Worlds is over, and the time for analysis is here! Now everyone can take the time to evaluate whether they made the right meta-calls. Since my meta-call was clearly perfect, and, uh, since I’m a colossal nerd, I’ve collected and collated the cold hard facts of the matter for your viewing pleasure.
Firstly, some caveats. All values presented below are taken from the Jousting Pavilion version of the Worlds data. As such, I cannot guarantee the absolute accuracy of each datapoint. I’ve also divided up the data only in terms of factions (with the exception of one graph). This has the unfortunate effect of grouping, for instance, both Builders and Alliance Qohor/Rose Jon Snow voltron decks under the joint category of “The Night’s Watch”; however, separating by faction AND agenda leads to sample sizes too small from which to collect worthwhile data – and we’re already only looking at one tournament. If anyone wishes to know the win-rates of any specific faction/agenda combos, please comment below or contact me privately.
With those hand-waves out of the way, on with the juicy bar charts!
As with all later graphs, the number in brackets is the size of the sample, so, for example, House Baratheon had 26 games in total at Worlds to generate its win ratio.
Going into Worlds, the perception was that Targaryen was the faction to beat, with Greyjoy and Stark the closest contenders. The other factions were deemed to be quite far off the pace, with Baratheon and Lannister in particular widely written off as unplayable.
From these stats, we can see that Targ was indeed worth consideration. Not only did it have the most entries of any faction, but it also had the highest win ratio. After that, however, it gets more surprising. Greyjoy, one of the most-fancied factions before the tournament, finished with the joint-lowest win ratio (something that the data will hopefully shed some light on later). Meanwhile, Stark finished with an exactly average win ratio, with that choice of faction neither increasing nor decreasing your chances for victory.
On the flipside, the lesser-fancied factions didn’t fare too badly. Baratheon, far from being rubbish, performed just as well as Stark did – and Lannister had the second-best win ratio of any faction. It is worth emphasising, of course, that sample size does play into this – Bara, Lanni, Tyrell and Free Folk each had fewer than 50 games across the tournament. Nonetheless, these stats speak to meta impressions playing into what people chose to take – anecdotally I can confirm that I heard of people expecting to see Greyjoy and/or Stark in large quantities and trying to include cards to counter these factions. Presumably this allowed the likes of Baratheon and Lannister to sneak somewhat under the radar – for instance, people perhaps put less of a focus on locations in case they were hit by Frozen Solid or We Do Not Sow, and Lannister and Baratheon can play better to the board.
Last it’s worth noting that above all these stats suggest to me that the meta is in a relatively healthy place from a faction-balance – there is only 9% separating the win-ratios all 8 factions (plus Free Folk). Making the worst possible meta-call faction-wise still gave a player ample opportunity to do well, and no faction, not even Targ, gave one close to an auto-victory.
You’ll note here that 8 agendas are not included in the graph. These are: Banner of the Stag; Banner of the Sun; Banner of the Watch; Greensight; Kings of Winter; The Brotherhood Without Banners; The Conclave; and The Faith Militant. As you may have guessed, this is because nobody took any of these agendas. Every other agenda, however, was taken by at least 2 people. It’s also worth noting that, as these datapoints are taken from The Jousting Pavilion, Alliance doesn’t take into account what the banners are.
The disparity between agendas is larger than it is for factions. This is perhaps unsurprising as the sample sizes are broadly smaller. Of note, the three agendas with the lowest win rates were taken by 3, 2 and 5 people respectively. Meanwhile, the three agendas with win rates over 60% tell slightly different stories. The top agenda, Banner of the Wolf, had its stats massively inflated by 3 specific players – Lennart Paga, who won with Martell Wolf (winning 10 games out of 12), Kevin Shannon, who made the top 4 with Targ Wolf (winning 8 games out of 11), and Gilles Molina, who made the top 16 with Bara Wolf (winning 6 games out of 8). Take out their results the agenda drops down to 46%. However, that’s also taking out 3 of the 5 players who played the agenda, leaving the 2 who played Night’s Watch Wolf.. Curse you, small sample sizes!
The other two top agendas, Banner of the Lion and The House with the Red Door, have another curious stat not shown here – namely, every single person who took these agendas used the same faction, House Targaryen. With these win ratios being above average for Targaryen it suggests these were good choices for agendas, but clearly only for one faction, not necessarily the field at large.
The main agenda people were focused on pre-tournament was The Lord of the Crossing – and perhaps people focused more on teching against it than with running it. With 43%, it’s one of the lowest win-ratios of any agenda, and the lowest of any agenda to have more than five people run it. Conversely, the agenda yours-truly recommended people run to answer it, Fealty, had the 4th highest win-ratio of any agenda in the field. It was still only 54% however, suggesting that much like with the factions, the agenda win ratios flattened significantly once substantial enough datapoints existed. Dare we whisper it – the field was somewhat balanced?
Moving on, I want to look at individual factions’ win-ratios by faction they’re facing.
