facebook

CoolStuffInc.com

Darkest Night Sale ends Sunday!
   Sign In
Create Account

Why You Need to Stop Using this Method to Choose Who Goes First

Reddit

Most matches of Magic start with the same phrase: "high roll okay?"

Just like rolling out the playmat and (hopefully) wishing your opponent luck, it's part of the ritual that accompanies the start of a game of Magic.

Well, "high roll" is part of my ritual. In the Magic community, there's a group of players that prefer another way of starting their games: "Odds or evens?"

For proponents of Odds/Evens, the advantage is that it is more efficient than High Roll. (If you're unfamiliar with Odds/Evens, read "Deciding Who Goes First" for an explanation). High Roll requires two rolls to start with, one for each player, and has the possibility of more re-rolls if both players roll the same number. In tournament Magic, where rounds are timed, shaving off a bit of time can stop a round from going to turns or buy you a few extra seconds to grab a bite before the next round.

And proponents of Odds/Evens are completely right about this efficiency. It is more efficient to use. But there's a problem. While Odds/Evens is a quicker way to start a match, it's also easier to cheat with because Odds/Evens is more susceptible to manipulation via weighted dice than High Roll is.

How to Cheat with Odds/Evens

In High Roll, players generally roll with the same pair of dice. This means it doesn't matter if the dice aren't perfectly balanced (and outside of a casino, there's a good chance they aren't) since any unfairness is canceled out by the fact that both players used the same dice and received the same amount of bias.

But in Odds/Evens, this doesn't happen since the dice are only rolled a single time.

Now, you might ask, "if the dice are weighted, shouldn't it not matter because the player who rolls the dice doesn't know if the other player is going to choose odd or even?"

The first problem is that humans are lousy at choosing things randomly. Given that people choose heads over tails 80% of the time in a coin flip and profess that seven is their favorite lucky number, it's naive to expect that players are just as likely to pick odd as they are even. To exploit this, a savvy cheater can simply note down whether people answer odd or even more and, once they've detected a pattern, buy dice that disfavors the chosen response.

But it's not even necessary to do that because you can cheat in a way that will allow you to win the roll regardless of if your opponent chooses odd or even.

How Would a Cheater Do This?

To start, they would go to their search engine, look for a pair of D20 loaded dice, and then navigate to one of the many sites that sell such things.

They would be looking for two dice in particular: one weighted to "high roll" (i.e., hit twenty) and the other weighted to "low roll" (i.e., hit one, less common but also available). They would make sure to pick dice that have evens clustered around the twenty and odds clustered around the ones to increase their chances of hitting the desired odd or even. I found a beautiful "high roll" one after about five minutes of searching on a popular eCommerce site.

Then, the cheater goes to their local tournament. They place the weighted dice in a bag along with an assortment of other dice - picture a purple Crown Royal bag with about ten other dice in it - and when they sit down to their match, they dump the lot onto their playmat.

Proactively, they bring up the topic of who goes first. "Odds or evens, OK?" the cheater says while shuffling their deck.

Whatever the opponent replies, the cheater then reaches into their pile and chooses the dice that will ensure their opponent loses the roll.

And their opponent is none the wiser because they were unaware, just like anyone else who has not read this article, that Odds/Evens could be manipulated in this way.

It might sound far-fetched that someone would go to all this trouble, but is it really more preposterous than a Hall of Fame player marking their cards to help find Tron or the numerous examples of players being dumb enough to cheat on camera? The weighted D20 dice I located had over 400 reviews (five stars, of course).

How Useful is This Cheating?

And worse, this style of cheating can produce a notable edge across the length of a tournament.

Imagine that an honest player enters their six-round local RCQ with a deck that is much more favored to win on the play than the draw - let's say Mono-Red Aggro.

If this deck is 60% to win a game on the play and 40% to win on the draw, at the start of the tournament we expect the player to emerge with a 50% match win rate or a 3-3 record because, since they are honest, they are just as likely to be on the play as on the draw.

But what if they used the methods above and, for the sake of argument, know they can win 90% of their flips?

As I show in this spreadsheet, this would boost their expected match win rate from 50% to 54.2%. This is equal to about a quarter of a match. That doesn't sound like too much, but as I write this, 54.9% is the win rate over the last few months of the deck with the highest win rate in the Pioneer format. So, using this cheat is roughly as effective as upgrading from an average deck in a format to the best - all at the cost of some dice.

What if we make the assumptions more aggressive? Let's say you find a deck that's 70% on the play and 30% on the draw and you perfect this cheat so that you now win 95% of your rolls. In that case, you can expect your win rate to climb from 50% to an astounding 60.4%. And that seems hard to sneeze at.

Solving the Problem for Good

Overall, I'm not sure if anyone is exploiting this weakness today - if they are, they probably aren't posting about it on the internet. But I think the fact that this could occur is enough to question its use. Personally, I'll always ask to use High Roll if you sit down across from me at a tournament.

However, I think the real answer is to get rid of players using dice to determine Play/Draw in the first place. We don't have players manually rolling dice in MTGO or Arena, so why do we do it in paper? Why can't tournament organizers be given the option to have the Companion App randomly determine who gets the choice of Play/Draw and send that along with pairings?

Let's move out of the dark ages - more efficient tournaments and more players confident that the game they're sitting down to play is off to an honest start.

- Richard 'Char' Galluzzi

Turn your cards and minis into cash! Maximize your value with our 25% store credit bonus!

Sell your cards and minis 25% credit bonus