PV’s Playhouse – Playing On Autopilot

StarCityGames.com 10K Open - Los Angeles

Thursday, December 24th – Last week, Paulo warned us about the dangers of overthinking. Today, he spins through 180 degrees and preaches against playing on autopilot. As with much in life, the key is to hitting that sweet spot, walking the fine line between each extreme.


Competitive Magic is, in general, a game of small edges. If you assume both players are competent and both decks decent, there is a high chance that the outcome of the match will be decided by the person who played slightly better than the other person on many different occasions. You can only play slightly better, though, if you recognize you have a choice of plays. Take, for example, this situation:

You are playing Yuuya Watanabe’s Dredge list:

2 Breeding Pool
2 Island
4 Misty Rainforest
4 Scalding Tarn
1 Steam Vents
3 Verdant Catacombs
3 Watery Grave

3 Bloodghast
4 Drowned Rusalka
1 Flame-Kin Zealot
4 Golgari Grave-Troll
4 Hedron Crab
1 Iona, Shield of Emeria
4 Narcomoeba
1 Sphinx of Lost Truths
4 Stinkweed Imp
4 Bridge from Below
3 Dread Return
4 Glimpse the Unthinkable
3 Ideas Unbound
1 Life from the Loam

You have a Hedron Crab in play and a Misty Rainforest in hand, and your graveyard is empty – you have only those two cards to work with. Your opponent is playing a Mono Blue Combo deck, and he will kill you next turn for sure, unless you happen to kill him this turn or reanimate Iona on Blue, which also wins the game for you. He is at one life (don‘t ask how… I write the article, I make the rules!), and you cannot mill him out.

Now, contrary to what some people might think, there are two possible plays here. It might be that you don’t even notice you have options – after all, you only really have one card to play – but you do have two choices, and one of them is strictly better than the other. Sure, you still need to get lucky to win – no one said you were looking good this game – but with one play you need to get less lucky, so it is the strictly superior play, and if you do anything different than that you are not giving yourself the best chances, and therefore making a mistake. So, what is your play?

Now, imagine that you are playing the finals of your PTQ – everyone is watching. You are considered the future of local Magic, everyone thinks you are good and deserving of a chance to go prove yourself in a Pro Tour, and so do you. It comes to the final turn – you know you need to get lucky, but what can you do? It’s your only chance. You play your Misty Rainforest and target yourself with the Crab ability. However, before the ability resolves, you decide to make the technically perfect play and sacrifice your land – if you are going to lose, at least everyone is going to see that you did everything in your power to win that game, but fate wasn‘t on your side. As you do that, you hear whispers in the background from people watching – they were right, you are the future of local Magic. None of them would have made this play – it is far beyond their skill level.

You take out an Island from your deck and target yourself again, content that you now have one less land in your deck, so the probability of milling an important card is slightly higher. You start milling your first three… they are Stinkweed ImpSwamp… and Bloodghast. Bloodghast! And your opponent is at one! Now all you need is to sacrifice that Fetchland and kill him! You’re going to your first Pro Tour! Except… you’ve already sacrificed your Fetchland. You then proceed to mill the next three cards, which are all irrelevant, and lose next turn. I guess the Pro Tour will have to wait a little longer…

Moral of the story: Don’t play on autopilot.

In my last article, I talked about how overthinking simple situations could lead to disastrous plays. Today, I’m going to talk about what is pretty much the opposite – playing on autopilot (and no, not small edges – maybe on another opportunity). Playing on autopilot is defined by not thinking through what you are doing, and instead just following your usual procedure, when the situation is different and warranted new thinking. In the above example, the player did not think through this specific situation; he did the thinking once and concluded that “sacrificing with the ability on the stack is better,” and in his brain that was automatically true for every time this decision came up.

“But PV,” you might ask, “in your last article, you actually told me to go with my instincts and not overthink situations, and now you tell me to stop and go against what I have pre-conceived, and not to play on Autopilot; do you want to drive me mad?”

The big point is that you must play instinctively when you are correct, and you must stop and think when you aren’t!

