It’s remarkable to think how far both the game of Magic and the culture around Magic have grown recently. What was rare and remarkable yesterday is common today, and old technology and edges are just part of the game. Likewise, old lessons have become integral to how we understand the game, but you’d be hard-pressed to get a newer player to explain why. What I’m getting at is that years ago it was rare to get in-game commentary from players as they played the game. The rise of streaming and video recording software has made it commonplace. Which is why when I tell newer players that Richard Feldman’s “One Game” is one of the most influential and important theory pieces in the history of the game, they usually respond with confusion.
After all, streamers commentate on their games all the time, and that’s all Feldman is doing in the article. While true, that reading does discount all the little things that he does during the match like considering his body language and how he personally interacts with his opponent, which aren’t possible on MTGO. It was the discussion of them and how to find the tiny, almost imperceptible edges that made the game seem closer than it actually was that made the article great. Despite the praise it received at the time and the continuing respect many theorists have for the article, you never see articles like “One Game” very often.
You’re probably expecting me to launch into how I’m going to follow Feldman’s formula now. I am going to do that, but it’s been nine years and I could have done considerably less work and just recorded a game for you and accomplished the same thing. So I’m adding retrospective commentary. I’ll guide you through how I approached the game in question and why I played the way I did and go through and analyze my decision-making in light of what I knew after the fact. As a side note, I do encourage this kind of note taking and postmortem introspection for those trying to improve their game. Figuring out why you did what you did at the time and whether it was actually correct improves your decision-making and helps you to find the correct lines more often. Also, if you haven’t read the linked article, do so now. His reasoning about why this is valuable is still very relevant, so much so that I’m just going to directly quote him:
There are all sorts of little edges you can gain in Magic beyond the confines of the tabletop. Mike Turian apparently once said that you could disqualify yourself from having played a “perfect” game of Magic by so much as blinking at the wrong time; like it or not the way you act as you play influences your ability to succeed. If you’re bad at it then you have “tells” that give the opponent bonus information about what’s in your hand, what you’re planning, and what you do and do not realize about the game state.
Today I’m going to walk you through one game. One game at a theoretical PTQ where I pause at every step of the way and think about all the different ways I might accidentally give myself away, throw the opponent off, or figure out what he’s up to based on the way we are acting. I will doubtless miss a good many of these interactions but hope to illuminate quite a few for those who are not in the habit of thinking about them. After all there’s a lot more to a game than the cardboard interactions.
Note on Methodology
This game was recorded during a testing session about a week and a half ago. I’ve been intending to do this piece for some time and had been taking notes for that purpose, looking for a good example game.
I am playing my Merfolk deck against a player who I know to be very goo— think RPTQ finalist-level good. I do not know what my opponent is playing, just that he’s got a new deck that he wants to try against me. I am on the draw. I recorded every play I made and why I made them. I will be clearly marking my retrospective comments. My opponent’s actual name will not be mentioned, but it’s rather awkward and impersonal to just use “the opponent” constantly, so I will substitute their actual name with one that nobody I know uses: Elliot.
As we sit down and shuffle up I take a moment to look at my opponent’s setup and particularly their sleeves. It’s possible to gain information and misinformation about a player based on the condition and brand of their sleeves. A player with high-quality sleeves that shows signs of wear is likely to be a very experienced player who practices a lot. This can be misleading, especially in later rounds of a tournament, so if I want to gauge my opponent I prefer to use their demeanor and obvious preparedness as a guide. Newer players tend to look somewhat uncomfortable and fumble with deckboxes and dice prior to a game while experienced players tend to have a system in place and quickly get settled. This is usually a clue as to how effective your bluffs and misinformation is likely to be in a given game.
None of this applies today because I know exactly who I’m playing and their skill level, but you should practice like you compete and since I do this at Competitive REL events, I do it for test games. Elliot has also just finished putting new sleeves on his deck so there’s really no information to gain. We sit down and start shuffling. I only do Rock, Paper, Scissors for the play since I caught someone using loaded dice, but since we’re testing Elliot’s deck they get to go first.
Since we practice like we compete we present and shuffle each other’s decks. I always shuffle my opponent’s deck, even at casual level, and I make sure to look away from the opponent’s cards but watch how my own are being shuffled. I’m paranoid about tricky shuffles and deck stacking. I may know that nothing will happen today, but it’s a good habit to be in. That complete, we each draw our starting seven.
