Spoiler season is rolling on. However, my prediction at the end of last week's article has been holding: Core 2021 spoiled the obviously Modern-playable cards early. Sure, there have been plenty of interesting cards, but they're all role-players or interesting build-arounds rather than massive shake-ups. And that's rather welcome, considering how often Modern's been churned up over the past year. A set with low impact is actually becoming novel.
I was planning to spend this week focusing on how the metagame has developed. And then Conspicuous Snoop was spoiled. I have a bit of a history brewing with Goblins in Modern, so it immediately drew my attention. It turned out I was a little late to the party. Snoop has a lot of potential due to an interesting interaction, and has set the world slightly abrew. I was not immune and have been working on making Snoop work in Modern Goblins. Jim Davis actually beat me to publication this time, made worse by us both thinking in similar threads. However, our conclusions are very different. Where's he's hopeful, I'm far more skeptical. Snoop itself is a decent card. Making it work in a good shell has been a problem.
The Snoop Scoop
If you haven't been following the hype around Snoop, my discussing the card is surprising. It doesn't look like much in a vacuum. For double red, Snoop is 2/2 with no combat abilities. Instead, it reveals the top card of the library. When that card is a goblin, it can be cast. Therefore, Snoop is built as a card advantage engine. Courser of Kruphix has shown that netting specific cards off the top of the library can be powerful. However, there is a world of difference between getting a land drop and casting spells and an even greater one, especially in Modern, between 2/2 and 2/4. Besides, Goblins also already has lots of card advantage.
And if that was all to Snoop, I'd have thrown it in with the other interesting, but likely unviable, cards. However, for reasons known only to Wizards, Snoop has a third effect which turns it from a mediocre beatdown creature into a combo piece. Snoop gets the activated abilities of any goblin on top of the library (and just the activated abilities--don't make the same initial mistake I did). And there is one specific goblin with an activated ability that goes infinite. And another goblin that sets up Snoop.
Since Snoop isn't legendary it can copy itself, the tokens have haste, so they keep copying themselves until you have an arbitrarily large number of Snoops.
3. The last Snoop then copies Harbinger to find Sling-Gang Lieutenant.
4. Throw Snoops at your opponent's face until they die.
It's clean, simple, and can happen on turn three.
The combo can't be assembled instantly because Snoop doesn't have haste. Thus, Snoop must either be played the turn before going off, or something must give it haste. These are not deal-breakers or barriers to playability. Rather, I expect it to be a common mistake that players should be aware of and try to avoid, as with Devoted Druid combo or Splinter Twin.
Required Obvious Comparison
Which means I must, wearily, compare Snoop combo to Twin and Counters combo. On the surface, this is a fair comparison. All these combos require two cards to pull off, classic A+B brokenness pioneered by Trix. However, only Twin is truly two cards. Deceiver Exarch or Pestermite plus Splinter Twin equals an actual win. Counters combo consists of Druid plus Vizier of Remedies, but winning requires at least one more card. The combo just makes infinite mana.
Snoop also combos off with one other card, but it doesn't win the game on its own. Kinda. There has to be a Kiki-Jiki on top to combo and a Lieutenant on top to win, meaning both have to be in your deck. That's not as demanding as needing to find then cast Walking Ballista, but also still not as clean as Twin. If Twin is a true 2-card combo and Druid is 3+ cards, then I'd say Snoop is a 2.5 card combo. Still, that's very good, and going off on turn three rather than Twin's four would suggest that Snoop is a new and potentially better version of Twin.
Evaluating the Combo
But I don't think that's actually the case. There are a lot of minor inefficiencies and some awkwardness associated with Snoop that I think makes it far worse than Twin. The glaring one is that it's very much a right-pieces-at-the-right-time combo deck. The play has to be Snoop, then Harbinger; going reverse-order means requiring another Harbinger. In the same vein, drawing Kiki-Jiki is a disaster. There is no combo without Kiki-Jiki exactly on top of the library. In my test decks, I never wanted to play mulitple Kikis, but had to as a hedge. It also meant I had to trim Goblin Ringleaders or risk mooting my own combo. That never happens with Twin or Druid, and in fact drawing extra combo pieces is good for them.
Then there's the issue of the cards themselves. Boggart Harbinger is not a good card. If it was, it would have seen play before Goblin Matron was reprinted. Demonic Tutor is far more powerful than Vampiric Tutor for a reason. Harbinger's only advantage is the extra point of power and the combo. Twin is also not a very good card in a vacuum, but at least it has some unique utility with any creature rather than being a worse version of another card.
Snoop combo may cost five mana compared to Twin's seven, but Snoop necessitates playing multiple five-mana cards in the deck, where Twin's doesn't. Snoop is just an aggro creature, where Deceiver Exarch is mildly disruptive. Snoop also dies to more removal than Exarch. And finally, since half of the Twin combo can be deployed at instant speed, pilots had the immense luxury of choosing whether to interact or go for it each turn cycle.
