percentage of matches, resulting in player frustration and boredom. Many decks with high win rates have moderate or even low population sizes, and thus despite being highly effective never threaten the balance of the meta or the fun of players, consequently tending not to become the subject of complaints. For example, in January 2017 an Anyfin Can Happen Paladin Murloc deck had the fourth highest win rate at 51.5%, but with only 0.7% of the overall population.

Balancing the meta
Decks that become too popular may threaten the balance of the meta, in the most extreme cases leading to card changes or nerfs. While the primary concern for meta balance is not win rate but popularity and diversity, win rate is taken into account when deciding whether to implement changes.

Laddering speed
According to Ben Brode, win rate is more important for increasing laddering speed than match length. This contradicts popular opinion, which frequently considers fast decks such as Face Hunter to be optimal for climbing the ladder (assuming they are currently favoured by the meta) due to their short matches, allowing the player to complete more matches - and therefore gain more stars - in the same period of time. While essentially correct, Brode states that this is only true if the win rates of the decks in question are "very close", stating that if a slower deck has a higher win rate, the player will gain rank more quickly by playing that, regardless of the longer game times. He also states that the average game lengths for typically fast or slow decks aren't "as different as you might assume".