Commander Dashboard
Magic: the Gathering is one of the most popular games in the world, and has been leading the TCG market since its inception.
The most popular format is known as commander, where 4 players with decks of 100 unique cards battle it out in a fun and casual environment.
But what if we tried to analyze the decks to find trends in what decks win and how they do it?
Thanks to the Command Zone for collecting the raw data for this analysis
How land count correlates with winning and deck types
Taking into account that a losing deck can end a game in the first few turns, and therefore will skew the losers numbers down, we still can draw insights on how to improve. By also including number of cards in deck at the end of the game we can slightly control for early player removal.
Decks that won consistently had 1-2 more lands in play compared to losing decks
What is immediately apparent is that there is a strong correlation with winning and having the most lands in play. As discussed before that could be related to not only the length of time in the game, but also the two of most valuable card effects putting lands into play and drawing.
Most deck types are in roughly the same pattern with mid-range at the top in terms of mana producing lands with aggro, voltron, and combo in a clump. Control is the big shift, moving from the lower clump in the losing decks to nearly matching mid-range in winning decks. Control does seek to win a long game and cast power house spells so it makes sense that when they are low on mana they are likely to lose.
Interestingly if we look at cards remaining in deck (the size of the bubbles) there is very little distinction between winning and losing decks, and particularly little when comparing deck types within the winners and losers.
Winning decks generally had 10 or so fewer cards left in their decks compared to losing decks and that was a consistent trend across all deck types. Comparing losing to losing and winning to winning the biggest difference is just 5 cards.
Now that we can look into cards remaining in deck we can compare lands at end of game to learn about the level of ramp and card draw that took place.
As expected the least cards left in the deck are control and combo decks, decks that highly value card draw. Interestingly control moving from the lower clump into the higher clump into the land count, while also maintaining relatively few number of cards in deck, indicates that it is mostly card draw and not land ramp happening in the losses.
With this insight we can look more closely at the control stats in the lands remaining. We can surmise that due to a high amount of card draw, a late game strategy, and a weakness in the early game that the jump in lands in play correlating to win rate is better attributed to a long game where the control player routinely hits land drops every single turn, due to high card draw, rather than any amount of ramp or lands in deck.
In contrast for mid-range decks the difference in a win and loss, in terms of cards left in deck, is relatively large. Since mid-range sees the most lands on the field in both wins and losses, yet in losses has many more cards left in the deck, winning mid-range decks implement more card draw than losing decks. Or at least are given the time for their ramp to pay off and stay in the game longer.
While card draw and ramp are seen as the two most valuable effects in the game, they need each other and also push each other away. In order to take advantage of the increased number of lands, you must have cards in hand to play. Conversely in order to play cards from your hand, you must have lands in play to pay for them. This illuminates the strength of cards such as Oracle of Mul Daya, Tatyova Benthic Druid, and Chulane Teller of Tales because they combine playing lands and drawing cards into one mechanic.
Control and mid-range are the standouts in disparity of the land count. Whether it is card draw, land ramp, or length of game.
HOw does color identity affect strategy?
Using how we analyzed land count and card draw as generically powerful, how does adding more colors pull decks towards or away from a given strategy?
For this analysis we will treat card draw and ramp as strategy agnostic, being powerful in nearly any strategy, and we will see if adding more colors pushes decks toward generically powerful good stuff.
A trend towards more mana winning
As we take a look at these three graphs we can see a general trend that as we add colors to a deck the more lands winning decks have compared to losing decks. Monocolored decks have two strategies beneath the median line, and we see the disparity grow with winning and losing decks grow to the point where for three color decks all winning decks are above the median and all losing decks are below the median.
There are a few initial takeaways, first of which is that the more colors you have in your deck the more potential there is for you not to have the colors that you need to cast your spells. This has the knock-on effect of reducing your ability to use ramp or card draw spells to maintain a healthy land total. Additionally the more colors are in your deck the wider the pool of cards to choose from is, meaning that you can take the best generically powerful cards from each of your colors rather than just one.
The result of this is that winning decks will tend to have more cards that are powerful in a broader sense, namely card draw and ramp, rather than just powerful due to the synergy with their own strategy.
To look at this through a different lens, let’s revisit cards remaining in deck.
two color decks represent consistency
Two colors decks stand out with how little difference there is in cards remaining in deck for winning and losing decks. As we look at the color identities of the two color decks, the top 4 combinations all include blue or green (with the top deck being blue and green), which means that the majority of decks have access to the most potent card draw and ramp effects. However as we have discussed before, having access to the card draw usually leads to a disparity for winning and losing decks because the winning deck will be the one using these effects the most. Another cause is that with only two colors to balance in terms of mana it is very unlikely to be unable to cast your spells, especially with green aiding in color fixing, so even the losing decks can cast the spells they need to stay in the game for longer. When we look a bit closer and control for no green decks that creates the greatest disparity in winning and losing decks, combining this with the knowledge that there is a great disparity in land count between winning and losing decks we can presume that losing decks were spending much more of their mana on card draw rather than ramp. Finally the final potential reason for the equality in cards remaining in deck is that two color decks tend to last longer in games, again likely due to consistent openings since there is a good variety of strategies. In contrast to the extreme consistency of two color decks mono and three color decks have wildly disparate cards remaining.
Mono color decks are plagued with inconsistent card draw, but it can be easily explained by color identity. Once we control for no blue and green we see that the values are much more clustered together. This is easily explained by the fact that the best ways to pull cards from your deck are in blue and green. Three color on the other hand could be explained by the additional color creating inconsistency in the manabase. When we bring attention to the much higher land count in winning decks we can see that for most decks ramp was seen as more important than card draw.
This is just the tip of the iceberg
This analysis on deck colors, composition, and winning is just a single lense to analyze this data set. The most valuable aspect of this project is that you can glean small optimizations about decks and how you can use the insights to make the deck you want. Whether it is to make something unique, something powerful, or learning deck building best practices this dashboard provides the visualizations to easily and effectively gain insights.