In Development – What’s Your Read?

In the fifth round of last month’s PTQ, we were in the middle of resolving mulligans. By “we” I really mean “my opponent,” as I’d reviewed my opening seven, decided it seemed sound, and placed my hand face down in front of me with a concise, “I’m good.”

My opponent checked his seven…and stopped to think. He stared. He put the cards down. He picked them up again, screwed up his face a bit in concentration, and then finally decided to mulligan. As he was shuffling up for his six, he asked me, “Do you think it’s right to mulligan a hand with five lands and two spells?”

You might call that a “tell.”

I’m not especially good at reading my opponent, and as a consequence don’t spend a lot of time looking for tells (real ones, I mean). Instead, I tend to play as if my opponent were some sort of Bayesian average of all opponents in similar situations.

It is not, as we say, one of my strengths. However, as we’ve discussed before, if it is one of your strengths, wouldn’t you like to learn that it is and spend more time cultivating it? At the same time, if it turns out you’re not much good at it, wouldn’t you like to plug any critical gaps and then move on to something you are good at?

Yes, today is a Magic Effectiveness Project (MEP) piece.

Today I’m going to discuss “reading” your opponents. There is no one skill of “reading” – and, in fact, it probably breaks down into at least two distinct major skill sets, where we may be good at one and not so great at another. Using information gleaned from the first big MEP survey and examples from high-level tournament play, I’m going to talk about these flavors of reading.

Your “job,” such as it is, is to read all about ‘reads’ with an eye toward understanding whether one of those flavors is one of your strengths, and an area where you can grow your game by leaps and bounds instead of by inches.

A reminder about strengths

One of the essential ideas that I’ve acquired from the nice people at Gallup is that we tend to have trouble recognizing our strengths. The reason is just that they are something we have an affinity for, something that comes so naturally to us that we tend to assume it’s just easy. I have a coworker who thinks linear algebra is intuitive, but is consistently perplexed by biology.

…and if both of those seem tricky to you, that doesn’t make you dumb, it just means that your cognitive strengths are far more suited to other tasks altogether. I have a very good friend who can’t find his way through either science or math, but can bust out Middle English and other arcane languages like nobody’s business. In fact, he’s so good at it that he has a book about it.

So a strength is something for which you have both aptitude (you’re good at it) and affinity (you like to do it). A Gallup-style strength isn’t one of these skills I just mentioned, but something higher-level like “Context.” As we go along through the MEP results, I think we’re starting to see something similar at work…but more on that later.

Since we are talking about strengths today, I’m not going to spend a lot of time lingering on how we can avoid being read – that is, getting rid of our tells. If we were to talk about a strength that really lends itself to not being read, we might want to focus on stoicism.

But that’s a topic for another day. In the meantime, just pay attention to how the “victims” of these reads are giving up information, and keep it in mind as you play.

People watching

Perhaps it sounds silly to name one of the categories of reads “people watching.” Aren’t all reads fundamentally about watching your opponent?

Well, sort of. As you’ll see, some are about watching the other player and tracking what they do. Others are about engaging interactively with your opponent to acquire information from them. Finally, there is a line of reading that has very little to do with your people watching skills.

This category – people watching – is about how we pay attention to our opponents and casually pick up cues to hidden information, such as their cards in hand or their intent. In the world of the imagination, picking up tells via watching works like Cal Lightman on Lie to Me. The person does their thing, then you tilt your head to one side and loudly declare that they totally have a Lightning Bolt in hand and they’re planning on bluffing countermagic if you look like you might cast something.

The reality is less magical and more practical, and it breaks down into two parts – mechanics and emotions.

Reading mechanics

Here’s a snippet from the glacial Karsten-Leong match in the quarterfinals of Worlds 2005:

So if you were ever wondering where that godawful flick-flick-flick continuous shuffling habit came from, there you go. My preference is an obscuring shuffle followed by putting my hand down until I need the cards.

Many of the MEP responders who say they read players talked about tracking a player’s mechanics – that is, the actual motions they make when playing the game. This includes things like sorting the cards in your hand, or immediately playing a card that you just drew off the top of your library without any intermediate shuffling to hide the fact that you just drew it.

If you were Ding Leong’s opponent in that quarterfinals match, and it was your inclination to watch his mechanics, you’d quickly notice that he was flicking some category of card to the right of his hand. You could then watch him play a land from that part of his hand, and soon enough you’re able to track how many land versus nonland cards he has in hand.

And – this is the important distinction here – this happens all without having to develop any kind of feel for the specific details of the person you’re playing. You don’t have to watch their eyes, or see if they react to their card draw. You just need to track what they do with their cards.

The other major targets for the “mechanically inclined” reader are pauses and mana. A big pause at end of turn or after drawing a card gives the reader information about what cards the opponent has in hand…and the lands that opponent leaves untapped also spell out the opponent’s available options. There can also be “special case” mechanical reads, such as catching someone reading one of their own split cards during Ravnica and Time Spiral blocks.

If the idea of knowing where the lands are in someone’s hand strikes you as a no-brainer, you may have a strength in this area of observing and recalling your opponent’s actions. You can try to cultivate and enhance that skill by picking new traits to try to pick up on. For example, can you figure out the CMCs in your opponent’s hand based on how they’re sorting their cards?

Reading emotions

Consider this moment from the quarterfinals of Pro Tour Valencia 2007:

If you were just playing the “what do I lose to?” game here, you’d name Enduring Ideal, the win card in Andre Mueller’s combo deck. Both Randy and BDM agree on Enduring Ideal as the correct card to “name blind” on that Cabal Therapy.

So why did Sam Stein name Solitary Confinement?

There’s some logic to the idea that Sam wanted to name a card that might trip up his game plan, but if so, he still has a lot of options. Solitary Confinement, Pernicious Deed, and Form of the Dragon were all in Andre’s deck and all represented cards that would put a massive damper on Sam’s aggro plan.

The cue was something that, unfortunately, you can’t really hear in the coverage. As Andre plays out his turn two, he asks Sam, “Can you kill me next turn?”

For many of us, this would be random noise, but for Sam, cued to pick up on all sorts of tells, it carried a lot of information. It meant that Andre was deciding between the play he eventually made – casting an accelerator – and a play that would make it relevant whether or not Sam had a kill the following turn. Translated, that can also be said as, “Whether or not Andre needed to spend this turn buying himself another turn.” Seen that way, there was only one card that absolutely bought Andre another turn.

Solitary Confinement.

So that’s what Sam named, and suddenly Andre found himself with one less fewer to work with and a giant horde of angry robots cruising across the board to kill him.

This side of the read is for the person who is, in the classical sense, a good listener. If you, like me, are inclined to treat comments like, “Can you kill me next turn?” as conversation meant to lighten the mood, then this isn’t one of your strengths. On the other hand, if you’re constantly trying to evaluate what someone meant by what they said, then you might want to spend time applying that skill to linking people’s comments and conversation during play with what they have in their hand, and what their game plan is.

There are a lot of opportunities to exercise this skill, and many of them come up outside the strict bounds of the individual games themselves.

Imagine you’ve just been seated for round three at a PTQ. You and your opponent are both 2-1. Your opponent complains about their loss last round to Faeries, and what a pain being timewalked by [card]Mistbind Clique[/card] is.

What deck might they be playing?

What if they complain about how Faeries is stupid because it’s “all about getting [card]Bitterblossom[/card] down.” What are they playing then?

In the same vein, if your inclination is to listen and evaluate like this, pay close attention to how people refer to the hands they’re disposing of when they mulligan. Do they just flat-out tell you how many lands it had? Do they talk about casting costs? What can you learn from that?

Note that this is all about active listening, but not about engaging with your opponent (which is, ironically, what Andre Mueller was probably trying to do when he accidentally gave up the goods in that video). If you’re the type of person who likes to chat ‘em up, that’s…

Working the crowd

Check out this excerpt from the quarterfinals match between Brian Kibler and Evangelos Papatsarouchas at Pro Tour Austin 2009:

Brian Kibler is an animated guy. He makes big, expansive gestures (seriously, go back and watch the opener to that video, as the camera pans down toward him). He chats with his opponent through the entire match, discussing their experiences at the Pro Tour, other games, how things might have gone differently in that last game.

Honestly, it’s fun to play with someone this good natured and lively. It makes for a good match.

But amidst all of the banter, two interesting exchanges happen.

In the first excerpt, Brian asks Evangelos if he has anything good coming off of the second Hypergenesis, and Van lets him know there’s an Angel of Despair coming.

In the second excerpt, a similar question from Brian gets Evangelos to let him know he has a Progenitus in hand.

Say what now? Van just told him what he had in hand? Twice? When it might conceivably impact which card Brian chose to play first off of Van’s Hypergenesis?

This is what we might think of as the “proactive” version of the emotional read. Instead of catching the tidbits of material that your opponent lets slide as they speak, you create a bridge between the two of you. It’s a lot like being the “good cop” in a stereotypical police negotiation. You create rapport, get the other person comfortable with the back and forth, and then should you happen to ask them something they perhaps shouldn’t answer, they might just answer anyway.

Lest this sound like a pure tactic, keep in mind that the only reason Brian can do it so easily is that he’s genuinely interested in interacting with the person across the table from him. As Keith Ferrazzi tells us, successful contact building requires an actual interest in the other person.

