fbpx

Analytical Failures

In this potpourri of important studies, Ari Lax combs Magic and its community for signs of improvement! See how a Pro Tour Champion reviews his own thoughts and analysis, as well as the greater Magic world!

Being wrong happens. In Magic, it happens a lot.

The best thing to do is figure out and discuss why.

Here are a few current examples of people being wrong, mostly myself.

Fate Reforged Draft

While I still haven’t drafted the set that much (or in more accurate terms, at all), I’ve noticed a bit of a flaw with my overall analysis last week. While the format still consists of a significant
number of the cards that trump basically the entire set of Fate Reforged, the overlap is killer.

I’ve talked a lot about drafting Five Color decks in Khans and the defining
characteristics that make this good. Overall, the deck has three required components: defensive creatures, powerful incentives, and fixing to all the top
notch cards you can get.

Fixing? Check. There are slightly less lands per pack than there were in Khans, but it’s not significantly so.

Defensive creatures? Outside of black there really aren’t a lot. No Monastery Flocks. No Sage-Eye Harriers. Not even a Bloodfire Mentor.

What about incentive cards? This is where Fate Reforged really falls through. There aren’t multicolored rares or uncommons. There’s a ton of great rares
and mythics, but there isn’t the same cost for picking them early. Someone might pass a Butcher of the Horde or a Charm because they don’t want to pick a
three-color card that early, but in Fate Reforged, the furthest a multicolored card goes is two colors. Even though the legendary dragons are allied
colors, and as a result, are only double on color in a specific clan, every other clan only needs a fixer or two of a single color to splash them. Compare
to something like Siege Rhino, where a Jeskai deck splashing it could be at best described as “undisciplined,” and more accurately described as “loose as
hell.”

So defensive, greedy decks are going to be strained on filling their important quotas in the later packs. Maybe you can say people in pack one of Khans
Draft are more open to picking up high-powered multicolored cards and it’s as big of a loss as I’m making it out to be, but the defensive creatures were
all almost incidental picks. You could always find more if you needed them. Now there just aren’t three packs where you get to say this, and those picks
will overlap with cards like Abomination of Gudul. Even if you hit all the numbers, your average card quality is going to drop. When you draw Sandsteppe
Outcast on turn 8 instead of War Behemoth, it’s a big difference.

You can adjust, play more to optimize your cheaper cards. But it is an adjustment, and I don’t really know how this changes the archetype’s positioning
without a lot more experience.

Valorous Stance

I feel really dumb.

How did I not see it?

Everyone just kept focusing on a one-use case, and I did too.

Valorous Stance in Heroic. Valorous Stance protects your Hero of Iroas and kills a Siege Rhino! Wow! Pow! Bam!

Wait a second. Valorous Stance doesn’t stop Abzan Charm. Or Utter End. Or Banishing Light. Or Perilous Vault. Or Chained to the Rocks. It doesn’t get
through blockers the way Feat of Resistance or Gods Willing does. We already have Ajani’s Presence, and it doesn’t see play for a reason.

For the bonus of being able to kill Siege Rhinos when they can race our creatures in the mirror, you are losing raw linear power. I’m generally a fan of
making my highly committed linear decks as good at what they do as possible rather than diluting them, and I didn’t really see this as a huge upgrade.

So I dismissed the card. It was a conditional removal spell when bigger decks wanted something unconditional. Except it’s still a two mana removal spell
that’s actually good and reasonably flexible. And those just don’t exist in this Standard format.

I really like Valorous Stance alongside Goblin Rabblemaster, but going down that road brings you to Chained to the Rocks as a likely better version of the
same effect. That card does create certain issues with your manabase, and it’s possible the deck is better with a slightly worse removal spell and better
mana for Wingmate Roc and other similar cards.

Of course, there’s another deck that might want this cheaper removal spell that hasn’t had access to one and also features two new Fate Reforged cards.

Last year we saw a land-light White Weenie deck pop up from time to time that was actually really reasonable in certain metagames. It was heavily based on
the concept of multiple early one-drops into an anthem effect, either Spear of Heliod or Hall of Triumph, putting a lot of power into play very fast.

