Feldman 04 30 2008

Deep Analysis – How to Think, by Richard Feldman at StarCityG...

1 of 7

http://www.starcitygames.com/php/news/print.php?Article=15799

Deep Analysis – How to Think by Richard Feldman

I am about to give you the best advice for competitive Magic: the Gathering you will ever receive. It transcends all theory and is the actual Holy Grail of every Magician who aspires to win at Magic. Ready? Here it is: “Win.” Got it? Great! Now go do it. Until next time, this is Richard Feldman, tapping the cards so… er, sorry. Obviously this is not the most practical advice. There is a slim chance you hadn’t already considered that the way to win at competitive Magic was… to win, so I’m not really telling you anything you didn’t know already. Still, we all agree (I hope) that the above piece of advice is good advice that everyone who wants to win at competitive Magic should heed. At the opposite end of the spectrum of indisputably good advice is something like the following. “If you want to win the game, never decline to counter a spell that will definitely kill you if it resolves.” Again, I hope we can all agree that this is good advice. If you want to win the game, and failing to counter a certain spell will definitely end the game with a loss, you had better damn well counter it if you can. Also, as with the previous example, this is not the most practical advice. Pretty much everyone who is reading this article already knew what to do in that situation. The trouble with indisputably good advice in Magic is that, with precious few exceptions, it’s either so general you already knew it, or so specific that if you didn’t already know it, it might not ever come up in your lifetime. For example, I could generate a million different permutations of rock-solid, hyper-specific pieces of advice along the lines of “If a 6/6 Hunted Wumpus with a +1/+1 counter on it is attacking you, and if you are at 7 life, and if not blocking it will definitely make you die, you should block it,” but chances are

5/1/2008 11:29 PM

Deep Analysis – How to Think, by Richard Feldman at StarCityG...

2 of 7

http://www.starcitygames.com/php/news/print.php?Article=15799

you could have either figured these things out on your own, or - when I start to get really absurdly specific - they’ll pretty much never come up. Take another oft-discussed piece of advice: “Focus only on what matters.” That’s slightly less broad than “Win,” but is it indisputably good advice? Honestly, I think it is. If you decline to focus on some aspect of the game that you don’t think matters, and the fact that you didn’t focus on it makes you lose, then I’d say it mattered. You can try to dispute that, but unless I’m missing something, I don’t think you’ll get very far. However, like “Win,” this advice is incredibly broad - and that makes it less useful. Yes, I know I should win at Magic, but how do I do that? Yes, I know I should focus only on what matters, but how do I know what matters? Figuring out what matters and what does not in Magic is an incredibly complex undertaking; if you’re fortunate enough to have the intuitive grasp of it that masters like Jonny Magic do, then good on ya, but for the rest of us, this maxim that is so often quoted is really not as helpful as we wish it were. Still, it is more helpful than “Win.” It gets you thinking about the relevance of your actions, asking questions about what you’re doing. “Does this matchup matter? Should I really be spending my energy here, or should I test something else? Does choosing between two similar plays here matter, or should I just pick one so I don’t run myself out of time?” This is a useful mindset to be in, so even if you can’t follow the advice to the letter (because you realistically just can’t know exactly what matters and what doesn’t when it’s decision-making time), it may still help you out by getting you to think about the relevance of your actions. The tricky part about Magic advice is that most of it falls somewhere between the very broad “Win” and the very specific “Don’t decline to counter a spell if it will kill you” - and in that space, it gets a lot less clear that the advice is actually worth following. “Never play a deck until you’ve thoroughly tested it.” Is that good advice? It’s not too general, not too specific, but it’s a far cry from indisputably good. In fact, it’s not even clear that it’s worth listening to at all. Theoretically, I could do an experiment to help test the merit of this piece of advice. I could round up 1,000 Magic players, have half of them play decks they’d thoroughly tested and half play decks they hadn’t thoroughly tested, and see which group did better. Still, even if I ran an experiment like that, I’d only know that the average Magician does better when he has thoroughly tested his deck, not necessarily that one should never play a deck unless they have thoroughly tested it. All the experiment would let me conclusively say was that “Magic players do worse, on average, when they have not

5/1/2008 11:29 PM

Deep Analysis – How to Think, by Richard Feldman at StarCityG...