This is why I’m pretty chatty during mulligans. It’s a chance to reconnect with people I know and get to know more about new opponents. But I get quiet during games because I’m not as much of a natural connector as Kibler.

Once again, this is going to be a difficult trait to force…but if you find that you already tend to be chatty throughout the match, and you’re a decent enough listener to create a genuine back-and-forth with your opponent, you may want to try throwing in the occasional question aimed at finding out something helpful about your opponent’s cards or game plan, just to see what you can harvest.

Of the various reads described so far, this one is probably the hardest to see coming and block out, at least for most people – so if you already have some affinity for this approach, it’s worth trying to cultivate it.

The procedural read

So far, the reads that we’ve covered feel like what we mean when we talk about “reading” and “tells.” You’re watching your opponent or chatting them up, picking up on their physical and emotional cues.

However, a lot of players don’t want to have anything to do with this kind of read. As one MEP responder said, “Reading people is very difficult and requires a lot of practice, so I don’t listen to most people when they say they’re really good at it.” That’s my bias, too – I bet a lot of us aren’t nearly as good at these reads as we think we are. And, per the Gallup idea, those of us who are good at it don’t think we are.

But let’s set this all aside to look at an entirely different kind of read, one that can feel like it’s based on the person. That’s the “procedural” read, and it comes in two flavors – logical and historical.

The logical read

Let’s return to that quarterfinal match from Worlds 2005:

Now, that feels like an excerpt from Lie to Me. Frank’s right there, the protagonist in our Magic-themed police procedural, tilting his head sideways, staring at Ding Leong, and just calling out the cards in Ding’s hand.

The thing is, Karsten isn’t paying attention to how Ding’s flicking his cards or what Ding is saying. Frank isn’t really chatting him up, either.

Instead, Karsten is running a logical evaluation of what cards Ding is likely to have in his hand. He’s seen a couple turns without any action from Ding, and, if Frank assumes that Ding is a reasonably rational opponent – a fair assumption in the top eight of Worlds – then he can start to evaluate which cards in Ding’s deck wouldn’t be worth playing given the board state.

Now, Frank’s excellent party trick of naming the exact cards in Leong’s hand is made possible by dint of his having full access to Ding’s deck list, but the principle holds even in normal, non-top-eight play. Consider pretty much any tournament report by our own PV, or any of the videos by LSV. They walk through the possibilities in a level of detail that might make you believe it was added after the fact – except that the videos clearly put the lie to that idea.

Doing this kind of read “in the wild” requires a fairly thorough, or perhaps obsessive, understanding of a format the likely cards in a given archetype. It also requires the ability to mentally model the contents of your opponent’s hand as you plan your own plays, and this doesn’t just mean “have a general idea of what they have” but “be able to construct and maintain a specific hand in your mind.”

If you’re already inclined to make educated guesses about what cards your opponents have, and can usually “unravel” those guess to explain why you made them, you may be on track to cultivating this style of read into a robust strength – and you’ll have an excellent party trick the next time you top eight an event.

Historical

We’ll end on this clip of future Hall of Famer Gab Nassif and future Pro Tour Paris winner Ben Stark at Pro Tour Kobe in 2004:

Those are games two and five, respectively.

Much like the logical read, the historical read is something that you could easily do on MTGO or any other online game, even without access to their physical tells, the things they say, or the ability to talk with them.

In AI, there’s this thing called a Hidden Markov Model (HMM). It’s not worth going into detail here, but HMMs are basically computational tools that try to figure out “what happens next?” for situations where the underlying rules are hidden (thus the “Hidden” in HMM).

Lest this seem tremendously esoteric, the point is that this is what you’re doing if you’re running a historical read on your opponent. This is what Nassif was doing in his fifth game against Stark. Knowing how Stark got locked out of game two and having watched him play out the games in between, Nassif formed a relatively solid evidence base – a set of rules – to help him figure out “what happens next?” when Stark started off with such a seemingly mediocre first turn in game five.

How quickly do you figure out what your opponent’s next play is going to be? Not just what a generally likely next play is, or even what cards must be in their hand, but actually how your specific opponent is going to respond to the board state and whatever you just did? As I mentioned above, I’m not particularly good at this – it’s not my strength, and I treat each opponent as a sort of aggregate of all likely opponents. But if you find that you rapidly form an idea of what your current opponent will do, then you may want to push to develop this strength.

A recap, and a suite of skills

The simple take-home from today’s piece is that there is no one “skill of reading your opponent.” It’s a suite of distinct abilities, many of which have nothing to do with each other. The upshot of all this is that you may find you’re good in one aspect of reading, even if you aren’t particularly suited to another.

After all, there’s no shame in being either a Karsten or a Stein instead of being both at once.

***
magic (at) alexandershearer.com
parakkum on twitter

Share this

Discussion

Scroll to Top