The creatures this year are sized better to fight this strategy, and the loss of Brave the Elements is a big deal, but these are certainly a lot of steps
in the right direction. More one-drops and cheap removal for their larger creatures that block your two and three-power attackers.

I don’t think the time is quite right for this deck, but if another good anthem like Honor of the Pure or even a good pump spell like Fortify pops up, I
would not be shocked to see this deck pop up from time to time. It’s also worth noting for future reference that this deck gets hit by a lot of the same
hate as the Hordeling Outburst decks, and is definitely a metagame call.

Also note that both of these creatures are Warriors. I’m unsure the mana in the format really supports an Orzhov Warriors deck, but that deck really wants
the lower curve and cheap removal that Fate Reforged offers. One bit of advice: don’t play four Obelisk of Urd. Flooding on that card is the worst, and I’m
almost inclined to say it’s worse than Sorin, Solemn Visitor as it’s not an actual threat.

Don’t Play Red Devotion

People see these cards, and they immediately build something with four Nykthos, Shrine to Nyx. It makes sense, right? Two new double red spells to cover
the losses of Frostburn Weird and Boros Reckoner last year. Mono-Red Devotion was always close to good then, and in a format with worse removal it has to
be better.

Right?

No, actually you would be wrong.

Red Devotion now has all of the costs, but none of the payoffs.

Red Devotion was always a highly inconsistent deck. This shouldn’t come as a shock when you look at the shell. Four of a legendary land, 25 land, and
twelve two-drops that don’t sink mana; four colorless lands in a deck needing triple red on turn 3, and so on. In return, you got raw power. You had draws
that ended up with ten power on turn 3, or that attacked with multiple Stormbreath Dragons on turn 4.

There is actually no way to do that with the cards that exist in Standard this year. Burning-Tree Emissary was the key to all of this. Without Emissary,
you can’t use Nykthos to significantly jump the curve into five and six-drops the way you could before.

If you play a turn 2 Mardu Scout and a turn 3 Flamewake Phoenix, your Nykthos is plus a mana on turn 4. That means you can turn 4 a Stormbreath Dragon.
Woo. You can also play Rattleclaw Mystic, Ashcloud Phoenix, and Stormbreath Dragon on the same curve to basically the same effect without playing worse
cards and mana just to hit Red Devotion pips.

But a Devotion deck has seen some success this year. Green Devotion has a number of notable finishes,

most recently a top 4 at Grand Prix Yokohama in the hands of Hall of Famer Makahito Mihara.

What’s the difference?

Green Devotion uses Nykthos in a different way. It isn’t trying to hit five, six, or seven mana. Green Devotion is using it to hit fifteen, sixteen, or
twenty-seven mana. You are using it to spiral your boardstate out of control.

The key in this case is some kind of way to keep the wheels turning. Eidolon of Blossoms this year, Garruk, Caller of Beasts last year. Even something as
simple as Polukranos, as red just doesn’t have that. It also doesn’t even have something along the lines of Hornet Queen: a single high end, near
unbeatable threat to ramp into. And it doesn’t have back up ramp, so even if it did have something massive, it would clunk up hands more than it would win
games.

Maybe these cards make Purphoros a playable threat, similar to how Thassa was excellent last year. Maybe they bring back Fanatic of Mogis. But the cost of
being a dedicated Nykthos deck is much higher than the reward here.

On this note, there are a couple interesting devotion enablers in other colors. I don’t think any archetype has quite enough to hit critical mass just yet,
but Cloudform is a notable blue creature that plays very well with Thassa, God of the Sea and Master of Waves. It plays especially well with Thassa, where
you have an indestructible threat enabled by hexproof Devotion pips.