3 of 7

http://www.starcitygames.com/php/news/print.php?Article=15799

thoroughly tested their decks,” not “You should never play a deck until you’ve thoroughly tested it.” Why is this distinction important? What difference does it make? Whenever I criticize a piece of theory, someone is quick to tell me I’m nit-picking, and that the distinction doesn’t matter. But it does matter. It is important. It is important because of Mark Herberholz. If Mark Herberholz had followed the “Never play it if you haven’t tested it a lot” maxim, he would have showed up to Pro Tour: Honolulu with the Beach House deck that so many of his masterful Pro Tour compatriots brought (and bombed out with). But he didn’t. He knew that it would be much riskier to play, say, a Green/Red decklist he invented basically on the morning of the tournament, but he saw a chance to make a metagame call that could beat the odds to death, and he took it. And so he won the Pro Tour. You can say that “well, okay, in that one case, you can make an exception,” but if you have to selectively ignore a piece of advice, analyzing it every time you make a decision to see if it actually applies or is totally misleading, then the advice is full of holes. It’s leaky. It’s not trustworthy. Worse, the fact that it only applies in some situations and not others calls into question whether or not the person who gave it to you actually knew what he was talking about; did he really know that “people do worse on average when they don’t test,” or did he genuinely believe that “you should never play a deck you haven’t tested a lot,” like he said? Is the person giving me this advice really so oblivious that he’d never heard of Herberholz’s PT win? If so, what else was he oblivious to when he gave me this advice? Sure, “not testing does worse on average” doesn’t have the same ring to it as “never play it if you haven’t tested it,” but it’s trustworthy. It’s telling you exactly what it means, so you don’t have to stop and wonder which part of it is - let’s face it - an outright lie that you’re expected to disbelieve in certain situations. It’s up-front about what it doesn’t cover, and doesn’t promise what it can’t deliver. It warns you that your chances might be lower if you don’t play the deck you’ve tested, but it leaves it up to you to figure out if you should actually take the plunge and run with the last-minute creation. And that’s a good thing, because - guess what? Sometimes you should play the last-minute creation. Just ask Herberholz. Unfortunately, advice-givers - from the guy at your card shop who wins just a bit more than you do, to writers like myself, all the way up to Kai and Jon themselves - don’t always give you the trustworthy version. In fact, a lot of them give you the slick-sounding version that is full of holes, and it’s up to you to figure out what you can and cannot take away from it. Several years ago I was drafting on MTGO, and a guy with a higher Limited rating than mine told me, “Drafting is easy. Just pick removal.” Should I have followed that advice?

5/1/2008 11:29 PM

Deep Analysis – How to Think, by Richard Feldman at StarCityG...

4 of 7

http://www.starcitygames.com/php/news/print.php?Article=15799

Not to the letter, certainly - it’s full of holes! Obviously I can’t just pick removal and expect to win, unless at least some of the removal damages my opponent in some way. That’s a simple way to prove that the theory has holes in it, but assuming I believe the guy knows what he’s talking about, what am I actually supposed to take away from his advice? Should I take removal and then a couple of finishers? Should I just always value removal more highly than creatures? If so, should I do that all the time, or should I make exceptions for bomb creatures? These are fundamental questions you have to ask yourself when you get a piece of advice. First, you ask if the source is credible. Do I trust this person to give me advice? Do I think they know what they are talking about? Did they do their homework? Next, you figure out if the advice makes sense. What holes can I poke in it? Can I think of one or more situations where following the advice would cause me to fail? How often are situations like those going to come up? Finally, you figure out which parts of the advice to follow. If the person didn’t give you a trustworthy piece of advice, chop off the absolutes until you make it trustworthy. If you think the source is credible and the advice is worth considering, you still shouldn’t leave “Just pick removal” as full of holes as it is when you file it away for inclusion in your decision-making process. Question it, examine it, revise it until you arrive at something more like “Favor decent removal over decent creatures, unless you are running out of finishers” that you can actually trust to inform your decision-making process. Now for a story. I defended a weird card choice awhile ago. At the time, this card - let's call it Card Alpha - was a draw spell from a new set, and its merit was fairly disputed. People weren't really sure if it was worthy of Constructed play or not. Patrick Chapin tried it out in of his decks, but mentioned that several other people were suggesting he put in the more conventional alternative let's call it Card Beta - instead. I was getting advice from all sorts of different sources. On the one hand, Patrick said it was good - but mentioned that other people thought it wasn’t. On the other hand, my testing went positively - but I hadn’t really given Beta that much of a try yet. On the other, other hand, most people around the Internet seemed to be saying Beta was better. What’s a guy to do? Part of the debate was that Card Alpha was powerful when drawn early but lackluster as a late-game topdeck, while Card Beta was fairly consistent in both the early game and late game. If consistency were the only metric, Beta would obviously get the slot every time, but Beta was also more expensive than Alpha, and less powerful in the early game besides. As it should be, a debate in the forums arose. One forum poster reported, "I literally doubled [this deck]'s win % with [Beta]

5/1/2008 11:29 PM

Deep Analysis – How to Think, by Richard Feldman at StarCityG...

