I’ve said it a lot, Pithing Needle is severely underplayed. If Revoker weren’t so fragile I would definitely maindeck it for that reason. I don’t know about Thorn against Jeskai because they tend to board in Engineered Explosives, but it might be worthwhile to test. Interesting results, I hope that this method serves you well.

I went through the process (to the best of my ability, I am by no means an expert) and found mostly information I already knew, but with a few interesting blips. First, I found that Thalia is in an interesting spot right now, and that between being a first striker, and having her ability, she is very close to being an all star. in fact, I found that in 50% of matchups I wanted a fourth copy of her (the list I was looking at had 3) so playing one in the board may actually be a smart move. Second I found that I was greatly underestimating pithing needle in modern. I found that close to 80% of matchups wanted the first pithing needle, and a further 50% of matchups wanted a second. that’s probably enough to warrant mainboarding needle (or revoker), which I had never anticipated.

Another tidbit I found interesting was that I did not find a matchup where neither pithing needle nor Thalia were good enough to justify boarding into. It’s pretty telling about the strength of DnT that these cards are applicable in a majority of matchups, but not a massive majority. A closing thought I have is that if thalia is good enough to board into, it may be more worthwhile to keep a thorn of amethyst sided since some decks (looking at you jeskai) have a much harder time removing artifacts than creatures.

What I mean is that if you are going to Elephant your deck you need to have a deck that is as flexible as possible. To make the deck work you have to be playing the best cards possible in your colors and have as few unchangeable slots as possible. You are looking to prepare the best configuration of your deck for all matchups, and this typically means that you are looking at a deck full of answers. There’s just not a lot of reason to do this if your plan is to kill them quickly all the time. You’re looking for a more interactive game where you will win via superior positioning. This tends to only be true for midrange decks like Jund or Standard decks. Is this universally true? Certainly not, but it is the general rule.

I tend to think of decks with cores as having a large number of cards that you absolutely cannot change for the deck to function or be recognizable as that deck. Affinity and Storm fall into this category. Delver does not. There is considerable variation between lists in terms of its total composition and general plan. This is either because the core has not yet been found or because the general URx plan lends itself well to tuning and building your deck to suit your own taste.

I hear this a lot and I do not believe that it is true. I think that players unnecessarily limit themselves to only the known good cards and do not explore the actual pool enough. We have tunnel vision on the card pool, and I’ll be talking about that next week.

I’ve had similar experiences to Rory as well. I would describe the approach I took as a “semi-Elephant”, where I basically identified the decks where Merfolk’s Game 1 aggro rush wasn’t going to necessarily get me there, and adjusted accordingly. I have to say I’ve been fairly pleased with the results, but I think that’s in part because I have the luxury of not having to metagame for every single deck thanks to having a solid, proactive gameplan. A full elephant, as David describes, seems like a daunting task for Modern.

I’ll also note that I’ve found this article very informative – I hadn’t heard of the Elephant term before, though I’ve seen the method applied, and I think that the description of how to employ it was very apt and well-put. Good stuff, David.

First, the concept of a dedicated core and it’s relationship to the Elephant needs some refinement. The best example of this comes from, I think, UR/x Delver (and more broadly, tempo) decks. UR/x Delver variants do have an extensive dedicated core, but since the core only requires you to have 27+ instants and sorceries to reliably flip your Delvers, you have a lot of room to Elephant in a deck that is neither really a midrange nor a goodstuff deck. You can elephant your instants and sorceries between your main and sideboards without compromising your core because the core of the deck does not require specific cards, but card types. The prior post’s Hatebears example is another good example, where your deck core consists of the types of hate you plan to employ, which are flexible and encompass a broad range of card types, strategies, and configurations. Once again, Hatebears is neither, really, a “midrange” or a “goodstuff” deck, but employs a defined pool of cards with certain restrictions to accomplish it’s game plan.

Second, I think that the point you make about Elephanting in Modern being problematic due to card pool size isn’t quite as damning as you posit. While it is true that Modern has access to exponentially more cards than a standard environment, it is also true that we have years and years of data on commonly played bullets, answers, and flex slot cards for specific matchups in all colors relatively easily identified. In other words, I would argue that there is a working “industry standard” that functions, at minimum, as a starting pool of cards not much bigger than Block Constructed, where Zvi wrote about the Elephant.

It would appear to be a good candidate with some caveats. There are a lot of different versions and if you’re going down this road you need to pick one baseline deck and go from there. Mono-white, Eldrazi, BW, etc. are all going to yield different results and will require different iterations of the Elephant. I think that it’s a good idea to try, though I’m not sure you’ll find anything out about the bad matchups you didn’t know before you started. It’s probably best to just focus on the winnable good matchups.

Yes, exactly, That’s why I was proposing the partial Elephant. I do think that methods like these need refinement because I’m not yet convinced that they’re analytically valid, so be careful. That said thinking along these lines is a productive way to improve your sideboarding.

Mainly, I found it useful for the following:
-Small Tweaks to the Main Deck. Honestly, the Main Deck of Mono-Blue Merfolk is mostly solved, so it’s really about making a few tweaks to the handful of flex slots. For example, if I found that Phantasmal Image almost never made it into my ideal deck, that would be a good argument for not running it or cutting down.
-Moderate Changes to the Sideboard: Mainly, this was about finding the right combination of popular sideboard slots to use and whether to favor Negate over Spell Pierce, for example. Stuff like that.
-Minimize Over-boarding: The biggest benefit to actually constructing the “best build” for each match up was that I realized I had been over boarding a lot of my match ups. I’d see that I had 8 or so cards in the SB that were good in the match up and feel the need to bring them all in, even if it hurt the core of the deck. So this process helped me take a bit more of a lighter touch.

Ultimately, in Modern it’s probably fine to work with the core of the main deck and put together a core of sideboard cards and then tinker a bit around the edges, going through each match-up to see how you’d sideboard and swapping 1-2 cards out here and there until you arrive at a good main and side.

Also, if pressed for time I think you can cheat a bit in any format and go to a site like mtgtop8 and do a side by side comparison of decks that have done well in the last couple of weeks. If 80%+ of the decks play a card in the main or the side, you can feel pretty safe including it, at least provisionally. Then you can tinker with the match-ups to figure out the rest.