What is luck?
“Luck is used as a convenient explanation for any event that we don’t want to believe was caused by another person’s effort.”
Marc Myers
One classic example of this is when the Red mage “gets lucky” and topdecks that burn spell he needed at the last possible moment. But was it luck? Aside from playing in such a way as to put himself in this situation, he obviously made the decision to put it in his deck, giving him the chance to “get lucky”. If it’s Limited, he did draft it. Even if it’s Sealed, you have to choose to play the color and the card to have a chance to draw it.
Magic players have a tendency to blame “Bad Luck” in one form or another for a disproportionately high number of loses or other Bad ThingsTM. Your expected win ratio, in any particular match, is a function of the skills you both possess (all types combined), and the cards you are playing (match-up in Constructed, card pool in Limited, etc.).
This will not tell you who will win, of course, but rather how likely it is that one in particular will beat the other. The difference between the favorite and the underdog reflects how much luck is needed for an upset. The luck is viewed as good or bad… depending on which side you are on, of course.
Relatively speaking, if we could quantify Player A’s advantage, it would take twice as much “luck” for Player B to overcome the odds, than if the match was 6 – 4 (a difference of 4 compared to 2).
This is all well and good, but where the inaccuracies slip in is when people try to evaluate this equation with regards to themselves, or some match they have a personal stake in (someone they want to win is playing, or a deck they designed is involved).
First of all, as mentioned in “Information Cascades in Magic”, people, especially experts, are chronically overconfident (in general). Almost everyone thinks they are better (more highly skilled) than they actually are. To a lesser degree, they also tend to think that people they have never heard of are worse (less skilled) than they are.
This is even more so regarding experts who may be pros or not. Experts have information that they are “better” than the average opponent. While this is true (if they are actually an expert player), people have a tendency to overreact to information they learn. This is part of the root of a sophomore jinx. When people first learn they are good (or experts) is when they are actually most likely to overestimate their own ability.
It is easier to see when you look at someone else. Ever notice how when someone qualifies for the Pro Tour, they are suddenly an authority on all sorts of things? If Joe Schmoe wins a pig with his triple Savage Twister deck he drafted, should he really be the authority in your playgroup on the Urzatron versus ‘Vore match-up?
When you combine this with natural overconfidence (a trait more often found in highly skilled players) this leads to the classic scenario of a “Pro” (or any good player for that matter) overestimating his or her chances of victory against a random opponent or a rogue deck.
If the opponent is “random” (as in a stranger, or at least not someone the player knows to be highly skilled), then the player will tend to associate that opponent with the average unskilled opponent, not the average opponent. If the opponent is being viewed as a “random”, it implies he is not highly skilled.
For instance, let’s say skill could be measured 1 – 10.
1 – 2: Noobie
3 – 4: Poor/Inexperienced
5 – 6: Average
7 – 8: Highly Skilled
9 – 10: Pro Caliber
Now, let’s use the example of a highly skilled player whose skill is a 7. To begin with, inherent overconfidence (possibly exacerbated by his expertise) creates the self-perception that his skill is an 8 or so.
In addition, let’s assume that player is piloting some Tier 1 deck such as Urzatron, ‘Vore, or even some homebrew that the player is convinced is excellent. As a result, the player overestimates his deck’s advantage, – in general +/- 0, putting him at a 9.
On the flipside, the player estimates his unknown opponent to be a 3 or 4, despite 5 – 6 being average. This assumes we are at the PTQ level. Why is this? Well, the player will generally assume that if the opponent is not a “Pro”, then they can’t be a 9 or 10. If they were a 7 or 8, the player should at least recognize them, if not personally then by their behavior, or so the subconscious logic goes.
As a result, “highly skilled” players have a tendency to assume random opponents at random tournaments (PTQs, etc) are 1 – 6’s. When the random opponent is running a traditional strategy or deck, he is given credit for at least knowing enough to ‘Net deck.
If the opponent is using an unorthodox strategy, or at least some unusual card choices, the “highly-skilled” player may tend to subconsciously assume the random opponent is less skilled for making those deck decisions. In addition, he may assume the match-up is in his favor. After all, his deck is tried and true. The opponent is rogue (i.e., probably not Tier 1).
As a result, a random opponent, who very well may be a 6 to begin with, is assumed to be a 4, on account of being an unknown. This drops to 3 for “being bad enough to change the deck”; then he is lowered to a 2 considering how “surely Urza beats his rogue pile.”
At this point, the “highly-skilled” player (who is actually only slightly more skilled than his opponent) perceives the differential in this match to be 9 – 2, in his favor. Obviously, this is a far cry from 7 – 6.
Because of this, the “highly-skilled” player assumes it will take a tremendous amount of “luck” for the random opponent to defeat him. After all, he is an expert with a Tier 1 deck and his opponent is some random scrub playing second-rate cards, not to mention a second-rate strategy, right?
