fbpx

Double or Nothing: Breaking the Tie

At some point, you’ve probably been at a Magic tournament staring at a list of standings: Maybe you’ve won all your games, but for some reason you’re not number one. You ask around and a helpful soul takes a look:”Your tiebreakers are awful!” You’re left staring at rows and rows of decimals; somehow they determine your position… But how?

At some point, you’ve probably been at a Magic tournament staring at a list of standings: Maybe you’ve won all your games, but for some reason you’re not number one. Maybe you and your friends have all won one game and lost two – but for some reason you haven’t been told, they’re all ahead of you! You ask around and a helpful soul takes a look.


“It’s your tiebreakers! They’re awful. You might as well drop.”


They wander off and you’re left staring at rows and rows of decimals; somehow they determine your position… But how? Why are they awful? Whose fault is it? Why should you drop?


In this article, I hope to demystify the myths.


Most people know that if they’re playing Magic in a tournament, they get three points for a win, one point for a draw, and no points if you lose. Even so, in large tournaments you’d still have a lot of people on the same number of points (match points) and so the DCI came up with a way to rank them: tiebreakers*.


There are three tiebreakers:


Opponent’s match-win percentages.

This tells you how good your opponents are. If they’re beating other players throughout the day, it goes up and up; if they lose, it goes down. You work it out by dividing the number of points each opponent has by three times the number of matches they’ve played and adding them all up. Then you divide that number by the number of your opponents.


Example: Jim has played Chris, Bob and Tarik so far in the day.


  • Chris is 2-1, having beaten Bob and Tarik but lost to Jim.

  • Bob is 0-3, having lost to Jim, Chris and Tarik.

  • Tarik is 1-2, having lost to Chris and Jim, but beating Bob.

So; Chris has 6 points, Bob has 0 points, and Tarik has 3 points.


We start by dividing all of these by three times the number of matches they’ve played. Chris = 6/(3*3) = 6/9. Bob = 0. Tarik = 3/(3*3) = 3/9. Next, we add these all up: 6/9 + 0 + 3/9 = 1, and then divide that number by the number of Jim’s opponent’s.


Jim’s Opponent match-win percentages is only 1/3 or 33%


Is 0.33 any good? No.


On average, Jim’s opponents are only winning one match out of every three they play. On the other hand, Bob has played Jim, Chris and Tarik and they’ve done okay – with Jim winning all of his matches – so Bob’s percentage is (9/(3*3) + 6/(3*3) + 3/(3*3)) / 3) = 0.67 or 67%. That’s because Bob has played better players than Jim has.


Opponent’s match-win percentage is the first tiebreaker to be used after match points; that way, the players who have played hardest matches against better players and won are rewarded, whilst the players who played and beat bad players can still win… But their tiebreakers won’t be as good.


Game-win percentages.

After looking at how many matches you have won and how good your opponents are, the next tiebreaker tells us how easily you have won games, and how much of a fight you put up if you lost.


Game-win percentages are worked out from the number of game points you win. Like match points, you get three points for each game you win and one point for each game you draw. So if I beat my opponent 2-0, I get six points. If I lose to him 1-2, I still get three points! If I draw 0-0, I’ll get a point, but if we draw 1-1, I’ll get four points.


We take all the points you win during the matches you’ve played so far and add them all up, then divide that number by the number of games we’ve played, multiplied by three.


Example: Jim has won all three of his matches: He beat Bob 2-1, Tarik 2-1, and Chris 2-0. Jim will have a Game-win percentage of ( 6 + 6 + 6 ) / ( ( 3 + 3 + 2 ) *3 ) = 0.75 or 75%. Jim is winning three out of every four games he plays. Not bad!


If Jim’s opponents put up less of a fight (say, Bob and Tarik both lost 2-0) Jim’s win ratio would rise to 100%. If Chris had put up more of a fight, Jim’s ratio would have dropped to 67%.


Game-win percentages show us how well we’re doing: The closer to one it is, the easier we’re beating our opponents. The lower it is, the closer our games are.


Opponent’s game-win percentages.

The last tiebreaker shows how much of a fight our opponents are putting up. Typically, the higher their match-win percentages are, the higher their game-win percentages are, too. We add up all of our opponents’ game-win percentages and divide it by the number of opponents we’ve faced.


The higher it is, the more of a fight your opponents are putting up. The lower it is, the more they are losing 0-2.


What does all of this mean? Well, if you win all of your matches, you’ll probably make top eight unless you’re in a really big tournament anyway. However, if you lose a few, you’ll probably need to be at the top at the group of people who have the same number of points as you… And that comes down to your tiebreakers.


It is very rare indeed for the Opponent game-win percentages to be used. DCI reporter, the program most large tournaments use for keeping track of results, keeps percentages to three decimal places!


What do Byes do to all of this.

