Tuesday, February 12, 2019

February 10 Pauper Challenge Breakdown

The other week I described the Pauper metagame as a game of ping-pong between Dimir Delver and Boros Monarch variants while the rest of the decks play musical chairs for third place. The February 10 Challenge did not do much to dispel that notion.

Just looking at these results, things look fine. The most popular Top 32 deck - Boros Monarch - didn't crack the Top 16 while Burn, Delver, and Dimir Delver all had the same Win+ score.
Win+ is a metric I use to help measure relative strength. It measures a deck's performance against the 32nd place list (or the final .500 or winning list in the case of a small challenge). Thus far, every Pauper Challenge has set the Win+ baseline, that is 0, at a X-3 record. One win above this is usually good enough for a Top 16 finish. 
In this particular challenge, five of the decks in the Top 8 went 6-1, meaning they had a Win+ of 2. In a discrete event this is important. However when looking at the format in aggregate those sorts of stats matter less. Instead we can look at how a deck is performing over a longer span of time. Burn has been very strong the past two weeks while Boros Monarch was also on the rise (until Sunday). Again, these are short term trends and help to inform ideas of metagame health.
In the micro, this Top 8 looks fairly diverse. Six distinct archetypes including two - Affinity and Orzhov Monarch - that cracked the elimination rounds for the first time this season. We see no Boros and no Tron, and last week's hot deck (Burn) continue to perform. How do things look when we talk about the four weeks of Ravnica Allegiance season?


Before diving too deep, there are a few things to know. The threshold for being included in this chart was a volume of 3 (2% of the total metagame comes out to roughly 2.56 appearances and for his purpose I round up). Both Izzet Control and Stompy made the Top 8 once but neither appeared enough times to make the cut. 
This chart is sorted by the last column - Delta Top 8. How the heck is this calculated? Remember what I said before about a Win+ of 1 being roughly equivalent to a Top 16? Some number of times a Win+ 1 deck will make the Top 8. Over the past two years it comes out to somewhere between twice a tournament and three times a tournament. Using this rough estimate (0.3125), you can estimate how many times a deck is expected to make Top 8 given it's cumulative Win+ score. On this metric, over four weeks Dimir Delver has three more Top 8s than it should all things being equal.
So what does this chart tell us? Only so much. I don't think anyone would dispute that Dimir Delver is one of the best decks this season and the same can be said for both Boros Monarch and Boros Bully. Yet Monarch, which just had a fairly down week, is outperforming its newer cousin. What gives?
Small sample size. That's the crux of the issue. These charts are great for identifying trends but less so for predictive. If they were stronger on the prognostication front, then we should have seen the rise of Tron weeks ago. Rather these charts all look back and try to make sense of what has happened. 
Well then, what happened at the end of last season? Dimir Delver and Boros Bully finished with 1.5 Top 8s over expectation. And the one before that? Dimir Delver had 4.13 Top 8s over expectation while Boros Bully had 1.63 (Tribe Combo finished between the two with 2.06). That takes us back to the start of October. That's a sustained success of two of the same macro archetypes for four months.
The Mythic Qualifier is three weeks away. Prepare for a format dominated by Dimir Delver and Boros decks. Have a plan for Elves, Tron, and Burn. Beyond that, pick something you know well. And maybe, just maybe, a week after someone makes it to the Mythic Championship on the back of Pauper, we will see something change. 

No comments: