The Riki Rules – Insufficient Randomization– err… Shuffling



Last week, I promised to bring you an article on Insufficient Shuffling, that is the infraction formerly known as Insufficient Randomization. Those are long infraction names to type out, so for the most part I will just abbreviate them as IS and IR in this article. The short version is that in addition to the name change, the penalty has gone down from a game loss to a warning. Let’s discuss why.


At first, I found it very odd that the name had changed to Insufficient Shuffling, mostly because the infraction was expanded to include partial deck randomization effects (read: cascade) that most definitely are not shuffle effects as far as the game (read: Psychogenic Probe) is concerned. Psychogenic Probe won’t trigger even if you miss on cascade and randomize your entire library; it still isn’t technically a shuffle. It’s one of those technicalities that judges like to get other judges with, even though we will likely never see it happen in a live game. No, not even in EDH.

When the new infraction name doesn’t make much sense from a strict game rules standpoint, it’s time to turn to the human element. As explained to me by the headstone of the policy pillar Toby Elliott himself (that sounds more morbid than I intended), the word randomization was a problem for many judges who would look at the end result of a deck.

Let’s say you deck check a player. Flipping through the deck you notice a pattern of land, spell, spell, land, a classic mana weave pattern. Except that you watched the player shuffle their deck, and they definitely did not do a mana weave unless they have David Copperfield-like powers. So what do you do with this perfectly stacked deck? Absolutely nothing. It is possible for a well-shuffled deck to end up in a mana weaved pattern.

This is where some judges would try to fall back upon the concept of Insufficient Randomization because look, the deck is clearly not randomized. This misses the point. The randomization that judges must look for is a process, not an end result. It’s been proven by mathemagics that the amount of shuffling that the average player does is not mathematically sufficient to produce true randomization, whatever that means. Judges, and by extension players, don’t care about true randomization. We’re not going to buy a bunch of casino shufflers for tournaments. Besides, I’ve heard that those things tear up cards like crazy, and you wouldn’t want to put your Baneslayer Angels through that.

The true desire is not for the deck to be perfectly randomized, but for players to make a good faith effort to randomize their decks, which is why IR focuses not on the end result but on what the players do. There are a few things that are flat out bad in this respect. The first is the face up riffle shuffle. Point blank, this is pointless. In the industry, we call it a “null operation.” If you do a hundred face up riffles, your deck must be close to mathematically randomized. However, a judge should hit you with IR (or IS now) every time because the true point of shuffling is to make it so that you and your opponent don’t know the positions of any of the cards. With face up riffles, you can see the bottom card of each half at the very least.

While it’s true that many players look up, up, and away from their deck when shuffling, this doesn’t guarantee that you can’t see those bottom cards. Try looking up while you shuffle a deck. Depending on the positioning, you can still see your deck with your peripheral vision. Ah, that tricky eyeball. What a marvelous piece of engineering. This is similar to the infraction Looking at Extra Cards. When I give this to people for flipping over a card from their library or starting to pick up two cards on their draw, their defense is often “But I didn’t actually see the card.” That’s great, and I hope players are being honest with me, but in terms of the infraction, we need to rule based on the possibility of the card being seen. If you riffle face up, it is just too damn likely that you can see those cards.

The pile shuffle is one of those nebulous actions. Just doing one pile shuffle isn’t sufficient because it isn’t so much a randomization as it is a reordering. If you know what the bottom card of your deck is when you start a pile, you are going to know exactly where that card ends up, in that very last pile whatever its position may be. And that amount of knowing goes against the spirit of shuffling. If you do ten pile shuffles, it’s still very possible that you can track the position of one or more cards, and indeed there are some methods of cheating that play off of this fact. I won’t get into them here, but suffice to say that piles alone don’t cut it.

However, this doesn’t mean that pile shuffles are completely worthless as some people seem to think. If you give your deck four or five quick riffles (or mashes, or side shuffles, or whatever new-fangled name you kids have for a shuffling method), it’s likely that you don’t know the position of any one card. Do a pile off of that and you’ve reordered a bunch of cards whose position you didn’t know. Sure, you can point to one of the piles and say “the top card of my library is on the bottom of that pile,” but if you didn’t know what the top card was going into the process, we’re good. Thus, ending with a pile shuffle is not a crime unto itself provided that sufficient shuffling happened beforehand.

