Too Much Information – Baltimore Legacy Open

Jesse and Alix Hatfield, the Legacy brothers who developed High Tide, analyze the Baltimore Open in preparation for the Cincinnati Open this weekend, with M12 for the first time. See which decks are on top in the metagame.

It’s been a quiet couple of weeks on the Open Series circuit, but we’re back to examine the results of the Baltimore Open in time for this weekend’s stop in Cincinnati. We’ve also decided that enough data has accumulated since the release of New Phyrexia (the last major shift in the Legacy format) that it will be useful to look at some aggregate numbers, as a stand-alone chart. You can find both, along with the raw data, in the spreadsheet.

We also have a minor change in our methodology to announce. Previously, when calculating win percentages, we have counted draws as a third of a win, in keeping with the ratio of points a draw or a win earns in a round. However, after some discussion with Jared Sylva and each other, we have decided to count draws as half of a win. Not only is this more intuitive; it also ensures that the results from every tournament will contain a total number of wins equal to the number of rounds played and that the baseline “average” win percentage for each event will be 50%, rather than some lower value.

Without further ado, here are the results from Baltimore:

Baltimore Legacy Open Breakdown

fig 1 breakdown

If you’ve been following the Open Series at all since the release of New Phyrexia, the first thing you’ll notice in this chart will probably be the first number it contains: 9.43%. This is the percentage of the field claimed by Merfolk, and it is the lowest such value we have seen for some time (since before NPH, actually). As if to drive this point home, the next most popular deck (Zoo) is only one player shy of matching Merfolk’s popularity. While it is probably unwise to assume that Merfolk will again fail to break the ten percent mark in Cincinnati, the low point in Baltimore does fall in with a downward popularity trend that began with the Orlando Open and includes the Grand Prix.

Zoo, on the other hand, experienced a significant spike in Baltimore. In Indianapolis, Wild Nacatl and Co. was only 5.39% of the field, and they were even less represented in the Invitational and in Denver. Whether this sharp rise is the result of regional irregularities, or a larger movement, is difficult to tell. What we can tell is that Zoo’s win percentage in Baltimore was below 50%, and this follows a mediocre performance in Indianapolis. In our last article, we concluded that Zoo had been hampered in that event by variance (because its EV was higher than its win percentage), but the results of Baltimore may indicate a hostile metagame. Note that Zoo’s EV for Baltimore very nearly matches its performance.

Next up in popularity, and also fairly close to Merfolk in that respect, is U/W Stoneblade. Every U/W control deck that contains Stoneforge Mystic is included in this category, which is further broken down into sub-archetypes, depending on the presence of Ancestral Vision, Standstill, or neither. When we last looked at these decks, we postulated that Stoneforge Mystic could become ubiquitous in U/W control, and it appears to have done just that. There were only six U/W decks without Stoneforge Mystic, four of which used Ancestral Vision (seen at the bottom of the chart) and two of which used Standstill. One of those Landstill decks made the finals of the tournament, but the archetype still made up less than one percent of the field, so it failed to make the chart at all!

After Merfolk, Zoo, and U/W Stoneblade, the field quickly disperses. B/W Stoneblade, Dredge, and NO RUG all come in at 4.4%, with fourteen players apiece, and each of them actually performed better than the more popular choices. Dredge and NO RUG, in particular, put up strong numbers, and it’s hard to ignore how consistent the latter has been across the last few events (from the Grand Prix through Baltimore, NO RUG’s win percentages have been 56.67%, 57.32%, 56.85%, 58.33%, and 56.18%).

Next, we see Burn showed up to do poorly, again, which is only worth mentioning because Hive Mind showed up in equal numbers. It does seem that the quirky combo deck’s transition to being mainstream is finally exacting its price, though, leading to an unexciting win percentage. It still boasted a high EV for the event, but we should remain skeptical of that value until more data has been accumulated.

A few more things stand out from the chart. Affinity looks a little bizarre sitting so far down, with only ten people playing it, but its dismal record at this and previous Opens is probably a good explanation for this drop in popularity.