I also may have missed another interaction in one of my earlier evaluations via false comparisons. I was so hung up on Palace Siege being a worse card
advantage engine than Liliana Vess that I failed to realize it does something special and unique: loop Gray Merchant of Asphodel. While you can chain
Merchants with Liliana Vess, the less vulnerable enchantment interacts much better with devotion counting. While there isn’t currently a good way to keep
putting your Gray Merchant into the graveyard, it’s very possible that one comes out in the next two sets. Kheru Bloodsucker is definitely on the way too
cute side of things.

De-incentivizing Collusion

This one is a bit of the reverse of the above situations. I have a proposed solution to an issue, and I want feedback on why it will or won’t work. Poke
any holes in my logic, find the scenarios where it fails.

A week or two ago there was some discussion of collusion after the disqualification of a known player from a PTQ for what boils down to saying the wrong
thing in the relatively common scenario of a pair down in the last round; one where one player is locked for top 8 with a win, and the other can’t possibly
make it. They were extremely shocked because they didn’t feel anything they said really crossed a line in terms of collusion (regardless of their intent).

The gray area here is a big issue. There’s a lot of collusion that isn’t technically collusion occurring, and there’s a lot of people attempting to
technically not collude who accidentally do. These kinds of operations really shouldn’t be happening, but it mostly exists because the current dance is
objectively easier than figuring out an answer.

I’m going to start with the fundamentals of why this issue exists. You can discuss it and offer solutions all you want, but unless you address the
root cause you aren’t getting anywhere.

Some easy ones to lay down the basics:

Players have to be able to concede at any time they want for the game of Magic to function. If someone decides they are done, not allowing them to be done
just doesn’t work. Imagine the world where concession isn’t allowed. Oh man, I have to leave. Sorry, there’s a penalty for that. Oh man, I don’t want to
have this game 1 drag on forever and not finish a game 2 and lose the match. Nope, the control player just gets to naturally deck you with Elixir of
Immortality. Not a good situation.

Many Magic event prizes aren’t really tied to a specific winner, and even the ones that are have some cash equivalent. Money can be exchanged for goods and
services!

The way Magic tournaments currently exist, you end up in matches where one result ends up creating more net value than the other by a large margin. If the
parties cooperate, there can easily be a way for both of them to end up with more than their expected value of playing the match.

You can’t go Big Brother and police players’ actions once they are away from the event. Player B scoops to Player A, the judge of the event doesn’t have
access to the fact that Player B then Facebook messages their Paypal address to Player A who then sends them some money.

When a collusion or bribery transaction completes, both parties involved are benefitted. This is contrary to cheating, which is zero-sum. As long as the
people colluding don’t make a mistake and give away their intent to a third party, they are both incentivized to deny it happened, at which point nothing
can really be proved.

We’ve proceeded down the “punish the people stupid enough to be caught” road for a long time, but that doesn’t seem to be removing collusion from the game.

If you push further, it gets real ugly. A player asks “Would you like to scoop?”, ends the conversation at that, and gets one. They pay the other person.
Is their phrase “Would you like to scoop?” assumed to imply “I will pay you to scoop” if this becomes public knowledge? You can’t really do that. So we
have a lot of people who understand the implications playing along, and people who don’t not being in.

Aside: I do think this is better than the in-crowd of a world where these incentives were actually impossible, where you just have friends/teammates
scooping to friends/teammates. The only thing required for this in-crowd is literally playing the same way.

This world is also non-ideal, as it creates a lot of awkward scenarios when people do scoop. Don’t compensate appropriately, or in one story I know, don’t
compensate at all? People will likely be reasonable and bring it up, but if you flat out hold your line it, will be seen as you being a non-cooperative
person and being denied the benefits of being assumed to be cooperative. It’s a prisoner’s dilemma where everyone votes to not defect unless they know
their opponent has a history of defecting. This person is now the scumbag for not supporting this system, and honestly for all parties involved, this is
the purely logical right play.

I feel like I’ve clearly demonstrated that as long as matches with mismatched incentives exist, collusion will exist and be virtually undetectable.

So the key is removing the scenarios that create the incentive.

Collusion commonly occurs in two scenarios:

  • Player A is live for a greater prize. Player B is not.
  • There is a prize that only matters to Player A and not Player B, such as Pro Points.