(I will also briefly point out that mana weaving or any other kind of deck stacking is not only pointless, but could get you into very dark depths of trouble.)

At the most rudimentary level, the change to IR/S (acronym pun unintended) is good news all around for players and judges. Players don’t like getting game losses, and believe it or not, judges don’t like giving out game losses. The IR game loss was particularly vexing to judges because of its subjectivity. Off the top of your head, what would you consider a sufficient amount of shuffling? Put a number and a style to it. I’ll bet your definition is different from mine, which is different from the next reader’s, which is different from most everyone else.

Hopefully we can agree that one riffle is not sufficient. On the flip side, twenty riffles is. Between that is a lot of gray. Mix in multiple shuffling methods and things get grayer. Two piles and two riffles? My guess is that’s about the 50/50 point in the spectrum where you will get a lot of debate. So ten judges watch a player do two piles and two riffles. Five of them give the player a game loss while the other five see no problem and do nothing. Interesting.

Actually, the first five wouldn’t necessarily give the game loss, which is where the infraction ran into even more problems. Judges were afraid to assess the penalty even if they believed they witnessed insufficient randomization. Some might hem and haw and end up doing nothing, while others would step in and tell the player to shuffle a little bit more. Now if you’re asking a player to shuffle a bit more, wouldn’t that mean that you don’t think what they just did was enough? Perhaps we might even call what they did”¦ insufficient.

I actually had one instance where I gave a game loss for IR (I think it was something like two riffles after a fetchland or similar), and not only had it appealed and overturned by the Head Judge, but the HJ told the player to shuffle some more. I was confused. Point of fact, I have never issued an IR since that day because frankly I didn’t know what to do anymore.

Having a lesser penalty will probably lead to judges calling it more often now, thus raising awareness that such an infraction exists, and in turn leading to players calling it more often on their opponents. I don’t think it will lead to any kind of wholesale change in judge-calling behavior. Heck, I’ve been trying to get you guys to call us over for far simpler things. I’m guessing that we might see one or two more instances of it per PTQ, and that seems about right. Where it will really save us some headaches is towards the end of the day, especially in a Top 8 when players are tired of all the shuffling. A player picks up his deck, searches for a land, riffles once and presents. At the SCG event in Dallas, this resulted in a controversial (but absolutely correct) game loss in the Top 8 even though it was highly unlikely that the player was trying to pull something shady in front of a crowd and on camera. Now we can say “Hey, I know you’re tired, but we still need you to shuffle a bit more. Here’s your warning. Thanks.” Seems better.

Finally, this change to IR/S has been part of a downward trend in penalties. Two years ago, Outside Assistance was dropped from a DQ to a match loss (and was more recently loosened yet again by allowing sideboarding notes between games). Drawing Extra Cards is still a game loss, but a vast majority of cases have been moved into the realm of Game Rule Violations, warnings.

Marked Cards – Pattern (game loss) and No Pattern (warning) have been collapsed into one infraction, unsurprisingly called Marked Cards (warning). This change was made for a similar reason as the IS change; judges (like me) were spending too much time debating whether 4 [card]Brainstorm[/card]s and 3 Islands constituted a “pattern.” There is still the option to upgrade MC to a game loss in cases where there’s some potential for abuse if the player were to notice the markings, but overall I’m guessing we’ll see fewer of those than the old Pattern game loss.

It seems like the IPG is losing its teeth. In the short term, this seems like a good thing all around, and it speaks to the health of the game and clean play that is going around. But is this kind of thing cyclical? I had planned on waxing nostalgic about the Wild West days and speculating a bit about how things might return to such a state if we grow too lax, but it seems I’ve run out of time for this week.

Rikipedia at Gmail dot shuffle
Risky on efnet and most major Magic forums
Japjedi47 on AIM


Scroll to Top