How to ‘Win’ a Tournament

Good evening Duelyst Fam! I was sitting down to write an article this evening, and I rolled Abyssian (written a lot of those recently), and re-rolled it into Songhai, which you sort of know my opinion on. So I started thinking, “hey instead of writing another deck tech article why don’t you give the people what they need now that these exciting weekends are coming up?” In my career as a magic player I have made the top eight more times than I can remember. When I started playing Duelyst I kept count of my top 8 finishes, 16 out of 19 events (not counting single elimination events). Alright here it is, and pay close attention because I’m about to tell you how to become a boss so that you can go full Marcus Brutus on me one day.

The first and most important lesson is to know what you are competing for. You noticed how I specifically talked about top 8 earlier? Why was that? Surely I might instead talk about the times I won? Only five! But GGH, five wins out of 16 appearances is not a very good track record… Oh, but I made money! You see the goal for me, my focus was to top 8. I didn’t necessarily need to win to make it worth my while. In fact I couldn’t have cared less about my seating going into the finals. If there was money for Top 8 that was the goal, if there was money for top 4 then that was the goal. I knew what I wanted and I shot for that. In a different tournament my goal might be to get 1st place, this is the goal most people have every tournament, but it is very different from what I was trying to do in most cases.

Lets talk about why your goal might be to Top 8 instead of win, and how this affects what you are going to do. In full disclosure because of some of the things I and others did tournament organizers have changed rules, I will still talk about all the things you can do because if they are not specifically stated in the rule book it is to your advantage to know them. If your goal is to top 8 then you are probably playing a swiss bracket. Normally there is a hard cut for top 8 followed by either a finals bracket or a day 2. Either way some important things happen here, you have to make the cut, and your tie breakers are based on the people you played(you have no control over this), and your game win-loss record(you have some control over this). If your game win-loss record is important than it is critical that you win 2-0 as many times as possible in the swiss rounds. Sparta once commented that I must be really lucky because I broke into top 8 on tie breakers a lot and rarely got cut on the breakers. Going 2-0 is partly good skill but also partly careful planning. There are plenty of decks I consider to be S-rank decks at any given time, but which one(s) did I decide to bring? When you are going to play with the sole intention of getting into top 8 you want to play “Mr. Consistency” the deck whose baseline(μ) game win percentage is very high. The deck that doesn’t rely a lot on chance and whose game to game variance is lower. Again, it is because in a swiss tournament you want your game win percentage to be as high as possible because your game win percentage might take you into or out of the top 8. You might be thinking, “well yeah GGH we know we want consistent decks,” hold on to that thought for later because I am going to convince you that wanting a consistent deck is not always the case.

Before we get into when you want a higher variance deck lets talk about ties. In Swiss tournaments you need to make sure you look for opportunities to take a draw in the last round. This requires some math magic, but an astounding amount of the time if you draw in the last round you can force you and your opponent into the top 8. A lot of guys think this can only happen if they have already won out. That isn’t true and I would rough guess that two out of three times you and your opponent can force yourself into top 8 if you draw the last round. For whatever reason tournament organizers in Duelyst (and Duelyst only) have decided they don’t like this. In one instance they even changed the rules on myself and Kolos mid-tournament (still salty). To check and see if a draw would be mutually beneficial for you and your opponent; in the last round of the tournament go look at the current standings to see how many people are competing for the top 8 by points and where you currently stand. If there are 14 dudes or less competing for the top 8 and you are currently ranked in the top 2 you can for sure take a draw and force yourself in. To be fair to your opponent they would also need to be in the top 2. Now there is zero risk you missing the cut to Top 8, personally I say congratulations on doing well enough to make top 8 by taking a draw in the last round. It gets easy to count backwards from those first two numbers, 2 and 14.  If you are in the Top 4 you need 12 or less, in the top 6 you need 10 or less, and in the top 8 you need 8 or less. Why would you do this? Lets pretend for a second that you are 5-2 and ranked #8. Why are you number 8, which is the bottom of the 5-2 bracket? Because one of the dudes you lost too isn’t doing well. You lost to Joe Blow the rando in round #1. The current #9 is 4-3 so you are out scoring him, but why is he the top of the 4-3 bracket? Because he lost to the #1, #2, and someone else in the top 8. If you lose this next round and he wins this next round you will both be 5-2 but his tiebreakers are better (Joe blow the rando probably dropped out 3 matches after beating you in round 1). If you take a tie both you and your opponent will finish 5-2-1 making both of you safe from the #9, who even if he wins can no longer catch your record. This is a pretty standard tournament practice in everything from Magic to 40K. It is mutually beneficial for both you and your opponent, and in other games it helps prevents intentional draws from being forced in match. Don’t message me about this, you will only make me angry. I know “why” they force people to play it out in Duelyst, I simply disagree emphatically.