In the second scenario, almost all of the time the transaction is irrelevant to everyone else in the event. It’s the last round of a Grand Prix, everyone
is paired by tiebreakers, the winner goes X-3 into 14th place and the loser goes X-4 into 56th regardless of who wins (barring tiebreaker fluctuations that
are basically a purely random, independent variable). If Player A gets the extra two Pro Points, it doesn’t detract from anyone else’s Pro Level or finish
in the event. Sure, maybe if the match result was random, that person would get less points, but odds are this would be universal and levels would be
shifted down to adjust. It’s completely random noise in the long term. This occurring is also basically random over a season, so it happening to one person
in line for the World Championship versus another is basically another random variable along the lines of getting a good matchup, your opponent not showing
up, etc. You could find some strange alternative to the current Pro Club that makes these matches not exist, but that’s likely more work that it’s worth to
solve an issue that doesn’t seem to cause actual harm.

The ones that people hate are from scenario 1. If the match was random, a different player could be in top 8. That’s an insanely huge impact.

But the reverse happens with pair ups and people getting draws or concessions the next match down can’t, and that’s a fine inherent part of the system. The
issue relative to the Pro Point scenario is the person rewarded isn’t random, and it’s backwards of how things should be incentivized. The person who gets
paired down has the worst tiebreaks of the bracket and gets rewarded for it. We are putting faith into the tiebreak system as it exists being a measure of
“top 8 worthiness” here, but considering that’s what it’s used for to begin with, you kind of have to.

Basically, we are punishing you for being second to last in tiebreaks instead of dead last most of the time.

Backwards and wrong.

The cause of this is the pair down.

Randomizing the bracket to make the person who receives the “free” win arbitrary eliminates some of the feel bad, but it creates more of those scenarios
with tiebreaker mismatches. See the disaster that is Limited Grand Prix pods.

What if that pair down didn’t have to exist?

I’ve thought a bit about this and haven’t come up with a perfect answer, but this is the best I have.

In the last round, all players with X-1-1 records or better are guaranteed to not be paired down. If there are an odd number of players in that bracket,
the player in first place gets a bye. Pairing of remaining players is done by tiebreaks.

Not perfect for a number of reasons.

There are still X-1-1 with 70% breaker versus X-1-1 with 40% breaker matches where if the lower breaker player wins, they will be drawn out of top 8 by
X-1-1s.

This ignores lower bracket pair downs that don’t affect top 8 that cause the same issue because adding a ton of byes messes a lot of things up and just
feels wrong. There are still scenarios where this occurs. For example, I conceded my last round of Grand Prix Omaha because X-5s were on the very fringe of
prizes and my opponent had a full percent of tiebreaks on me.

This doesn’t account for Grand Prix or Pro Tours. The latter is usually fine, as the paired up X-4 is playing for top 8 as well, but in Grand Prix there
can easily be an X-2-1 who is or isn’t in contention based on size. And don’t even ask about Limited Grand Prix and Draft pods. I’m pretty sure those are
basically random because no one ever figures out if they are dead to begin with, only if they are 115% safe to draw.

Giving the first seed player a bye has weird tiebreaker effects in the late rounds where by winning, the 2nd seed player might get a chance to jump them
that is out of their control. An 8-0-1 beats an 8-1, which averages in an 8-2 or an 80% to their breaks, while the bye player holds their previous value.

I’m not sure of the exact scenario, but it might create scenarios where two X-0-1s could have drawn in the last round, but now one based on tiebreaks is
paired down and forced to play. This doesn’t break Swiss rules, but it does feel awkward.

I don’t think it can mess up the Swiss round numbers, but it might. I think that every case results in all X-1s making it, as the previous case accounted
for the X-1 pair down winning.

But it’s a start, and I think something along these lines is doable. Poke holes, suggest alternatives, but this is a fixable problem.

And by “this is doable,” I mean it would involve recoding Wizards Event Reporter. Best of luck to whoever has to do that next.