What is Farsig.ht?
Farsig.ht runs free competitions to predict which Hearthstone cards will be popular - the "meta" - in a new expansion. The current competition is for The Boomsday Project.
Many Hearthstone players like predicting the impact of new expansion cards but there isn't currently a way to track these predictions. We think a competition is a fun way to learn who is actually good (and not!) at predicting the Hearthstone meta.
How do I play?
Sign up with your Battle.net account (we can't access your password) and your email address. Then rate each card from 1 to 5 stars, based on how often you think the card will be in standard ranked decks.
The more accurate your predictions, the more points you earn. The top scorers will win prizes.
The competition is free and no purchase is necessary to play.
When is the submission deadline?
Entries for The Boomsday Project closed on August 7th, just before the expansion launched. Entries for the next expansion will open soon after its first cards are announced around November.
What is the rating and scoring system?
You rate each card from 1 to 5 stars. Stars reflect how much you think a card will be used in the standard ranked meta.
A 5-star prediction means that you think the card will be in the top fifth (20%) of usage across all standard ranked cards. A 1-star prediction means you think it will be in the bottom fifth.
For every correct prediction of a card’s actual star rating, you earn 5 points. If your prediction is 1 star off, you earn 4 points; if you're 2 stars off, you earn 3 points, and so on. Every card prediction earns at least 1 point.
|Your Prediction||Points Earned|
|Correct # of stars||5|
|1 star off||4|
|2 stars off||3|
|3 stars off||2|
|4 stars off||1|
|Your Prediction||Actual Meta||Points Earned|
For the 135 cards in a new expansion, a perfect score would be 675 points.
In the future, we might score sleeper picks higher - e.g. correctly predicting Stonehill Defender as a 5-star card when the community thought it would be 1-star - but we want to keep the scoring simple for now.
What’s your source of data on the meta?
We’ve partnered with our friends at HSReplay to use their statistics. They have the largest data set: over 8 million games per week as of August 2018. Players contribute this data by using Hearthstone Data Tracker. We recommend installing it to improve your game and contribute more data.
This isn’t a perfect measure of the meta - only Blizzard has that - but it’s the best currently available.
When are prizes awarded and what can I win?
At the end of each month, we’ll compare everyone’s predictions to the actual Hearthstone meta. Players will earn points according to our scoring system.
On September 30th and at the end of the expansion (early December), the top three scorers will win the following:
|September||End of Expansion|
|Perfect Score||100 packs||$1,000|
|1st place||60 packs||60 packs|
|2nd place||40 packs||40 packs|
|3rd place||15 packs||15 packs|
|477th place||47.7 packs*|
* We're serious. Thanks to NanashiSaito on Reddit for donating this prize.
If there’s a tie and the prize can be split, it will be split between winners. If a prize can’t be split, it will be awarded randomly among the tied players.
As the meta changes, the winners at each point may change as well. Everyone has a chance to win.
Why are you scoring cards on their frequency in decks?
There are many ways to judge cards and each has pros and cons. We wanted a metric that was objective, measurable, simple, and reflects how the Hearthstone community likes to predict the meta.
Some metrics we considered:
Frequency in decks
This is how often a card is in played decks. We think it's the best reflection of the meta - what people are actually playing - and a reflection of perceived quality.
However, it skews toward lower-rarity cards that people are more likely to have. It also doesn't reflect quality as much as winrate. Cards like Yogg-Saron can be in decks because they're fun or hyped, but not necessarily the most effective.
We chose frequency in decks because many people seem to like predicting the meta. If you can predict the meta, you can beat it. In the future, we might try a contest on a different metric.
This is how often decks win when containing the card. It may be the best measure of a card's quality.
However, it's also closely related to a deck's quality and synergies. Crystal Lion went from a 1-star card to a 5-star card once Dude Paladin became strong. Mysterious Challenger turned previously unplayed secrets into strong cards.
While a few cards like Dr. Boom were strong in almost every deck and meta, most cards are meta-dependent. We think a card's quality should be judged on its effectiveness within a meta, not in a vacuum.
This is how often a player wins when the card is played. It may be the best measure of a card's game-ending (“finishing”) ability.
This skews toward niche, higher-mana cards like Pyroblast and Bloodlust, which have had played winrates over 70%. It doesn't measure a card's frequency in the meta or how decks containing it perform when the card isn't played.
This measures how much people like a card. It’s probably what's being measured most when players rate cards without any instructions.
The aggregate of card popularity ratings may measure how much people are enjoying Hearthstone, but it doesn't directly measure the meta or card quality.
Why isn't there a division for arena, wild, casual, or legend-only?
We’d like to keep things simplen. We’re also not sure if people want to rate cards for multiple modes. Let us know if you do.
Why are you doing this? How do you make money?
We just like Hearthstone and wanted to create something fun. We're paying for the prizes out-of-pocket for now. We’re open to advertising or sponsorships to increase the prize pool but only if we can maintain a good user experience.
What if I didn't submit my entry on time?
Sorry, entries must be submitted by the deadline. If you missed the current competition, you can play in the next one.
How do you handle nerfed cards?
We can’t predict whether any cards will be nerfed so we’ll just measure the monthly data of a nerfed card like any other. A 5-star card in one month may be nerfed and become a 1-star card in the next month. If you think a card is so strong that it will be nerfed, you can rate it for the expansion accordingly.
How do I get pro player status on the site?
We define pros as people who either a) are ranked in the top 500 standings of the World Championship Tour in any region, or b) have over 50,000 followers on Twitch.
Please share your prediction from a verified social media channel, such as your Twitter or Facebook account. If you meet our pro player criteria, we'll update your status.
How can I suggest a new feature?
Feel free to contact us about potential features.
How can I help out?
We appreciate any support! We’re open to:
Programmers who want to help us deliver features
Desginers who want to improve user experience
Contact us if you have any questions. Thanks!