Of course, it is very hard to recognize when you should or when you should not think more – if it wasn’t, things such as overthinking and underthinking would not happen, and we would simply think as much as the situation demanded in all the cases. The best way to differentiate those situations is, to me, to know why you are doing the things you are doing. The moment you know why you play on autopilot, then you are able to recognize the situations where you shouldn’t.

With the Dredge example, you must know why you sacrifice that land with the Crab ability on the stack… you do it to take a land out of your deck and improve your chances to hit relevant cards. You don’t do it because your friend told you to, or because you watched Yuuya Watanabe do this in the PT top 8 videos. The reason you sometimes eschew a Bloodghast entering the battlefield in favor of that small percentile advantage is that, in the big picture, the two damage is very rarely going to matter in this deck. If you know that this is why you do it, and if you are consciously giving up the chance to get a 2/1 in play a turn earlier because winning the game is more important than dealing two damage, then once you are faced with the situation where dealing two damage actually wins the game, you will be able to recognize it and prevent playing on autopilot from destroying you. If you do it simply because it is better, there will be no differentiation from you between the cases it is actually better and the cases it isn’t.

Another example:

A few days ago, a friend of mine was playing in an online tournament. Then he pasted a link to the following image:


His opponent hasn’t taken the damage yet in this screenshot – he is going to drop to six this turn. My friend has a Nimana Sell-Sword, a Blood Seeker, and a Vampire Lacerator in his graveyard.

Then he asked a group of people which two creatures he should target with his Expedition – the Sell-Sword and the Lacerator, or the Sell-Sword and the Blood Seeker. People gave him different answers, and they all had good points – the Lacerator is better if the opponent has a removal spell that can target the Sell-Sword, since it kills him anyway (or forces him to stay home with the Bog Tatters, which works just as well), but the Blood Seeker is better if he has two creatures to block. If he only has one creature, they both are pretty much the same. Which one do you return?

In this case, the smart person will tell you that there is no way of knowing, because you did not play the entire game. Was there a situation where your opponent could have played a removal spell but didn’t? Can he have a Disfigure in hand to kill the Sell-Sword with the trigger on the stack? Did you pass five Journeys to Nowhere to your left? Or is it possible that your opponent has two creatures in hand and the mana to play them? Which is more likely? I can’t tell, and neither can you. However, the even smarter person would instead return both the Blood Seeker and the Lacerator, and play those two and the Puma. Once you think of it, this play is just so much better – it beats a removal spell like the Sell-Sword/Lacerator play, and it beats double blockers, like the Blood Seeker play. It even beats a removal spell AND a blocker (i.e. a Heartstabber Mosquito, which is not an unlikely card since your opponent has six lands out and three cards in hand). The only thing it does not beat that the other play does is Marsh Casualties, but that is overwhelmingly less likely than Disfigure, Mosquito, Journey, two blockers… you get the picture. Once you stop to think about it, it’s rather surprising that my friend even asked about it – there is really only one play you can make here, there are no decisions. It is even more surprising that no one suggested it – all the suggestions were Sell-Sword and Blood Seeker or Sell-Sword and Lacerator, never both.

The reason people mess up this play is because they play on autopilot – they have it in their minds that Sell-Sword is a better card, so it is your auto first target. However, once you stop to think, why is Sell-Sword better? It is better because it is 3/3, but in this situation, being 3/3 is completely irrelevant! If you consider your opponent being at two life (since he is taking four from the Skyfishers), then the 2/2 and the 3/3 body are just the same, and so are the 1/1 + 1 damage from Blood Seeker – there is no combat involved. All that it matters is that each of your cards deals two damage to the opponent, and that you have as many of them as possible so they are harder to deal with.

Sell-Sword might not be screaming “get me” enough, but imagine that you have a Vampire Nighthawk in there – I’m sure the overwhelming majority of people would just get the Nighthawk back, but in reality your decision does not change at all – you still lose to Mosquito, Disfigure, and Journey. You have to see past the fact that he is a much better card and he gains you life and he is a first pick and the best card in the set and blah blah… For this situation, the only thing that matters is that he is only one attacker. Think of it this way: if you are on five life and your opponent has a Felidar Sovereign and a Bramble Creeper, you must kill the Bramble Creeper, even if the Sovereign is an infinitely better card, because at this precise moment all that matters is the five power. The situation is pretty similar here, except much harder to recognize. Remember, being good is not an intrinsic quality of a card in a game of Magic. It is something that is earned by a game state or game strategy. The fact that some cards are good a lot more than others doesn’t mean it’s always so.

