Don't Judge a Book By Its Cover, and Don't Judge a Card by It's Win Rate
I constantly see people quoting card win rates. Cards with high Win Rates are good, and low Win Rates are bad...
...right?
Let's dive in!
What is a card's "Win Rate"?
If you go on GUdecks.com you can view a cards Win Rate over the last X days and at Mythic (or all) ranks. This will tell you the percentage of games won with that card in the deck in that timeframe.
People have quoted this number to me many times in the last two days, so I was compelled to write this.
One person, in response to me calling Obelisk and Necroscepter OP saying they will likely soon be nerfed, said "you're wrong, Obelisks Win Rate is only 49%". Therefore Obelisk is a bad card, right? (Spoiler alert, it is good!)
These cards are "bad!"
Another person said they knew someone who would purposely use off-meta cards to pump their Win Rate, then sell them claiming they had a high Win Rate therefore they were good.
Yet another person asked me to play a deck they had made, built from all the highest Win Rate cards, so it must be good, right? (Oof, the deck was a complete mess.)
All these things could be avoided if people understood what card WR really meant, and what it doesn't mean.
What does it tell you
The Win Rate tells you that the card was in a deck before a match started, and if the deck won or lost. That is all.
What it doesn't tell you
Was the card even played? The card could've remained in your deck the entire game, having no bearing on the outcome whatsoever, and yet win or lose, you're changing that cards Win Rate.
What cards were played in those games? Was the deck a meta deck? Or was it a starter deck with one extra card thrown in? If it is the former, the card could easily stand on the decks merits and not its own, and if it is the latter, even having the best card in the game won't get you wins if you're using a starter deck in Mythic.
The Experiment
I'm going to be playing Dense Fog, one of the worst cards in Gods Unchained. Any high level player can participate as well, with the sole goal of getting a > 70% Win Rate in Mythic for Dense Fog to completely demonstrate the illusion of a "great card" based on Win Rate. I'll report back when I have successfully made Dense Fog the best card in Gods Unchained!
How am I going to accomplish this? Play a 29 card meta nature deck and add in 1 single Dense Fog. So far I'm 9-1 and the community has bumped the Win Rate to 81% over 21 games.
Before someone says, "but sample size!" Hold up, we'll get to why a large sample size isn't enough.
But Genesis Cards Are So Good Look at their Win Rate!
One big reason WRs are skewed is money. Some cards cost a boatload of money. For instance, Reap is more expensive then other single target removal in death, such as Ray of Disintegration. But guess what that means? The likelihood of the person playing that Reap also having an insanely expensive deck is very high. Likewise the person playing the completely free Ray is likely playing a completely free, or at least far cheaper, deck.
Ex. You're not going to find a slightly modified beginner deck that also includes a single Reap, but absolutely will see Ray of Disintegration in these decks. So things like cost of a card completely change the Win Rate based on the decks the cards are being played in.
Before you say, "Copper, just filter down to Mythic to get rid of all the bad players" I will just say two things.
Matchmaking is so bad that obscene amounts of players who aren't good enough to compete at the top level make it into Mythic. This is compounded by the insane amounts of deranking (while conversely simultaneously ranking up players who shouldn't be ranked up by giving out copious undeserved wins!)
Cards "played at Mythic" include "played against players beneath Mythic". And if community sample size is any indication, between 25-40% of matches for Mythic players are against players who are not in Mythic. Guess who's gonna win that one 80% of the time, regardless of which cards are being used!
The Nemean Lion / Legendary Issue
Once upon a time I famously built a Nemean Lion meme deck which I played 30 times over the course of 2 days on stream. I drew the Nemean Lion exactly ZERO TIMES. (We were, and I still am, convinced it was a bug with the shuffler and it kept putting it on the very bottom of the deck).
The meme was a total failure, but more importantly, I drove the Nemean Lions Win Rate way down (cuz the deck was very bad without the Lion who all combos revolved around for fun), without ever even drawing it a single time!
Legendary cards can only have a single copy in a deck, so they can easily go multiple games without being seen. So their Win Rates are more likely to just hover around whatever the decks Win Rate happens to be. Which brings us to the next point...