After the “highly skilled” player losses to his random opponent’s Rise/Fall hitting the two perfect cards, he proceeds to find every reason in the world why he had bad luck while his opponent had good. He has to do this, in order to explain the upset when he perceived the skill and deck differential to be 9 – 2.
Let’s look a little closer at that “highly-skilled” player’s loss with Urzatron versus some random opponent playing a Rise/Fall deck. To begin with, while Urzatron is a very good deck, it is also widely known. There is no telling what is in the Rise/Fall deck… not even the colors it plays (for instance, some B/R/W decks run Rise/Fall). The Rise/Fall player likely tested against Urzatron strategy, whereas the Urzatron player may have never seen the Rise/Fall player’s strategy before.
Consequently, while the Urzatron deck may be stronger in the abstract (which is not necessarily the case), the Rise/Fall player may have compensation for this. This may equalize the match-up, or even tilt it in Rise/Fall’s favor if he is tuned to beat Urzatron.
I am not suggesting that Rise/Fall decks beat Urzatron (though the card is good versus the deck). That I am suggesting is that right now, in Standard, the “50th best” archetype is not that much worse than the best. Be careful not to overestimate your position on deck strength alone. Individual match-ups are far more telling.
Aside from deck advantage, there is also the issue of underestimating the random opponent. There once was a day when there were literally only 50 “good” Magic players on Earth (early Pro Tour days).
Now, with 13 years of Magic playing and theory, an ever-growing online supply of articles and information to draw upon, and in recent years the rise of MTGO, the random opponent is far better than he used to be. The average regular PTQ player, these days, could have made it on the Pro Circuit 10 years ago.
Competition is fiercer than ever, and you face more threats at every PTQ than you may realize. The random PTQ player’s near average is probably a 5. If he were a 10, you are likely to have heard of him (if this is a local PTQ and he is a local player). In addition, high-level Pro-caliber players are unlikely to be playing in PTQs… though many have to, and do. Just because you don’t know someone or they act eccentrically, do not assume they couldn’t be a 7, 8, or even higher.
In our example of the highly skilled Urzatron player getting “unlucky” and losing to “lucky” Rise/Fall opponent, Urza finds ways to try to explain the discrepancy between the perceived differential and the outcome. To explain a 9 – 2 upset, many unusual things probably needed to happen. As a result, Urza blames not drawing the ‘Tron, not having double Red when he wanted, his opponent having an ideal mix of lands and spells, and losing the coin flip… in addition to Rise/Fall hitting the two perfect cards on the critical turn.
Let’s examine all that so-called “luck.” Urza is mad that he didn’t get his three-card combo (the ‘Tron) with minimal library manipulation. Is it really that unlucky he didn’t draw it? Does he “deserve” to draw the ‘Tron?
Urza didn’t have double Red for the turn 5 Wildfire? You know, twelve sources of Red is fewer than what Zoo plays, and Zoo can’t count on getting a single Red all the time.
His opponent had a good mix of spells and land? Should he not? Never mind if he did something to give himself an edge here, such as play a Rakdos Carnarium.
How about losing the flip before the match? Surely, that is the game-losing luck, right? Sure… Assuming you are willing to chalk up over half of your wins to when you get lucky in the same way.
So, we are left with the perfect Rise/Fall. Yes, he was fortunate to have it at an opportune moment and to hit the two cards he needed. Yes, he may have needed this little but of luck to overcome the 7 – 6 differential. However, he also may have put himself in a position to “get lucky.”
Aside from picking the right moment to cast it (and the right half!), he made the decision to put it in his deck at all. Was that luck? As soon as he made that decision, he effectively had used his skill (deck building, predicting the metagame, etc.) to determine how he would be compensated for playing this particular card, and when. The “when” is randomly distributed in terms of bound states and which cards you pull. However, a card that’s compensation would be distributed this way.
While the Urza player perceives the Rise/Fall player as being lucky, in reality he was merely reaping the randomly distributed rewards of playing those particular cards. It is no different than the randomly high value of a Wrath of God (do they play a lot of men?) or a Karoo (are all the advantages worth the cost in tempo in this game state?).
Simply put, the Rise/Fall player chose to play Rise/Fall. The Urza player didn’t. The Rise/Fall player played to put himself in a situation where he could win if he hit the right cards. The Urza player chose a strategy, both in deck choice and game play; that left him vulnerable to “the perfect Rise/Fall” (which may have been the best in the abstract, but which didn’t work out in this instance).
The Rise/Fall player gave himself a chance to “get lucky”. There are countless ways to give yourself changes to get lucky. Michael J. Flores would say you are looking for an edge.