By the same token, there are some cards we are simply taught to fear, and sometimes we will go out of our way to deal with such cards just because they are “good cards,” and not because we actually have to do so. One very good example of this is LSV versus Nassif finals match in Kyoto. There was one game where LSV played Ajani Goldmane with a bunch of dudes out and kept Windbrisk Heights untapped, and Nassif had Negate. A lot of people would just Negate it; the Ajani was a threatening card at that point, and tapping two mana to deal with a four-mana threat is supposedly an advantage. Nassif, however, knew what really mattered: that he didn’t get Head Gamed. He reckoned that he could beat Ajani Goldmane with what he had access to, but he could not beat Head Games, and if LSV had a Plains in hand and a Head Games under that Windbrisk, then that would happen. He chose to let the Ajani resolve – contrary to the what I believe to be the autopilot play of just Negating it – and as a result won a game that, in the words of LSV himself, most people in the world would have lost (since he did have the Plains and the Head Games under the Windbrisk). Just like yours, your opponent’s cards are not intrinsically good or must-answers – before you think something is a must-handle threat, actually stop and ponder whether you really have to deal with it.

One other common occurrence of playing on autopilot is when an opportunity to have an advantageous play presents itself. Since we started to learn the game, people have told us that getting card advantage is a good thing, so it’s a very hard habit to break, but sometimes you simply don’t need card advantage. To avoid playing on autopilot in those situations, you must have the ability to recognize what matters in each match or game state – you cannot blindly go for card advantage just because most of the time card advantage is good, because sometimes it is not the most important thing.

Take the scenario in which you have a Baneslayer Angel in play, and Double Negative and two Lightning Bolts in hand. Your opponent then plays Bloodbraid Elf and cascades into Blightning. A lot of people would just instinctively play the Double Negative there – it is, after all, killing a creature, gaining three life and saving two Lightning Bolts. You could see it as a four-for-one. The autopilot play is to just counter it instantly – after all, this is the kind of situation you play Double Negative for. However, in a situation like this, most of the time those cards don’t matter at all, and you are much better saving your counter for something that deals with your Baneslayer Angel. Now, what if it is a Swerve you had in hand? The temptation to play it would be even greater, but it would still probably be correct to hold it to protect your Angel, as counter-intuitively as that might seen (though that depends on the situation – the key point here is that you don‘t HAVE to play Swerve, and that it doesn‘t necessarily say “change the target of target Blightning”).

Another situation we might play on autopilot is when we play decks that have one clear plan, and then sometimes we get too focused on that plan and we ignore other ways we can win the game. You can take, for example, my Scapeshift opponent at Worlds – he played his deck for numerous rounds (at least), and in all of them his plan was so play Scapeshift and win. In our last game, I had a Meddling Mage on Scapeshift so he couldn’t do that, and his entire game was committed to finding a Firespout or Engineered Explosives to deal with it. He was focused so much in this game plan that he didn’t realize he could have simply dropped his sixth Mountain to Valakut it away!

It is not because you usually play a card or a combination of cards in a way that it always has to be so – a lot of people would win a lot more games if they only remembered things such as the other two modes of Jund Charm. At Grand Prix: Columbus, my opponent played Meddling Mage on Flash, so I simply hardcast Hulk and killed him, and Tiago Chan did the same thing in the finals of the Invitational versus Richard Hoaen. The easiest way to prevent that, if you are not particularly skilled for thinking outside the box, is to just play a lot of games, so that every possible situation comes up – the first time you kill someone by playing a Protean Hulk, or by manually playing your sixth Mountain with Valakut, the option will be carved in your brain forever and you’ll always have access to it.