Meta Analysis
One card nerf can completely change everything. Death Board Wipe was the absolute best deck a few weeks back, two players went perfect with it, and the following weekend Villibesti followed up his 25/25 weekend with a 24/25.
Then a single card was nerfed - a card that WASN'T EVEN IN HIS DECK - and Death Board Wipe ceased to be the best deck. (Don't get me wrong, Death Board Wipe is still strong and boring, but that's another topic! Suffice to say, it is no longer the best deck.)
What? How could this happen?
Guild Enforcer had 2 Armor and Frontline. For a few weeks, Aggro as we knew it ceased to be playable. Guild Enforcer could easily shut down entire games by itself against fast aggro decks where most creatures only have 2 or less strength. Because aggro was shut down, Death Board Wipe didn't have to be concerned with its least friendly matchup - aggro Slayer - and could easily play the most insanely greedy list I've seen in a long time (two 9 drops!).
Finally Guild Enforcer was nerfed down to only having 1 Armor and Aggro came back.
So if you're attempting to identify "good cards" but there are certain imbalances in the games meta, card Win Rates can easily be skewed. Were all aggro cards "bad cards" for a few weeks? No, they just needed an opportunity to shine, and Guild enforcer was holding them back!
This goes to my point about being told the Obelisk being a "bad card" because of its Win Rate. Just because zombies aren't top tier meta right now doesn't mean Obelisk and Necroscepter don't completely have the ability to singlehandedly win you games. Having 4x three Mana cards in your deck that can just instantly win you a game if your opponent doesn't have an answer is pretty insane. Add just a little support for zombies and you have bona fide broken cards on your hands.
Net-decking
A large portion of the Gods Unchained community net-deck. What does that mean? It means they look up a deck that a top player played and then play the same exact deck themselves.
There was recently someone posting a decklist in my discord and the list was INSANELY AWFUL. But not only had they played it a bunch themselves, it had nearly 1000 plays from all players!!!! Why were all these people playing this god awful deck??
Someone on YouTube told them it was good.
Narrator: It wasn't.
It had so many anti-synergies for a magic deck (ex. Lacking low cost spells but including Mana Thirst Golems; including Deuteria which will thin out their card draw and they were already lacking any ways to draw), a horrible curve, and a ton of singletons. Yet 1000 matches were played with it!
If you run this creature, you MUST be running a ton of spells!
Never underestimate the power of Net-decking on a cards Win Rate. If you build an expensive deck that only serious players who put in money can play, it is less likely that the Win Rate will be brought down by new players. New players won't destroy the Win Rates of cards that cost a lot of money, but boy oh boy were they destroying the Win Rate of Starshard Bolt (which they only ran 1x of, and no Tracking Bolts at all!).
I won't say more about this deck as I'm not trying to shame the person who built it, nor the people who netdeck. Both of these things are fine, just be aware Starshard Bolt is a PHENOMENAL card, don't worry about what the Win Rate tells you!
Which by the way, Starshard Bolt at Mythic over the last 3 days is at 48% Win Rate.
What a trash card!
At 30 Days, Starshard Bolt is also at 49%, so no cherry picking data here.
Large Sample Size // Closing Thoughts
But Copper, come on, we have large sample sizes now! Therefore the cards Win Rates will all balance out!
Sure! To some extent this is true. On the other hand, things like the price of cards, availability to new players, the meta, Net-decking, Legendary cards being tougher to draw, and many more underlying factors all play into a cards Win Rate.
The one stat that could make Win Rate slightly more accurate is a "was the card played?" flag. If the card wasn't even played, it probably shouldn't be tanking or boosting the Win Rate of a card!
I'm sure people will have a lot of thoughts on this, and rest assured, I don't believe Win Rate is completely worthless. But just take it as a loose guideline and use common sense when considering other factors like "is a 3 mana card that has backline and perpetually spits out zombies good?" or "is a 2 mana spell that deals 4 damage to any creature good?"
~Copper
Source for card Win Rates: www.GUDecks.com
Check out Copper's YouTube
Follow on Twitch, Live M/W/Th 6-9 EST
Join Copper's Crypto Cave Discord for P2E info
Follow Copper on Twitter