Elliot takes a hard look at their hand and thinks for several seconds before taking a mulligan. My hand is a solid keep with Aether Vial, Island, two Wanderwine Hub, Silvergill Adept, Lord of Atlantis, and Master of Waves. I wait a few seconds after Elliot’s announcement to make mine. There is value to be found in misrepresenting the strength of your hand during the mulligan process, but not every opponent will fall for it when you take the first mulligan or make a quick keep.
One problem is that players will often remember to hide strength or weakness on the initial seven, but drop the act afterwards. I will often act weak or dispirited when I go below six so that my opponent will get some confirmation bias (when you mulligan a lot, the opponent expects that you’ll keep marginal hands) and will try to play accordingly, but don’t always succeed. This is why I normally try to be as neutral as possible. I always take about five seconds on the play to make my mulligan announcement and wait a few seconds after my opponent makes theirs on the draw to hide as much as I can about my hand’s actual strength. My hand is good, but there’s no reason to deviate from my normal procedure.
Elliot keeps six and scrys to the bottom after a few seconds of thought.
This tells me almost nothing about what I’m actually up against, as many decks will make that play, but it does eliminate a number of other decks. I’m certainly not playing a control deck or any deck that runs counterspells, so I should have no problem resolving a spell turn two. Which will depend on what I draw and what Elliot does on turn two.
I draw a Tectonic Edge, shuffle it into my hand and think. I will be playing Vial this turn, the question is which land. The Lord of Atlantis needs double blue, and it’s possible that Elliot could Qasali Pridemage and kill my Vial next turn so Tec Edge is out. One thing to consider is that your opponent will not know the exact content of your deck at a tournament but will be familiar with the general archetype. Therefore if you have an atypical list you should hide that as long as possible and let your opponent plan against the wrong deck. For this reason given the choice I will play Wanderwine Hub before Seachrome Coast and Island before both. This makes my play simple.
Play Island and Aether Vial, say go.
Once I say go I lay my hand on the table. I do this whether or not I have mana open or could make a play on my opponent’s turn. Part of this is to keep me from fidgeting but mostly it keeps me from looking at specific cards which might give something away. Experienced players will watch how you react and can sometimes be bluffed, but I’ve found that giving them as little information as possible is also effective. If you’re as unreadable as possible then the opponent has more opportunity to make false reads or assumptions and play into your hands. Letting someone’s own imagination work on their reason is just as potent as deceiving their reason, so I try to minimize my tells. It also lets me more closely observe my opponent, and I know the cards in my hand so I don’t need to constantly watch them.
Me: 20 life, six cards in hand
Elliot: 17 life, four cards in hand
While I still have no idea what deck I’m against, I now know quite a bit about Elliot’s hand. It has nothing proactive that costs three mana and there are no reactive spells that cost one or two that Elliot is willing to play next turn. The fact that Birds attacked for one indicates a desire to race, which points to something aggressive. Given this information, Elliot is playing a threat-heavy deck but has threats that cost four or more in hand. This could be a number of decks but given what I’ve seen I’m leaning towards a Chord of Calling deck. Those need a lot of mana birds, don’t play much interaction and have many plays that need a lot of mana. I put Elliot on having multiple costly creatures and/or Chord in hand. I need to find some interaction and get a board together in a hurry.
I charge Vial and draw another Adept.
While I could play Lord here, there’s no reason to expose it to a Path to Exile or Fiend Hunter yet, so I’ll be playing Silvergill. Given that I need to reveal Merfolk anyway, I should play Hub as well.
Play Hub revealing Silvergill, cast Silvergill revealing Lord, drawing Path.
In retrospect: Revealing Lord was wrong, I should have revealed the other Adept. I gave my opponent far more useful information than I needed to about how I could attack next turn. Revealing Adept is comparatively irrelevant.
Me: 19 life, five cards in hand
Elliot: 17 life, three cards in hand
Elliot draws and slams the freshly drawn Brushland down, which allows Thought-Knot Seer to come down. Guess my read was off; this is Bant Eldrazi. Once I’m actually targeted by the ability I flip my hand over, revealing Lord, Master, Hub, Adept, and Path. Elliot thinks for a while before taking the Path. It’s hard to say whether this was a case of actually being unsure of what to take or if Elliot was deciding whether to play the control deck and remove a creature, or protect their only threat from my removal. BoP then flies over for another point.
At this point my guess is that Elliot’s hand is worse than I thought and the scry was looking for a colorless source. I guess that there are few real threats remaining in Elliot’s hand and that the plan is to race me. It’s the only way tapping out makes sense here.
I charge Vial and draw another Adept.