Likewise, Team America has fallen out of favor after struggling in Providence. In Baltimore, the deck served its pilots reasonably, but it doesn’t really look poised for a comeback anytime soon.

Several decks at the bottom of the list sport very high win percentages. These include Ad Nauseam (noteworthy just for being this far down the list), Mesmeric Monolith, and U/W Visions, one of the non-Stoneforge U/W decks mentioned earlier. The two combo decks did significantly better than their EVs would have predicted, but U/W Visions had an EV to match its results. In the last case, this may be warped by a lack of data, but it may also show a deck that is very well-positioned right now.

For the chart of aggregate numbers, we have stuck to the same model we use for every field breakdown, since all we’re doing is combining all of the post-New Phyrexia data into one “field.” Thus, we have included every deck that makes up at least one percent of that field, and the numbers corresponding to each deck have been calculated using this combined field, as well.

Here are the results:

Aggregate Breakdown: 5/15/2011 to 6/26/2011

fig 2 aggregate

As expected, Merfolk sits atop the format in popularity and boasts a pretty good win percentage to boot. With more than 2000 matches-worth of data, Merfolk’s numbers all carry a good bit of statistical weight, and this is bolstered by how close the win percentage and EV are. What that means is that there were enough matches played as to nearly eliminate the variance from pairings, and that the win percentage across this time period is likely a very accurate representation of the deck’s position in the metagame, on average.

Keep in mind, though, that recent tournaments have been somewhat harder on the Lord of Atlantis, and his popularity may be waning somewhat.

The next-most popular deck since New Phyrexia was released is Team America, which is interesting because it has been much less popular in recent events than it was prior to Grand Prix Providence. Perhaps it is more telling that the deck has failed to put up any spectacular finishes since then, and a low aggregate win percentage doesn’t bode well for anyone trying to change that.

Despite being less than half as prevalent as Merfolk, Zoo is the third deck in our chart. We have repeatedly highlighted Zoo as a successful and well-positioned deck, both before New Phyrexia, and after, and here we can see more good numbers to support this. Only in the last few Opens, in Indianapolis and Baltimore, have we seen Zoo struggle to post positive win percentages, although those results may be more reflective of the current format than this chart.

U/W Landstill comes next, although this clearly contrasts with recent results. In fact, the deck was almost non-existent in Baltimore. This is likely due to the widespread adoption of Stoneforge Mystic in U/W control decks; U/W Stoneblade, as an archetype, appears to have displaced U/W Landstill.

Affinity is present, of course, although it hasn’t really justified its presence for some time, now. The deck has one of the worst win percentages in the chart, and this does agree with the most recent data. With more than 600 recorded matches to draw information from, we can be more certain than ever that Affinity is not a good deck in the current Legacy format.

After Dredge and Junk, which each made up around four percent of the total field, there are a multitude of archetypes that made up smaller and smaller portions of the metagame. Standouts include NO RUG, with its win percentage hovering around 56% (an astonishing number, really) in every event it was played in; Hive Mind, which has a great overall win percentage even with Baltimore factored in; and High Tide, barely making the list at all and with a win percentage even worse than Affinity’s.

Merfolk — 9.43% of Field — Won 52.17% of Matches

Example: Ryan Pawlik, 3rd Place

fig 3 merfolk

Merfolk’s level of success in Baltimore is surprising in light of these matchups. The deck managed to win most of its matches overall, but failed to post a winning record against any of the top decks (though, historically, it seems to be favored against Dredge). Most strikingly, its worst matchup, Zoo, was the next most popular deck. Merfolk broke even against both U/W and B/W Stoneblade, but it looks like both decks are slightly favored against Merfolk overall. Merfolk did poorly against NO RUG in Baltimore, bringing the overall matchup very close to even. Once again, Merfolk’s success in this tournament cannot be attributed to its matchups against the popular decks. Given its low expected value (49.83%), the field may simply not have been favorable for Merfolk at all.