Bracket Clean Sweep
We’ve talked about the strategy of Swiss and playing to top 8 instead of playing to win. What are some instances when you want to play to win and how does your strategy change? When we are playing to win game win-loss record no longer matters. This could mean we are playing a single or double elimination tournament or a tournament were 2nd place is functionally the same to us as dead last. This second example is more often the case in qualifying tournaments which is what the bracket I featured above is an image of. When we are playing a tournament where winning is all that matters we are still interested in an S-rank deck with a high overall win percentage but now we are interested in the S-rank deck with the highest variance(ξ). Why are we interested in that high variance? Because we don’t care if we lose a game as long as we win the match. If the best case scenario is nearly unbeatable we are likely to get some amount of free wins, and of course we will have some games were we hit the low end and lose some games we may have otherwise won. The average becomes the same as other S-ranked decks but something important happens here, the number of games we actually “play” is significantly reduced. We are changing the game from a best of 3, because neither player actually “played” during the very low end or the very “high” end of your decks performance. I want to make perfectly clear I am talking about an S-ranked deck that has a 60% or higher overall win percentage, but its best case gets into the 90’s while its worse case might be as low as the 20’s. We don’t want to try something that is 100% on the omega(Ω) but that high end is extremely unlikely, and the deck sucks outside of the high end. I feel like you can start to see why game win-loss either being a factor or not can have a big impact on what you should play, but I want to talk a little bit about double elimination specifically and while it still plays into the “must win scenario” instead of “okay to be consistent.” For the most part an X-2 record is borderline top 8 even in smaller swiss tournaments and in those tournaments we still want to play to Top 8, so why is a double elimination which essentially boils down to a smaller swiss bracket still a “play to win” scenario? We need to examine a small 16 man double elimination bracket to answer that question. If you are the very last dude to drop from the winners bracket you are going to have to win your next three rounds in a row to beat the guy who pushed you out. 3-0 is not an easy record to achieve in any format.  Comparatively the guy who beat you can go 1-1 or 1-0. Lets assume for a moment that you were 10% better than your opponent (an insanely significant skill margin). And your deck is favored 60% to 40%. So collectively you have a 70% chance of winning any given game. For simplicity sake we will assume the same about the guy you will play  in the losers bracket. Even under these outrageously favorable conditions you are only 48.17% likely to win the tournament by going 3-0 after your initial loss. For you math guys you can track my model as a 3-0 record assuming each match is best of 3.  Hopefully you can see the case that I am making that essentially a double elimination bracket is a single elimination bracket with a lottery ticket attached. In the end you want to play decks that will occasionally give you an unfair advantage so that you do not have to “play” those games. To increase the likelihood of winning the tournament we want to decrease the number of games we need to actually play.


Guys this was a long article even after I trimmed out a bunch of personal accounts and I hope you found it helpful. If people are interested I may make a second article talking about some of the non-play based mistakes I have seen made made or made myself at a later date. At the moment I am struggling to sort out what of the dozens of small things deserve a mention. It is hard to know what knowledge other people have, and I don’t want to accidentally slip into “mansplannin” tournaments to you. I’ve got too much respect for the intellect level of the Duelyst family to do that.

As always much love and I’ll see you on the field,




Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.