I will definitely play Lord and attack with my islandwalking Adept this turn, so I will cast the Lord normally. Elliot knows about Master of Waves so I might as well play my Hub untapped by revealing it.
Cast Lord of Atlantis, play Hub untapped revealing Master, attack for three.
In retrospect: Why didn’t I play Hub by revealing Lord? There was no reason not to. By not doing that I reveal that I was thinking of playing it tapped which signals that I don’t have another Path in hand, and thus makes playing the Hub untapped at all irrelevant.
Me: 18 life, three cards in hand
Elliot: 14 life, two cards in hand
Elliot draws and then plays Ancient Stirrings, which was obviously the draw and there’s really no point in trying to hide it. Sometimes you can hide your draw and sometimes you can’t and there’s no reason to try. Stirrings finds Matter Reshaper, which Elliot plays along with another Noble off the first Hierarch. This leaves just the other Bird untapped when Seer attacks for five. Elliot may be representing Path here, but the sequencing is odd given that they didn’t leave up mana for it last turn.
At the end of Elliot’s turn I activate Vial, play Adept and draw another Vial.
I don’t charge Vial and draw Mutavault.
I now have the mana to play Master of Waves, and if I’m going to do that I want to maximize my token generation, so I need to Vial in Adept. Playing another Vial at this point is worthless. I cannot attack with Mutavault and Eldrazi does not maindeck sweepers, so I’ll definitely play the Master here.
Activate Vial, play Silvergill, draw Lord of Atlantis.
Now I actually have something to think about. I could just cast the Lord, which lets me attack for four through the Reshaper. I wouldn’t attack with both eligible creatures because the threat of Path turning into a blowout is too great. On the other hand, if I’m worried about Path I should still just cast Master precombat. This forces the Path since there is no reason to Path afterwards and give me extra tokens. If Elliot doesn’t Path in response to Master I know it isn’t the last card in their hand. If they do then I don’t run Silvergill into Reshaper and have a follow-up Lord.
Cast Master of Waves, in response Elliot plays Path on Lord. I fetch an Island and Master resolves, giving me four elemental tokens. I pass without attacking.
I don’t want to attack and give my opponent a free draw off Reshaper. Sometimes it’s right to do so, but in this case Elliot’s life total is too high.
Me: 13 life, one card in hand
Elliot: 14 life, no cards in hand
Elliot draws, thinks for a couple of seconds without looking at the card, and passes.
The card in Elliot’s hand can only be a land or another Path. There is no reason to play a land and remove all doubt from my mind at this stage of the game. Elliot doesn’t know about the Lord in my hand, but does know that I could draw another one and win the race with Adepts and Mutavault, so it makes sense not to Path here. If it was another creature I expect that it would have gotten played as another blocker for tokens.
I don’t know the exact composition of Elliot’s deck, but I do know that Bant Eldrazi decks run 24-25 lands and only rarely more than four Paths for removal. There are 49 cards left in Elliot’s deck, three lands on the board, and one Path already used. This is odds of 43%-45% of a draw being a land vs. 6% for removal. I decide Elliot has drawn a land.
I don’t charge Vial and draw Cursecatcher.
There’s no reason to play the Cursecatcher pre-combat, so I won’t be doing that. I think for quite a while before making my attack. With the Lord in my hand my islandwalkers are lethal. If I’m wrong about Elliot’s hand then attacking only with Merfolk gets me blown out. I don’t think Elliot has removal, but I decide to play as though I’m wrong, which means maximizing Elliot’s need to block and attack with the tokens as well. At worst this will put Elliot to four when the unblocked tokens get through, with only Thought-Knot and potentially an unknown card on the battlefield while I’ll have some number of Adepts left.
I play Tec Edge, cast Lord, activate Mutavault, and attack with everything. Elliot scoops.
It was a land, so I won the game.
In retrospect: I think I played that last turn wrong. I should have used Vial to play Lord. When I won the game I left a Hub and the Tec Edge untapped, thinking that I could always try to attack his lands next turn. However, this cuts me off from playing Cursecatcher. I had nothing else to do with Vial, so I should have maximized my untapped lands post-combat.
An Alternate Story
To blatantly rip off Feldman, I drew well and Elliot didn’t.
Magic is so simple!
David began playing Magic during Odyssey block, quit playing Magic when Caw Blade ruled the world, and returned to Modern shortly before Deathrite was banned. He’s made an appearance at the Pro Tour, made money at GP Denver, and is constantly grinding and brewing in Modern.