Sometimes, you find yourself wondering if your crazy brew is competitive.

Okay, it’s possible that my crazy brews aren’t quite as crazy as some of the Saffron Olive or CovertGoBlue brews that are on their face non-competitive decks. It’s also possible that I am trying to cast Quintorius and storm off while not dying the Heartfire Hero by turn 5. Regardless, I find myself awake in EST time, in the middle of standard testing season preparing for the next Pro Tour. One advantage of having a global testing team is that we all join together on Discord, it unifies us. On the other hand, surprisingly often the Eastern time late evening slot is a no go for people wanting to play. Which is why I, somewhat belatedly, came to the conclusion of loading up two decks on Cockatrice and playing them against myself.

At first, I didn’t realize this was possible without having two instances of Cockatrice open, each one connected to one of the servers, one hosting the match and the other joining. Once done, you can set them up split screen on a monitor and alt-tab back and forth between the two. The trick is that the window on the left you have to go into the View menu and disable floating card images, log, and text to clear the screen space. Belatedly, I learned that you can also just Host Local Game and select 2 players as the option. However, while less clunky to set up, I prefer the double client implementation as it allows you to use hotkeys for each player without requiring moving through turn phases to the other deck’s turn. For instance, if I wanted to use Ctrl-W to scry cards on my opponents end step with a Consider, the UI would think the active player has the hotkeys and would scry for the wrong deck. This issue is remedied by Alt-Tabbing back and forth, but requires more familiarity.

Solo Magic

Beyond the technical difficulties, playing the game against oneself is a very compartmentalized process. While at the beginning of a match, the deck that goes first chooses whether to keep or mulligan. Then priority switches to deck 2 (like a normal game of Magic). If deck 1 kept, this is where the first example of doublethink comes in. You must forget deck one’s hand; try to remove the bias of overcorrecting for the cards present in the hand that is visible to you in the opposing deck. It can be hard to measure how successful I am in doing this. In fact, oftentimes knowing about a bias is (citation needed) not even helpful in avoiding it. One strategy is to look at the full opposing decklist while mulliganing, though this can slow things down. I can rationalize that my endless hours of experience mulliganing might help build up an immunity, and ultimately, the effect on keeps is usually small and matchup dependent. I felt this pull most with cards like spell pierce.

As an aside, playing around one-of copies of cards follows a similar logic, with the caveat that you have to try and calculate whether playing around them will maximize your winrate*. The gameplay I found dynamic. You get the full conception of the graph of decisions that could be made throughout a game. Sometimes in a match, you play your 1 drop 2 drop and 3 drop on curve, and the other player has to decide to play their 2 drop or play their removal spell. You will always be the one player having to decide. Making twice as many decisions can feel taxing, but it allows for play with utmost intentionality. Playing with a purpose to your actions is a great quality to have, especially in a training game. I felt like I better understood the reason going for a tempo gameplan might be better than a value gameplan in a matchup. This a skill I have tried to internalize more, especially since Worlds 2024.

At worlds, I played UW oculus, alongside Eli Kassis. Eli has a gift (cough, 30+ years of experience) for deck building. He often has an easier time than myself understanding what a deck should be trying to do. At this worlds, we were discussing deck building choices and Eli spoke about deck velocity. UR phoenix in pioneer is a premier example of a high deck with high velocity. Individual card quality can sometimes suffer, as the rewards of seeing your key pieces more consistently outweigh the air that having a bunch of cards that mill or draw cards in your deck can present. I wanted to go bigger, Eli wanted to go smaller. This analogy was quickly cemented when I began building a limited deck (Duskmourn) that was focused on reanimator. Again, the strongest move the dec kcould make was reanimating the Valgavoth I had taken in P1P1. I had built a version that was more value heavy, UW with black splash. Eli rebuilt my pool into a BW splash g for 2 broodspinners deck that focused on churn and setting up one big play. All this to say, in the Cockatrice games against myself, I could feel the pressure of the deck needing somethign to build towards when in the driver’s seat much more clearly than an opponent who held 7 cards in their hand, face down.

Cockatrice matches also allow for replayability from specific spots, and for putting cards directly in hand to test them in situations quickly and efficiently. This is why Cocktrice is especially useful for testing new cards. If you think a card should only be a 1 or 2 of in a deck, playing a league or queueing up on Arena will lead to few games with that card in hand. However, 4 copies may lead to games with multiple copies. Simply putting the card in your starting hand on Cockatrice allows us to rapidly prototype decks including the new card. Returning to the concept of a graph or tree of decisions, Cocktraice allows you to set up the same situation and play it out in various ways, testing strategies and counterstrategies that might have led to different outcomes. This is an underutilized technique in Magic practice, but is extremely common in games like Chess. Part of the reason for this is the software we have available. Additionally, each card you draw adds another level of variability, and it’s dubious that replaying games with the same order of cards drawn is more useful than full randomization each playthrough, to get more of a sense for the multiverse of possible games playable.

Sometimes playing against yourself will reinforce bad habits, make you overthink situations, or lead you to improve slower than against an opponent who can quickly point out decisions you miss or new opportunities to improve your game. I would recommend recording your playtesting session, and having someone review it (though finding this person may be tricky, perhaps a swap? Also requires that you somewhat stay on track in the session instead of going down confusing rabbitholes, or voice your rabbitholes out loud).Self-practice has unique value that can be combined with normal two player practice for maximum effect.

Other extraneous notes on testing process:

A feel for a deck that works can be more valuable than a small quantity of results that suggest otherwise. We are constantly extrapolating from small samples (quote here). Brewing decks or trying to understand new cards can be difficult to get an accurate read. Sometimes, I get a feel for a card just having it in my hand that will let me realize “huh, this card feels like I always want to draw it” or to a lesser extent a feel for situations when it is good “this card feels great in my opening hand, but bad when I draw it after turn 2”. Bootstrapping these initial impressions can be one way to reduce the bias inherent with one’s first time playing a card or deck. For draft specific analysis, tools like 17 lands opening hand win rate can allow one to quantify these feelings. A 17lands for constructed could still be useful, though the overall trends would have a lot more variability than your typical limited set. For example, the existence of dedicated combo decks throws a lot of draft wisdom out the window in deck building and play patterns.

*EDIT THIS: If I am in a dominating spot in a game that I am 95% to win if my spell isn’t spell pierced, and 80% to win if it does get spell pierces, the roughly 25% percent of the time (This logic is comparing wrong - we should be comparing them having spell pierce to them not having spell pierce and usplaying around vs us not - 4 scenarios, add up the ev of both playing around and not playing around)