Hey everybody! I want to bring you a conceptual article this week in order to give you some perspective on the mentality I have when I approach play testing and deckbuilding. From what I have seen very few, even among top players, use this approach, yet I would say this approach leads to a considerable number of round wins when we actually get to the tournament.
The basic concept is simple: don’t try to win every game, but instead try to prepare for what will win you the most games out of 1000.
What’s the Difference?
On the surface, it seems like they’re almost the same thing. “If I try to win every game, I’ll have won the most after playing 1000.” This isn’t actually true, but it is a popular mentality among the player base.
The entire notion that luck exists is comes from the misconception that there isn’t a difference between the two.
Luck does not exist. Your opponent drawing “the one out!” wasn’t them getting lucky or you getting unlucky. If they had a 10% chance of drawing the card, people see the fact that 90% of the time; they would not have drawn it. That doesn’t mean that they never draw it. Simply, they would draw it 10% of the time.
“Luck” gives the assertion that some outside force is acting against you. It doesn’t “always” happen to you, it’s just a preconditioned bias among humans to remember more distinctly when bad things happen than when the average or probable thing happens. You can maximize the probability of achieving your desire outcome and the entire notion of luck just acts as an excuse for people not acting on the things they could control.
Because of this, it’s entirely possible to make a play that didn’t work out, but was the correct play. If you could choose that they have a 10% chance of drawing an out or a 20% chance, you’d clearly pick the smaller chance. If each of those percentages represents a different card, the 10% being a 1 of and the 20% being a two of, what happens when you chose to play around the 2 cards and they still drew the 1?
This still isn’t an example of them getting lucky. It’s supposed to happen some of the time. Knowing this, we can make a distinction between trying to win every game and trying to win the most out of 1000.
The difference is whether you are looking at the short term or the long term. In the short run, it’s entirely possible that putting them on 1 out instead of 2 outs didn’t work because they drew the 1 and didn’t draw the 2.
In the long run, this isn’t the case. Probability will hold to be true. If you look at 1000 games, they will have drawn the 10% 1 of about 100 of those games and the 20% 2 of about 200 of those games, thus making the rational correct play to make it so they have to draw the 1 of.
Tournaments aren’t won and lost by playing for individual games. They’re won by playing to maximize your probability of a desirable outcome in the long run.
Application in Deckbuilding
Why do I play Upstart Goblin in most of my decks? It’s because I deckbuild for the long-run, not the short-run. Maxing out on Upstart increases your likelihood of drawing any card by roughly 3%. That may seem insignificant, but really it’s not.
Look at Tour Guide From the Underworld. Over the course of 1000 games, you’d draw it in your opening hand roughly 30 more games than you would have if you didn’t play it. Let’s not forget, you gain 3% on each card, not just 3% overall. You’d also draw other desirable at that increased rate.
In order to win a YCS you don’t need to not lose a match. You need to go at least 10-2 and then 4-0 in draft. When you’re preparing your deck for an event, the first part of that is all you’re concerned about.
If you were to look at deckbuilding on an individual game level, three percent may seem insignificant, but looking at the its impact on your deck’s performance for the entire tournament and it becomes much more significant. It’s very realistic for that three percent to account for the difference of going x-3 and just missing and x-2 and doing what it took to make it to draft.
This also applies to cards that “win you games.” Just because a card is capable of winning a game, doesn’t mean that it’s worth playing. When I am determining what cards to play or not play, I try to think about how that card would play out in the long run, not just for individual games.
This is what I’m talking about when I say that I don’t think Snatch Steal is worth playing in Burning Abyss. It’s not that it won’t win you games, because it most certainly will. I’m saying that over the course of 1000 games, I don’t think that scenario will occur often enough to compensate for the downsides of the card.
Application to Play Testing
I had an interesting conversation with Chris Mattiske last week, where I told him I was playing Masked Hero Koga as an out to Towers. He pointed out that they could just play around the card by not summoning any monsters besides Towers until it was possible for them to OTK you. I told him that while he was right and that people could play around it, they simply wouldn’t play around it.
“Surely you test against optimal plays, right?”
“No, absolutely not.”
I’d argue that it’s incorrect to assume that your opponent will always make the best play. Quite simply, they won’t and your testing will be inaccurate if you don’t adjust for it.
Anything basic, we’ll have the opponent follow the correct line of play. Anything complex, we acknowledge a typical opponent wouldn’t see it and try to instead make the play that they do see. I’d argue that the typical player isn’t going to be actively thinking about how to play around Koga, and thus will put another monster on board if it is seemingly appropriate to do so.
The Pareto Principle is a rule of thumb that says 80% of effects come from 20% of causes. For Yu-Gi-Oh, that means that 80% of the results will come from 20% of the effort. It makes sense if you think about it. Most people don’t spend 30-40 hours a week testing, but if there’s a top 32 cut someone has to top. Whenever we test, we try to make the plays of the opponent on par with what we think the typical person will know to do.
“What would you do if you played against a top Qliphort player?”
“I’d lose if they played around Koga and summoned Towers. That’s not the majority of the tournament though.”
This translates back to deck building as well. If I concede that most people will not actively play around Koga, I can justify having Koga as an out to Towers and not playing any additional outs.
If I didn’t make this distinction in my mind, I would likely subject myself to strictly worse cards. Let’s say I use Rank-Up Magic instead of Mask Change II for Koga. Even though I can rank up into C69 regardless of whether or not my opponent summoned an extra monster, I would say that the card has less utility over the course of 1000 games than Mask Change II. Mask Change is stronger in most scenarios, except my opponent summoning Towers and not summoning additional monsters. That one scenario alone will not justify the choice in the long run. I can acknowledge that they will make suboptimal plays a considerable number of those games and still summon an extra monster. The few times that my opponent thinks to play around it in those 1000 games isn’t worth the games I give up by playing a worse card, Rank-Up, for 1000 games.
Next week’s article will be a follow-up to this week’s article. I’m going to be writing on the importance of not becoming complacent and to always be seeking active improvement, something losing in the finals of YCS Charleston this past weekend served as a solemn reminder of. I look forward to seeing you all at the Circuit Series in Fort Worth, Texas this weekend. Until next time, play hard or go home!