Zoo — 9.12% of Field — Won 49.69% of Matches

Example: Kemper Pogue, 10th Place

fig 4 zoo

The Baltimore Open was an interesting tournament for Zoo, which posted its highest field presence and one of its lowest win percentages this year. Just like the surge of Zoo players likely had an effect on Merfolk’s performance, the drop in Merfolk’s popularity wasn’t a good thing for Zoo. We don’t have a lot of data on Zoo’s matchup against the next most popular deck, U/W Stoneblade, but it seems favorable for Stoneblade, which was ahead in this tournament and is ahead overall. Zoo did well against the next three decks, however. The B/W Stoneblade matchup looks particularly good for Zoo, based on past evidence. The Dredge and NO RUG matchups, on the other hand, look close overall.

U/W Stoneblade — 7.86% of Field — Won 49.63% of Matches

Example: David Shiels, 4th Place

fig 5 UW Stoneblade

U/W Stoneblade didn’t have a very good day in Baltimore, and that may partially be because its matchups against most of the other top decks are fairly close. It does look like U/W Stoneblade is weak against NO RUG, with twice as many losses as wins overall. U/W Stoneblade is relatively new, however, so we don’t have a lot of historical data for its matchups. With Stoneblade emerging as the most popular U/W deck, we are likely to get a lot more information about its matchups over the next few events.

B/W Stoneblade — 4.40% of Field — Won 53.89% of Matches

Example: Tim Frank, 15th Place

fig 6 BW Stoneblade

B/W Stoneblade has shown up in modest numbers in the last several tournaments, so we have some data to work with, even if we don’t have much from Baltimore itself. In particular, we see that B/W tends to beat Merfolk and lose to Zoo. It’s also behind against Dredge. The U/W Stoneblade and NO RUG matchups look close, but more data may reveal one deck to be favored.

Dredge — 4.40% of Field — Won 56.82% of Matches

Example: Mitchell Zelmanovich, 16th Place

Dredge’s matchups all appear very close. It remains slightly behind against Merfolk in overall numbers, and it did very well against B/W Stoneblade in Baltimore, making the overall record favorable. Dredge’s overall records against the other decks are all within one match of being even.

NO RUG — 4.40% of Field — Won 56.18% of Matches

Example: Alex Bertoncini, 13th Place

fig 8 NO RUG

This is the first time we’ve been able to look at NO RUG’s matchups, and there still isn’t quite enough data to draw many strong conclusions. We can see that the Merfolk matchup is very close. The Zoo, B/W Stoneblade, and Dredge matchups all appear close as well, though the samples are somewhat small. It looks like NO RUG has a good matchup against U/W Stoneblade, which likely contributes to its repeated success.

NO RUG’s high win percentages are one of the few consistent elements of the current metagame. The Baltimore data shows several departures from previous tournaments, and time will tell which of these are outliers and which are the beginnings of trends. This makes it very difficult to determine what decks are likely to be good in Cincinnati. We’ve recommended (and played) Zoo repeatedly, thanks to its consistently good performances since New Phyrexia. Similarly, U/W Stoneblade and Merfolk have, more often than not, posted solid records. In Baltimore, however, it looks like none of the top three decks were good choices at all. Most of the decks that did perform well, like B/W Stoneblade and Dredge, don’t have a consistent track record. NO RUG really is the only deck with a history of doing well that still found itself well-positioned in Baltimore.

But the most striking results from Baltimore might be the unexpected surge of Zoo players and the continuing decline of Merfolk’s popularity. We’ll probably see these numbers move back towards their previous values in Cincinnati, but there’s a chance that these are true metagame shifts. If Merfolk finally loses its position as the most popular deck, we may see a very different format start to emerge.

The Baltimore Open contained several surprises and raises several interesting questions about the future of the metagame. We’re excited to see how Legacy continues to develop in Cincinnati and beyond.

Alix Hatfield
Jesse Hatfield