This is obviously a story of small sample sizes providing unreliable data. Stop the presses, Baratheon always beats Martell because it beat it the one time they played! It always loses to Tyrell! Interestingly though, for the two matchups we have the most data on, Stark and Targ, Bara actually has positive win ratios, with the 0% ones coming against some of the least-played and least-successful factions. That suggests that, while Bara wasn’t necessarily a good choice in terms of faction efficiency, it was at the very least a good meta-call, because it actually fared well against the top decks.
One look at this and the reason for Greyjoy’s poor win-ratio is clear – the two main factions it needed to handle, Stark and Targ, both proved tricky for it. In particular, a 41% win ratio against the most-played and most-successful faction spells a bad tournament ahead. The incredibly poor record against Martell is also surprising – historically, with its location hate and ability to run The King in the North, Greyjoy have been one of Martell’s worst matchups. Yet, two times out of three, Martell came out on top.
Like Bara, Lanni’s sample sizes are quite small. Unlike Bara, it seems like Lanni was a less good meta-call, with a narrow losing record to Targaryen. The 4-2 record against Stark speaks well for the faction, as does the 4-3 record against Greyjoy to an extent, but broadly the data on Lannister is too limited to read much into this.
Martell is also a touch lower on datapoints, but there’s enough here to work with. We can see the faction was able to go toe-to-toe with the two main factions, Stark and Targaryen, with positive matchups against most of the also-rans. Bara and Free Folk aren’t really worth looking at, but the one worrying stat for Martell would be that 1-4 record against Tyrell – it looks like the Tyrell decks remain a thorn in Martell’s side, with their ultra-constructive play very difficult to consistently prevent.
With the exception of the ever-irrelevant Free Folk and Bara stats, Night’s Watch has a dubious honour – the only faction not to have a positive win-record against any faction. Some loyalists were able to push relatively far into the tournament, mostly with more traditional defence decks rather than recruitment or voltron builds, but for the most part this was just not a tournament you wanted to take the Watch to. Maybe next year!
Stark are an interesting one. With the exception of the low datapoint city that is Bara and Lannister, they have winning records against every faction – except Targaryen. This is a particularly crucial matchup to have a poor win ratio in, as Stark Vs Targaryen was the most-played matchup of all across the whole tournament. Of the 48 times that matchup was played, Stark triumphed in only 20, losing 28. To put it another way, they didn’t even PLAY another faction as many times as they lost to Targaryen! However, the 42% win-ratio is far from unwinnable, and in their other matchups Stark generally came out on top.
Targ were widely regarded as the top faction going into worlds, and so it proved. With the exception of Bara and it’s piddly 8 datapoints, Targ had a winning record against all but Martell, where it had a still-respectable 48%. Of particular note are its performances against the second- and third-most important factions, Stark and Greyjoy, with 58% and 59% respective win-ratios. Surprising me, albeit with, again, few datapoints, Targ finally seems to have conquered its boogeyman of Tyrell. It seems like the recent months’ shift away from burn-focused decks towards aggro/military-focused decks has paid off for the dragon in this matchup. The Bara stat does remain surprising though, even with the low sample size, and somewhat tough to explain – was the faction struggling to handle Bob? Kneel and dominance don’t particularly seem like they should present problems, and cards like Second Sons make it quite easy for Targ to overrun Bara’s power challenges, so I’m at something of a loss here. This isn’t the first tournament this happened at either, as the faction-based Thrones WAR competitions in America also mirror this outcome. Answers on a postcard please!
“Why are Tyrell in a bad way, these all seem- oh”. Yeah, low datapoints abound here, but comfortably the most telling are those two against Stark and Targaryen – the most- and second-most played factions, and a worse than 1-in-3 win record against both. We can also see from this that Tyrell perhaps got a touch unlucky with its matchups – it played Stark disproportionately more than it should have, and Stark was evidently a terrible matchup for it. My hypothesis on the reason for it having a bad matchup here is that Stark can play to the board just as well as Tyrell can, and it can draw cards just as well, but Stark can do both at the same time. Tyrell typically have to take a tempo hit to get their draw off, meaning they either run out of steam in hand or on board allowing Stark to seize control.
But yeah, overall, they seem in a good place, except for the part where they lose badly to the two factions that make up like half the entire meta by themselves. Oh dear.
The Free Folk
Missing from this graph is the ever-useless mirror matchup data, for the simple reason that out of all the factions (well, ‘factions’ in this case), Free Folk mirror is the only combination that was played 0 times across the weekend.
This is a relatively small sample size all-round, but of note are the 3 100% win records against some of the less tempo-oriented factions, and the losing records against the top two factions. Free Folk isn’t quite Tyrell-bad in those matchups, but it’s certainly not a pretty picture for them, and probably speaks to why of the 5 in the field, only 1 Free Folk deck made the cut (and that one lost the win-and-in to make the top 32).
So there we have it, my breakdown of datapoints from Worlds. As mentioned before, I have the data for agendas and faction/agenda-specific combinations so if you want to know any specific breakdown please let me know in the comments below and I will happily furnish them for you. I will, if time permits, make a companion article to this one for Stahleck’s stats, so we can see how different a predominantly-European tournament looks (and also see the effect of the latest chapter pack). Until then, thanks for reading, and I’ll see you soon!