About
Learn about this site
About
Hello fellow Magic player,
I'm Dee. I write MTG strategy articles for Magic Game Plan. I was inspired by Zach Lowe of Grantland to improve my articles by adding statistical analysis. Zach is one of the best NBA writers today. One of the reasons why is his use of advanced metrics to help people understand basketball.
I wanted to add stats to my articles like Zach. However, I couldn't find the MTG stats I was looking for. So, I learned how to code and built my own stats site. The result was MTG Analytics.
I used a responsive design so the site looks decent on tablets and smartphones. For example, on your smartphone, go to one of the decklist pages like this one and then click a card. The card image will pop up. :) Or open the site with a desktop or laptop and resize the window. The site will resize to fit your screen.
On some pages, you will have to use the landscape or wide display on your smartphone to see everything.
I hope you find MTG Analytics useful. I'd love to know how you use it. You can contact me here. Also, please contact me if you see any bugs. Thanks!
You can support the site by clicking the TCGplayer or eBay links on the decklist and card pages and buying something. I get a small comission for each sale. Also, I would really appreciate it if you tell your friends about MTG Analytics.
Below, I talk about how the stats are calculated and different features of the site.
Deck Metagame
With Starcitygames running a big Standard tournament almost every week, the metagame usually changes pretty quickly. You'll often have a deck that does well early in the season but stops producing top results as the weeks go by.
On the other hand, you'll have other decks that don't do much early but after a few weeks, they become a major part of the metagame.
I wanted a site that kept up to date with these metagame shifts so I came up with an algorithm that takes into account the age of the results. Newer results are weighted more heavily than older results.
For example, when the season is two weeks old, the results of the second weekend are weighted at 60% while the first weekend is at 40%. (A new season starts whenever a new card set becomes legal for tournament play.)
At three weeks old, the algorithm uses these weights: third weekend is 50%, second weekend is 30%, and first weekend is 20%.
Once a season reaches four weeks or more, the site looks back at the last four weekends and adds the following weights.
| Time Period | Weight |
|---|---|
| Total | 100% |
| Last weekend | 40% |
| Two weekends ago | 30% |
| Three weekends ago | 20% |
| Four weekends ago | 10% |
| Five or more weekends ago | 0% |
Notice the 0% for results five or more weekends ago. Players are quick to move away from non-performing decks so if a deck does not have any results in the past four weekends of a season, it gets dropped from the metagame.
Along with the age algorithm, MTG Analytics also takes into account different finishes and tournaments. For instance, first place decks are weighted more than 16th place decks and a Starcitygames Standard Open is given more weight than a Starcitygames Classic tournament.
The site includes the following tournaments to calculate the metagame percentages.
Tier One- Pro Tour
- Grand Prix
- MaxPoint Series Championship
- Starcitygames Standard Open
- Starcitygames Invitational
- TCGplayer Open 5K
- Starcitygames Classic
Tier one tournaments use the following weights for finishes.
| Finish | Points |
|---|---|
| 1st | 8 points |
| 2nd | 7 points |
| Top 4 (3rd-4th) | 6 points |
| Top 8 (5th-8th) | 5 points |
| Top 16 (9th-16th) | 4 points |
For tier two tournaments, the site uses the weights below.
| Finish | Points |
|---|---|
| 1st | 2 points |
| 2nd | 1.75 points |
| Top 4 (3rd-4th) | 1.50 points |
| Top 8 (5th-8th) | 1.25 points |
| Top 16 (9th-16th) | 1 points |
For each weekend, the site adds up the points per deck and divides that number by the total number of points to get the metagame percentages.
I don't include Magic Online tournaments because they have a different metagame than real life. For example, here were the top 10 Magic Online decks back in June 2013.
| Archetype | Percentage |
|---|---|
| R/g Aggro | 15.4% |
| Junk Reanimator | 13.2% |
| Junk Aristocrats | 7.4% |
| Jund Midrange | 7.3% |
| The Aristocrats | 5.7% |
| UWR Control | 5.7% |
| UWR Flash | 4.8% |
| Bant Hexproof | 4.5% |
| Naya Aggro | 3.5% |
| Naya Midrange | 3.5% |
Compare that to the real life metagame below.
| Archetype | Percentage |
|---|---|
| Jund Midrange | 13.43% |
| Junk Reanimator | 10.68% |
| Naya Aggro | 9.46% |
| Bant Hexproof | 9.27% |
| Esper Control | 8.68% |
| R/g Aggro | 6.05% |
| Four Color Reanimator (GWBR) | 5.48% |
| B/G Control | 5.38% |
| The Aristocrats | 4.71% |
| Naya Midrange | 4.08% |
As you can see, the metagames are pretty different. My theory is that cost is more of a factor online so players tend to play a cheaper deck like R/g Aggro. Also, I bet there are a good number of online players playing multiple games at once. It’s much easier to do that well with a fast aggro deck.
Including Magic Online tournaments will skew the metagame data and make it less accurate. Plus, I think most real life players don’t follow the online metagame so it won’t influence their deck choices.
Card Metagame
When players talk about the metagame, they are talking about decks. But what if we dug deeper and figured out which cards showed up the most in the top finishing decks? What if we analyzed the "card metagame" as well as the deck metagame?
This type of data mining gives us more info to make good deck choices.
MTG Analytics calculates a percentage for every card in the deck metagame. The percentage is basically the total number of copies of the card divided by the maximum number of copies that players could have had in their decks.
For example, let’s say the percentage for Abrupt Decay is 20% and there are 100 decks in the metagame. The maximum number of Abrupt Decays in all the decks is 400 (100 X 4). 20% means there are 80 Abrupt Decays in the metagame (400 X 20%).
All those Abrupt Decays could be split up in various ways. There could be 20 decks with 4 copies each and 80 decks with 0 copies.
(20 X 4) + (75 X 0) = 80
Or there could be 10 decks with 4 Abrupt Decays, 13 decks with 3 copies, 1 deck with 1 copy, and 76 decks with 0 copies.
(10 X 4) + (13 X 3) + (1 X 1) + (76 X 0) = 80
40 decks with 2 Abrupt Decays and 60 decks without any copies is also a possibility.
(40 X 2) + (60 X 0) = 80
Basically, the higher the percentage, the more likely you will see the card over a tournament.
Another way to look at it is copies per deck. You can multiply the percentage by 4 to get the average number of copies per deck. So, in the above example, the average copies per deck of Abrupt Decay is 0.8 (20% X 4 = 0.8).
Also, the card metagame is modified by the same algorithm as the deck metagame. In other words, the cards of a newer, higher finishing deck from a Grand Prix have more weight than the cards of an older, lower finishing deck from a Starcitygames Classic tournament.
I didn’t include lands that don’t have spell effects because that data doesn’t really help.
You can sort the card metagame by main deck, sideboard, and price by clicking the table headings. The price is the midprice from TCGplayer.
If you click a card name, you'll see the deck metagame of the card. On that page, you can dig deeper into the data by clicking a deck. This will give you a list of decklists along with the number of copies of the card for each decklist. At the top, you'll see the average copies of the card for the deck.