Yep, the that’s my standard advice to players “afraid of taking mulligans.” That includes most newer players. These players just don’t have the mulligan experience to understand or even believe any of the logical points you made in your comment. The only way for them to accumulate that experience is to practice taking mulligans. As I point out in the article, this advice isn’t intended for all players.

And I’ll say this: more decks can win with a four-card hand than you think. It may be rare, but I’ve occasionally found it correct to mulligan crummy five-card hands in Modern with a variety of different decks. Temur Delver mulligans particularly well, and the Eldrazi deck was notorious for the advantages it gained by aggressively mulliganing into Sol lands. For a great example of a player correctly shipping a “shady five,” check out the video linked in the first sentence of the article.

Mulliganing has everything to do with what you believe will happen with one less card. If you know there’s no way you win with a four card hand you keep the shady five and hope for good topdecks. Every card fewer in your hand increase the chance your next mull will just be worse, and very few decks can go below five with any hope. Seven to six is practically free with the scry, six to five is a much harder call if the hand has any potential at all because if the five comes up a total dud you just lost the game without even playing.

Intentionally. This article was already quite long. But I should have mentioned earlier in the article that all these hands are based on the following:

– The deck is on the play.
– The opponent is unknown.
– The deck is pretty much a stock list.

I think, this way, we can get into mulligan discussion and theory without having to write a novel. Things get much more complicated when you factor in matchups, tech choices, etc.

I think that hand is fine. Yeah you lose to combo but there aren’t many combo decks in this format. Modern is mostly creature decks and creature-based combo decks, and you can interact with those, plus you have a fantastic mid-late game against just about any deck that doesn’t want to kill you on turn four.

I don’t know if intentionally or not, you left out a deeper analysis of these sample hands regarding the MU and being on the play or draw. These two factors should also be considered whenever you should mulligan or not (especially the first one).

It was game one and I was 2-0, runs three

My open was 3x lingering souls, path to Exile, and 3 fetchlands.

My normal rule with my esper deck is ship any hand without interaction before turn 3. This deck has that, but against any combo deck, burn, delver, scapeshift etc this hand is completely dead. (Damn control struggles)

I ended up keeping the hand and it happened to be infect which I soundly beat with an endless wave of souls after pathing his blighted agent.

But infect is almost always a win for my deck (so far about 80% win rate due to being well aligned game one and an endless stream of disruption) I’m curious what other people might have done.

Has it been considered for Modern Nexus to group the articles like these together? Articles about understanding and improving a player’s game, not unlike Level One by Reid Duke? I think it would be great for the collective works to be easily accessed for newcomers to the format at any given time.

Another point to note is that all of these arguments can be backed up with math if you frame the arguments correctly and use a hypergeometric calculator (like this one: http://stattrek.com/online-calculator/hypergeometric.aspx ). Forexample, I know it was helpful for me to see that I had a 66% cumulative chance to come up with 2+ lands on a 6-lander in my 20-land deck (as opposed to the 59% chance that I would pull a land in the following 2 turns), and those odds don’t even count the scry. A quick rule of thumb that I use is that any option that has above 5% greater odds in occurring is a course of action you should strongly consider, and any option that’s above 10% should be the default. Hopefully, others find this way to validate Jordan’s arguments as helpful as I did.