You can have up to three byes in a Grand Prix, or sometimes you’ll get a bye at random if you’re the odd player out in the first round, or the lowest-ranked player. Byes affect you in two ways:


Firstly, whenever you have a bye, you effectively win 2-0. That means your Game-win percentages go up by as much as they can that round.


Secondly, any Byes you have aren’t counted when working out your opponent match-win and game-win percentages. That means that, with three byes, your very worst opponents will still have won at least three matches and so your match-win percentages – the most important of your tiebreakers – will be higher than someone who had to fight their way through the first three rounds.


Can you ID in?


The last topic I’d like to look at is Intentional Drawing, or IDing, to get into the top eight – quite often, near the end of a large tournament, you may not have to actually play your last round before the cut to the Top 8 if you have enough wins. If you have enough match points, you can agree to a draw with your opponent; each of you will get a single match point from the draw, which in turn may guarantee you a berth in the Top 8 depending on how the tiebreakers and match points pan out.


And yes, this is totally legal and allowed in the official rules.


If you’ve won all your matches, you’re there automatically… But if you’re 5-1 or even 5-1-1, there are two things you need to look at: Match points and your opponents’ match-win percentages. Have a look at the table below:












































































Name


Pts


OMW


GW


OGW


Roy Williams


19


62.58


68.42


60.79


Jim Grimmett


16


61.90


75.00


56.36


Madog Williams


16


53.06


70.59


52.41


Richard Williams


16


51.02


73.33


47.87


Nat James


15


59.86


61.11


56.05


Manveer Samra


15


58.50


66.67


54.14


Ian Roberts


15


51.02


62.50


49.32


Anwar Tarafder


15


44.22


63.16


44.63


Neil Pearce


13


59.86


55.56


61.28


Mark Knight


13


58.40


64.29


56.57


Shane Silk-Reeves


12


52.38


56.25


53.41

 


This table is a snapshot from a tournament after round seven. Who can afford to ID to make the top eight if there is one more round to go? Who is certainly in the top eight?


There are only three people who can catch Roy, as he’s at 19 points; even if all of the 16-point players played against 15-point players and won – which is unlikely – at the end of the round, the four highest-ranked players would have 19 points each. One of those players, naturally, would be Roy. Roy will make the top eight no matter what happens, so he can afford to win, lose, or draw.


Jim, Madog and Richard are all at 16 points. If they lose a match, some of the players on 15 points might win and overtake them – but there are only three 15-point players, and one of them is guaranteed to lose, so there aren’t enough players to knock all three of the 16-point players out of the top eight.


Even so, if five players somehow manage to catch up on 16 points, someone will have to go… And that’s where your opponents’ match-win percentages come in.


I’m at the top of the sixteen point players, and so maybe I’m safe. After all – I have a first tie-breaker of nearly 62%!


But am I safe? What if all of my previous opponents lose, and I lose to my opponent?


Let’s have a look. It’s round 8, so I’ll have 8 opponents by the end of it… And if they all lose their last games, my opponents’ match-win percentage would slip to about 55%! If half of them win, I only drop down to about 60%.


A fair guess is that during the last few rounds of a tournament your match-win percentage won’t change by much more than 5% or so.: If your percentage is very high you’ve played good players who won’t lose. If you’d played bad players, they probably won’t win, and if you have a middle of the road percentage, some will win and some will lose – keeping it pretty much the same. So if even if my opponents all do really badly, I’ll slip to around 55% – which would probably keep me in the top eight.


Even so, Roy, Jim, Madog, and Richard would probably all ID if they could – as with at least 17 points and only four people who have a chance of overtaking them, they’d certainly be in.


The real dilemma comes for the 15-point players. If they draw, they’ll have 16 points each… And although only two 13-point players could get up to 16 points with a three-point win, those two players have quite high opponent match-win percentages. If they win and Anwar and Ian decide to draw, then all four of them will end up with 16 match points, but the people who won will have higher tiebreakers – meaning that in all likelihood, Anwar and Ian will find themselves in ninth and tenth places.


At the end of the day, IDing can be quite simple. If IDing will keep you in the clear thanks to match-points – as it will with Jim, Madog, and Richard above – then you’re safe. If IDing means you’re going to have to rely on your tiebreakers, make sure you’re at the top of the group of people who have the same points that you do! Check you’ll get in even if your results drop by 5%; otherwise, play for it.


Hopefully, tiebreakers aren’t so mysterious now: They show you how good your opponents are and how well you’re doing with a three simple numbers. Byes give you a free win and help your tiebreakers out and IDing…


Well, IDing isn’t for the faint hearted when it comes to tiebreakers – just try to win your way in if you can.


Cheers, Jim Grimmett

Team Phatbeats,

Team Diaspora &

Level 2 DCI judge.


* – If you’re looking for the DCI’s official tiebreakers document, look no further! You can find it here.