The bigger issue is the deck itself. Twin slotted perfectly into a reasonable midrange shell. The natural home of Snoop is a fine, but unsuccessful, midrange beatdown deck, which already has several options to combo off. Both contextually and in a vacuum, I'd therefore rate Snoop combo below Twin. The primary advantage of Snoop is a faster goldfish, while deckbuilding constraints make it a more awkward deck. And the fact that it's not banned is a plus.
Finding a Home
However, slightly-worse-Twin is not an indictment. When it comes together, it's phenomenal. And plugs a strategic hole in Goblins. Whether in Modern or Legacy, Goblins is strong against slow decks, but folds to combo. Goblins's clock is surprisingly slow and can only really race with multiple Goblin Piledrivers. Instead, Goblins grinds with Mogg War Marshall, Goblin Matron, and Goblin Ringleader. My previous efforts incorporated combo kills to plug this gap. However, they were too complicated and expensive to really threaten Storm. Requiring one fewer card in play, costing less mana, and being faster should improve the deck's combo matchups while not impacting the midrange and control. Should.
I started the same place I imagine everyone did and just plugged the combo into tribal Goblins. It's the obvious place and I thought it would be easy. It wasn't, and I'd definitely counsel against trying this unfinished build in a tournament setting.
This version forced the combo and played full sets of Snoop and Harbinger. This was not a great idea on my part. Snoop is a fine beater, and sometimes an army in a can. Testing against Miracles led to several late-game wins following Terminus where I Vialed in Snoop, then cast multiple Chieftans into Ringleader and attacked for the win. Going forward, I'd cut a mountain for another fetchland for another shuffle to help find goblins to cast, but to me Snoop has proven itself to be a staple in at least some capacity.
As for the combo, Harbinger was as poor as previously described. And also worse because it was clogging up my hand a lot. Goblins is already a top-heavy deck, and Harbinger made it worse. Not just by also costing three, but by making the card flow worse. Since Matron tutors to hand, it is actually card advantage, and more importantly leaves the chance to draw something you need open. Harbinger isn't card advantage, just delayed selection, and dictates your draw step. This isn't inherently bad unless you need to find a string of cards to get back in the game. Harbinger can be actively harmful in those instances. And shaving a Ringleader to stop drawing all my Kiki-Jiki's was a huge mistake.
Fitting in the combo diluted the deck's aggro capacity. This meant that in testing, the Jund matchup was closer, because I had to work harder for the win more often. However, it didn't feel like the combo was adding much overall. Goblins doesn't have the cantrips to make the early combo happen consistently, so the combo matchups didn't improve very much. This deck is an aggro-combo deck where the glue isn't quite strong enough. I think there's potential here if you cut down on the combo and treat it as a "whoops, I win" plan rather than anything integral. I'll shave on Harbinger but keep the Snoops in the future.
If the aggro deck didn't quite work, what about the more dedicated combo version? This was an easier deck to work with, mostly because addition tutor effects are decent in combo decks but also because I didn't have to jump through hoops or worry about the deck's identity as I was making it. Just shave a few other pieces and voila!
I stuck with Metallic Mimic for my persist combo piece rather play Grumgully, the Generous. Part of that was fitting the green proved tricky. Goblin decks are mana-hungry, and I was really straining the color balance and life total to make it happen. The primary reason was curve. Again, Goblins has a lot of three drops as is, and hands can get clogged when they're not just clunky. Even when I ran Vial, there were lots of hands that just didn't do anything because of all the threes. Grumgully is very good because it can be tutored, but until the curve problems get worked out, I'm staying away.
The combo worked better here, unsurprisingly. However, this deck felt worse than more aggro versions. The main reason is this version is much harder to play. With so many combos that don't really overlap and those combos being easy to break up, knowing what to Matron for and when proved vexing. I was also frequently in positions where I'd gone the combo route and then failed. Had I instead ignored the combo for value, I would have done better, if not won. Snoop was a bit worse on its own because the higher average cost and additional land meant I got fewer goblins, but with some refinement this could work. I just don't know that it's actually better than normal Goblins.
The Twin Route
I didn't think of taking Snoop out of a tribal shell until Davis' article. The idea is that, much like Splinter Twin, the combo doesn't take up an entire deck and doesn't necessarily need to be built around, allowing you to run an interactive shell. I took Davis' list and started tweaking it, ending with this:
This was not a good deck. It didn't have enough interaction to shut down opposing decks, and the threats were too weak to push through. It had to combo to win most of the time, and that didn't happen consistently enough. The problem is that outside the tribal shell, Snoop is just a 2/2 that can combo. It's far too anemic to be a threat on its own, and the goblin density is too low to consistently hit anything in a typical game. Harbinger is in the same boat. Perhaps in a Grixis list with cantrips it could work, but I can't imagine that deck being better than Death's Shadow.
You Got Snooped
Conspicuous Snoop is a good card, and will be played in tribal Goblins strategies. I don't know that the combo is good enough for these or other decks. Harbinger is such a poor card on its own, and the deck is already so top-heavy, that adding in another three-drop makes it clunk out more than I like. There's some consistency tool missing. If that can be found, then Snoop combo will be good. Until then, I'm sticking with the tribal beatdown version.