Though this usually happens with combo decks, it doesn’t have to be so – take Faeries, for example. Even after it was legal for years, people still played Mistbind Clique on their opponent’s upkeep on autopilot because that’s the way it was played the most, but most of the time it was correct to play it on their attack step, or on their end step, or on your own mainphase, or to not even play it – the old “your cards don’t have Provoke.” To recognize all this, it is imperative that you know all your cards’ powers and limitations, and that you don’t limit them any more than they already limit themselves – there is no “Play Mistbind Clique only during your opponent’s Upkeep” clause, and there is no “If Protean Hulk comes into play, if you played it from your hand, you lose the game,” so don’t act like they exist.

Other sources of playing on autopilot are thinking something does not matter, when in fact it does. Sometime ago, my friend told me of a PTQ match he was playing, with GW Greater Good versus Heartbeat. My friend had Yosei in hand, access to ten mana on his turn, and had Greater Good in play already, when his opponent played Weird Harvest for 3. My friend just went and grabbed his remaining three Yoseis without giving it much thought, to ensure that he would win if his opponent ever passed the turn. His opponent then played Naturalize on the Greater Good and passed. My friend could only drop one Yosei, and his opponent killed him next turn with a Maga for 27. Now if my friend had not played so fast and had actually given it some thought, he would have realized that there was no need for four Yoseis – one or two are already going to win him the game at this point if his Greater Good stays in play, and if it doesn’t then the Yoseis are not doing anything. If my friend had grabbed two Loxodon Hierarchs instead of two redundant Yoseis, then he would be at a convenient 28 life, and his opponent would have been unable to kill him. If he only had given it some thought he would have realized it, but he thought what he grabbed simply wouldn’t matter – either he was dead that turn, or he was going to win.

The truth is, we don’t always know if what we are doing is relevant or not – it might have been that my friend would think for 5 minutes, find the play of grabbing two Hierarchs, only to have his opponent either kill him, or pass without destroying the Greater Good, or play Maga for 60, but as it turns out it might be the decision of the game. You don’t actually lose anything to stop and think about it for a while, as long as you don’t dwell on it for five hours and end up falling into the overthinking trap. If it turns out it was irrelevant, then in the end it doesn’t matter that you thought about it, right? It’s all irrelevant. But if it ends up being relevant, you will be glad you spent those couple of seconds thinking. Sometimes, it is a comforting thought to have things outside our control – if you accept that whatever you do won’t make a difference and the outcome of the match is in the hands of fate, then you have no pressure whatsoever – but that is rarely the case. Have you never had this conversation with a friend?

Friend: I lost; my opponent played this, this, and that, and I couldn’t do anything.
You: But why didn’t you block with this other guy instead? That way if you draw a Thirst for Knowledge into Plains and Wrath, you have a shot.
Friend: Oh, but the top card was a land, so it didn’t matter the way I played…

Yes, in this situation it didn’t matter – but as you have no idea of what is going to happen, always play as if what you are doing is relevant. If it is not, you lose nothing, but if it is, you have a lot to gain.

So, to recap, the situations you should watch for playing on autopilot generally are:

– You make plays but you don’t really know why, so you are unable to recognize the situations where you should NOT make them.

– You have it set in your mind that certain cards are better or worse than the others, and that makes you assign priorities on what to kill/counter/resolve/protect not based on what the game demands, but on how good a certain card is in the abstract.

– You value concepts such as Card Advantage when they are not relevant for the particular situation.

– You are used to your deck working in a way that you are blinded to the other possibilities that present themselves.

– You presume you don’t have to think because your choice is irrelevant, but sometimes it is not and you have no way of knowing.

The way around most of those is, really, just to understand what is going on, and why. Think of it as learning a Math concept. You might learn, for example, that Sin90 = 1. You can just memorize that, and it’s going to be enough to go through any problems in which you need to know the Sin of 90, but at some point you are invariably going to be faced with Sin180, and then you’ll be doomed. If you understand why Sin90 = 1, then, even if you have never heard of Sin180 before, you will be able to clearly see the value of it if you must. That does not mean you have to draw the Cartesian Plan every time you come across Sin90 – if you know it by heart, just use the number – but the knowledge of the reason behind it helps when the situation is not exactly what you prepared for. The same is true for Magic – the more general knowledge of how things work you have, the better prepared you will be to face a completely new situation, and those are the situations where playing on Autopilot kills you.

I hope you’ve enjoyed this, and Feliz Natal to you all!