Why was Mark Herberholz so lucky in Hawaii? Why is Kai Budde the luckiest Magic player ever? Both captured every opportunity to “get lucky”. They maximized intangible and subtle edges, as well as obvious ones. These little edges pay you back in the form of “luck.”
What are some ways to get lucky?
1) Play with good cards. Good cards (Wrath of God, Loxodon Hierarch, Keiga, Sakura-Tribe Elder, Dark Confidant, Char) are, in general, better than not-as-good cards (Culling Sun, Suntail Hawk, Windreaver, Civic Wayfinder, Yamabushi’s Flame). This is not the end all rule. It is simply one of the ways to get an edge. People with Keigas win more than people with Windreavers. It is a fact.
This may seem obvious, but it needs to be said. Despite being allowed to play the best, people chronically play subpar cards. I realize good cards are often expensive, but if you choose not to play Izzet Boilerworks, that is on you. If you enjoy substandard cards, as many do, more power to you. Do what makes you happy, but weigh your happiness of playing your pet cards versus your happiness from winning more. For the record, I play strategies I know to not be the best, but that I enjoy greatly, sometimes. It is a game after all, and sometimes the fun route is almost as good as Spike’s.
If you want a chance to get lucky and out-topdeck your opponent, play with better cards… Mysteriously, you’ll consistently draw better cards. If you can’t decide between cards, the final tiebreaker is power level. You need a good reason to play with any six-drop that is not a Dragon Legend, for instance.
2) Use cards with synergy. Again, this may seem obvious, but this is why decks like Husk and Ghost Dad are possible. If you are going to use anything but the most powerful cards available, it should be because of their synergy with the rest of your deck.
The “synergy versus power” debate is one as old as the game itself (well, after we got past 21 Black Lotus, 20 Wheel of Fortune, 2 Fireball). An obvious modern day example is the Eye of Nowhere versus Boomerang debate in ‘Vore. Synergy versus power… Which is better? That must be answered on a case-by-case basis, though the burden of proof is always on synergy, since power is trump in a vacuum.
Playing with cards that are synergistic is like playing incredibly powerful cards… when you draw your combos. If you want a chance to get lucky and draw a combo early, you have to choose to play it. In addition, you must make decisions about whether to use library manipulation, have a backup plan, or use redundant combo parts.
3) Use cards or strategies people haven’t seen or aren’t familiar with. This is the advantage of “going rogue”. The primary edge here is that you are, for the most part, prepared for your opponent. He might not know what to expect or haw to fight your strategy.
If you want to have a chance to get lucky and play someone who has never seen your strategy, or who misplays due to not having what to play around, you need to have new tech. Beware overestimating the edge you gain in this way, though.
Most rogue decks do not feature enough power or synergy to be considered Tier 1 once the cat is out of the bag. Truth be told, most aren’t actually that great even during their maiden voyage when they are typically most potent. If you are going to innovate, remember power and synergy, etc.
4) Employ a strategy that is good versus the metagame. Whether it’s playing ‘Vore in an Urzatron metagame or pre-side boarding Cranial Extraction; playing the metagame can be very profitable. The key is how useful your information is, and how well you take advantage of it.
If you want to increase your chances of getting “lucky parings,” obtain as much information on the metagame as your can. Then, make deckbuilding decisions based on it. Is it lucky to be paired up against Urzatron when you run ‘Vore? Or, is it part of the random distribution of rewards for your deck choice? Is it unlucky for the guy who gets his Keigas, Magnivores, Early Harvests, or Firemane Angels Extracted Game 1? Or were both sides just reaping what they had? One side played a narrow strategy, the other was pre-sided boarded. Remember, both the guy with main deck Cranials will have to pay the piper if Mr. Herberholz and friends start summoning Kird Apes.
Figuring out what to play (and what the metagame is) is akin to deciding which hands to keep in Texas Hold-Em, in which situations. Sure, it may seem like luck when the guy with 7 – 8 suited catches a straight or a flush, but what did you think he was hoping for?
Perhaps he gave himself a chance to “get lucky” by paying to see the last card when the pot odds warranted it.
When you maindeck a Cranial Extraction, you are paying to see the last card. The potential pay-off is huge, even if it is worthless often (as 7- 8 suited is). The key is figuring out the pot odds. Just how many wins will you get as a result in relation to losses.
An interesting point to remember is that drawing one can win you a significant number of games. At worst, drawing one is a dead card. While this is bad, it is hardly a loss. A mulligan is not a loss; if you have ways to make up for it (Wildfire, Phyrexian Arena, Tidings, Karoos, Bottled Cloister) or ways to get some use out of it (Sickening Shoal, Compulsive Research, Boomerang), it may be less than a dead draw.
5) Play a consistent deck. While this often flies in the face of power (i.e. Heezy is less powerful than Zoo), consistency needs to be kept in mind if you want to increase your chances of “getting lucky” and drawing a good mix, your colors, or even a curve. Likewise, you will get less “unlucky” if your deck is consistent. Remember, power is still trump, but consistency is another dial to turn.
