On Losing and Sample Size

I have plans to write an article about an M12 sealed league I'm in at work. We buy a few boxes, dish out the packs, and for six weeks battle it out in a conference room over lunch. I was going to wait until the league was over before the full report but two weeks in I already have a theme I want to write about: losing. Me losing. Badly.

My match record right now is 0-4. In a sealed environment it's easy to attribute your success or failure to your pool. Sometimes a pro couldn't build a winning deck with the cards you get. But the thing is, I don't think my deck is that bad. It surely isn't great but I can't really complain too much with a Grave Titan and a Serra Angel.




So, I thought, maybe I'm terrible at Magic? I can't win a match! But why did I lose? Thinking back on it... is it possible I just got a string of bad luck? I remember drawing the all-powerful Grave Titan twice out of ten games (one of those times I watched him rot in my hand stuck on four lands the entire game). I remember facing off against Chandra all three games in one match and four fireballs across two games in another. I remember a top deck Sorin's Thirst copied by Chandra to take out my Serra Angel one turn before she would have won the match. I remember my opponent, out of cards and one turn away from death, drawing a Jace and milling my last 9 cards away. What, am I cursed?

Time for an experiment. Thanks to high quality random data from random.org, I simulated 100 matches for an average Magic player. There was an equal chance for 0 (loss) or 1 (win) meaning this player's average win ratio is 50%. In this particular set of 100 matches his average was actually 51%. The first match is in the top left, the second match is one square to right... left to right, to top bottom.

I highlighted stretches of four or more wins or losses in a row, and as you can see it doesn't take a voodoo curse to get four losses in a row. Look at that stretch of nine!

This image demonstrates it is a mistake to extrapolate sets of data with small sample sizes to reach broad conclusions. You shouldn't conclude you are a terrible Magic player because you lost four matches in a row. Similarly, you shouldn't think you're going to become a Pro Tour champion because you managed to win a Friday Night Magic.

You can make the same mistake when evaluating a deck. If you put a new deck together and lose three games in a row you cannot necessarily conclude the deck is bad. You have to look back at why you lost. It may be that you drew the wrong cards and your opponent drew the right ones. However, it's also possible that those games highlighted fundamental problems with the deck. This is the kind of information you're looking for. Understanding the games you play is much more important than remembering your win percent.

Remember those red and green streaks when you win or lose a few in a row! The last games we play are always freshest on our mind but don't let your recent win percent get in your way of focusing on what matters: understanding the game and playing well. Do these and you will win more over time.