5 of 7

http://www.starcitygames.com/php/news/print.php?Article=15799

over [Alpha]. [Alpha] was bad versus UG. I found I was stumbling on the mana a lot at 22 lands and [Beta] helped to fix that as well as helped to win topdeck wars in the midgame." This is a good, solid argument. It notes a problem the deck is having, and shows how Beta helps to fix that problem. That is a grade-A argument in a vacuum... but remember, just as we can’t take advice at face value without questioning it, nor can we take peoples’ arguments at face value either. If we ever aspire to win Pro Tours, Herberholz-style, we have to question things. I’ll start with the obvious question: is the source credible? This is important, because his argument is based on the premise that the deck has trouble with UG, and that he was stumbling a lot at 22 lands. What if the real reason the deck was stumbling was that he had a poor mulliganing strategy? Or that he was mis-piloting the deck and did not realize he should not have needed as much early mana as he thought he did? The caliber of the player impacts the foundation of the argument. In this particular case, the poster was reputable; we can take his testing result at face value and move on. What holes can I poke in this argument? I can’t think of any off the top of my head, which is why I’d say it was a solid argument. His claim was simply that he doubled the deck’s win percentage in the UG matchup by playing Beta over Alpha, and gave a plausible explanation for the different results. Another poster quoted the previous comment and asked, "Is that versus just UG or versus any other decks?" Is this a reasonable question to ask? Certainly. At the time, UG was one of the Decks to Beat, so it was definitely important how your deck did against it - but it was not the only deck, and there was no mention of how the deck did with Alpha versus Beta in the other matchups. What if the 22-land Beta version did twice as well against UG but half as well against the rest of the top tier? That would certainly change things! This wasn’t exposing a flaw in the original poster’s argument so much as it was pointing out that the validity of that argument wasn’t enough to settle the dispute on its own. A third poster had a response: "It's probably against all decks. [Alpha] is just poor." What do you think of that? For my money, it's leaky as hell. This card is always poor? Really? No matter what? It takes me a split-second to come up with a situation where the card is the best for the job. Has this person really done their homework? Maybe it turns out Alpha is poor. Maybe Patrick and I test it out some more

5/1/2008 11:29 PM

Deep Analysis – How to Think, by Richard Feldman at StarCityG...

6 of 7

http://www.starcitygames.com/php/news/print.php?Article=15799

and it is ultimately a real junker in every situation, and maybe years from now we look back and say "yeah, obviously Alpha is just poor and you shouldn't play it." However, unless you have a time machine and already know what the outcome will be, by jumping right to the end of the analysis, you're flipping a coin as to whether you’re misleading people or helping them out. When it comes to whether the card is poor or not in the format, when a new set has just been released, the fact is, you don’t know for sure, and you shouldn’t pretend to. In the end, I felt there were not enough counter-arguments to justify playing Beta over Alpha, so I went with Alpha and promptly punted my way out of the tournament. Should I have played Beta instead? We don’t know. We don’t know, and we shouldn’t pretend to. I could have lost for one of a billion different reasons; personally, I happen to think the deciding factor was my poor play towards the end, but lots of other things could have been the driving force instead - only one of which was the choice to play Alpha over Beta. For all I know, I would have done even worse with Beta. As my friend Aaron “Darth” Hauptmann once astutely pointed out, it’s very difficult to concretely prove a theory in Magic because so often the outcome is decided more because of random noise than because of what you were actually testing for. While that uncomfortable reality might make us squirm a bit, trying to pretend it doesn’t exist makes us overconfident and prone to make frustrating errors when we think we know more than we actually do. In a world of so much uncertainty, you can’t take very much at face value. You have to think about every piece of advice you get, every observation you make, every lesson you learn, with a very critical eye. If you don’t, you can end up thinking that cards like Ancestral Vision - er, I mean, Card Alpha - are “just poor,” and you’ll instead be playing Think Twice - er, Beta - because you just followed the advice you heard without examining it closely enough. I titled this article “How to Think” for a reason. Did I actually teach you everything you needed to know about how to think? (I sure hope not.) What holes can you poke in this article? How can you take what I’ve said and distill it down into just the elements that you are certain you can use to make decisions in the future? Now you’re thinking. See you next week!

5/1/2008 11:29 PM

Deep Analysis – How to Think, by Richard Feldman at StarCityG...

7 of 7

http://www.starcitygames.com/php/news/print.php?Article=15799

Richard Feldman Team :S [email protected]

5/1/2008 11:29 PM