The most obvious way to help your consistency is to play wonderful mana. Many dual lands offer power and consistency. Is it any surprise they are so vital? In these days of limitless options for mana, there is no excuse not to make yours simply beautiful.
Here are some tips for random deck (of course, these are only guidelines).
- Play 3 – 5 Karoos. We all know how great shock lands are, but remember, Karoos are overpowered. (Yeah, I said it!) They are so good it is worth warping many decks to accommodate them. For more on why, check out Classic U/W Control – Karoos > Good.
- If your deck is two colors (or close), start with the full package of 4 shock lands, 4 Karoos, and 4 pain lands. Then, if there is a compelling reason, deviate. Typically, you’ll want all 23. You can always vie for better mana. The amount of pain two color decks inflict is negligible. Personally, I don’t mind the pain of playing 15 or more lands that hurt me, as long as most are Shock lands.
- If your deck is three colors, things are much trickier. May I suggest only playing 1 or 2 basics (Ghost Quarter much?), unless your deck is primarily one color (usually Blue) or you fetch basics (Elder, etc.). A starting point is 8 – 12 Shock lands, 3 – 5 Karoos, 0 – 5 pain lands, perhaps Legendary Lands, or Signets.
- Copy the manabase of a deck with a similar curve, even if its colors are different. For instance, switch Orzhov Basilica to Rakdos Carnarium. If you are adding a third color, just turn almost all the basics to duals and Ice Bridges and mess around with them until you get a mix you like. It is better to err on the side of too many duals than too few.
These days, while mana is easy in the sense that color-fixers are plentiful, it is much harder to perfect a manabase due to the plethora of options. Perhaps, if there is interest, I will dedicate an article to an in depth look, as I can’t even scratched the surface here.
6) Maintain a good reputation while networking heavily. This gives you countless opportunities to “get lucky.” If you are on friendly terms with most of the guys and gals at a tournament, you are more likely to get that draw or concession you need (want).
You are more likely to get lucky and hear about some tech others aren’t up on yet. You’ll have more information on the metagame, which you can use to increase your chances of getting lucky pairings. You will hear about more of what people are up to, taking away your opponent’s ability to get lucky and hit you with a Shining Shoal you weren’t expecting.
Your “luck” might even come in the form of someone trading or loaning you cards your need, right before a tournament. Let me tell you, I was very lucky at the Saga-Legacy Constructed Pro Tour while Round 11 was starting. I had suddenly found myself deckless. A group of players and dealers built my rare-packed deck from scratch for me in 3 minutes. A situation in which I was very “unlucky” was salvaged due to my good fortune of knowing kind people.
7) Play as though you’ll draw what you need, if that is your only or best plan. Conversely, if you will lose to a particular card and it is unreasonable to play around it, don’t. Just assume they won’t draw it.
If you want to get lucky and topdeck that you need, play so that you have a chance for it to matter. If you can only win by having a Char on top and 4-for-1’ing yourself this attack phase… Do it! You probably won’t draw the Char, but at least you are giving yourself a chance to get lucky.
If you can risk losing now to Char or Lighting Helix, or you can play around Helix (still losing to Char) but it takes you four more turns, it may be worth risking dying now. This is a situation where you must decide how to play based on the information you possess. By playing around Helix, you may be giving him a chance to get lucky and draw Char.
Also in this category are situations where you can only win if your opponent misplays. If you cast Wildfire and your opponents Remands it, if you are holding a Remand and you will lose if Wildfire doesn’t resolve, Remand his Remand… Just assume he won’t replay it. He probably will and you’ll lose if he does, but you’d lose with any other plan. This one gives you a chance to get lucky.
Luck in Magic is the result of all our actions playing out with relation to our opponents. Every card you play with, every person you talk to, every article you read, and every spell you cast affect the probability of things (such as winning) happening.
Think of reality as a matrix of an incredible number of universes (whatever that is). In each universe, one thing is different from every other universe touching it. Every time you make a decision, you are navigating your consciousness through this multiverse, altering the number of universes around you in which you win (among other things).
Even if you lose in 97% of the universes than the one you are currently in could lead to, you could “get lucky” and end up in one of the 3% you win. Good luck is arriving at a universe you want to be in. When you do things to improve your “luck,” you are in essence steering towards a collection of universes that is more likely populated by yours that win.
Are there actually an incredible number of universes that you are maneuvering though? It doesn’t matter, because it can be useful to operate as though there are, regardless of what the truth is.
At risk of going to far out there, I will say goodbye for now. Just remember who typically wins: the guy who is really good or the guy who got really lucky.
“Better lucky than good.”
Patrick Chapin
“The Innovator”
— Getting Lucky Since the Summer of ’96 —