Wednesday, September 14, 2016

NOVA Recap Part 2 - 40k GT Stats

For this part of my NOVA recap, I wanted to dive back into my stat analysis role from Torrent of Fire and break down the results. Field of Fire took a crack at the stats from the event, and I'll probably be retracing some of that ground but looking to dig a little deeper. Part of this is massaging the data a bit to get slightly more accurate results. For example, Chaos Space Marines as listed did very well, but we need to differentiate between any Chaos Space Marine primary lists and those that were actually Renegades primary.


One of the first steps I did was try to clean up some of the Army/Faction entries, particularly those in the top bracket. I would love if I could have done this will everyone, but not exactly feasible. For most of my analysis I'm also dropping all factions with fewer than five players. This amounts to, after my scrub, dropping Orks (3), Sisters of Battle (2), Blood Angels (2), Chaos Space Marines (1), Skitarii (1), Corsairs (1), and Dark Eldar (1).

Of those drops, CSM is notable because the aforementioned Field of Fire article showed them having promising numbers. There were initially 7 players reported as CSM, so I dug into those and asked the players and/or someone they played what their actual lists were. I found that only one was a CSM primary, and from the sound of it, would probably be a Daemon Primary in ITC terms. Five were Renegades, so they will be reported as such here.

To start, let's simply look at win percentage, taking a factions number of wins divided by the number of games played with that faction. To note, this excludes drops and forfeits.


The blue bars represent win % while the orange bar and gray line show popularity of an army. Interestingly, we see a pretty sharp trend here, with the high end holding an extremely high winning percentage while the low end has an extremely low win percentage.

If we compare to last year this shows a lot more imbalance than previously, where the high performer was Eldar at 57.3% win rate and the low was Blood Angels at 40%. This year Chaos forces book-ended this extreme trend, with Renegades taking highest with 72.2% win rate and Daemons in third with 61.4%, while Khorne Daemonkin brought up the rear, winning just 25% of their games. To dig just a little further, the gap between first and last armies went from 17.3% to 47.2% this year.

Interestingly, Dark Angels and Space Wolves, many people's favorites to win (mine included) in some form of Thunderwolf Star, were at the bottom end, both coming in well below 30% win rates.

Imperial Knights brought up second place in win percentage, but they didn't make top bracket, which is not very surprising given their historical track record. That said, half of the Imperial Knights players (6) won their bracket, a seriously impressive feat. Speaking of, let's take a look at how armies did within their brackets by looking at average final bracket position.


Again, we see Renegades on top with not only the highest average bracket position, but also the highest bracket win percentage. Knights are also interesting, keeping in mind that half the Knight players were first in their bracket, an average position of 7-8 means the other three Knights players finished toward the bottom of their bracket.

To go with the bracket performance as a whole, we can also look at what factions won their brackets:

Army
# of Bracket Champs
Brackets Won
Imperial Knights
3
2, 4, 8
Eldar
2
1, 3
Adeptus Mechanicus (War Convocation)
1
5
Space Marines
1
6
Chaos Daemons
1
7
Renegades
1

That's a pretty impressive showing by Imperial Knights, winning 3 brackets. And while we saw Eldar hovering around 50% in the W/L column, at the high end they took two of the top brackets, while also filling 5 of the top 16 spots in Bracket 1.

Next, I wanted to take a look at strength of schedule. This was essentially just ranking the armies by the average bracket they were in. That is, the higher the average bracket, the stiffer the competition faced overall by a faction.

Army
SOS Rank
Avg Bracket
Imperial Knights
1
3.83
Chaos Daemons
3
4.13
Eldar
4
4.27
Dark Angels
5
4.29
Khorne Daemonkin
6
4.33
Adeptus Mechanicus
7
4.50
Tyranids
8
4.80
Renegades
8
4.80
Space Marines
10
5.29
Imperial Guard
11
5.67
Necrons
14
6.20
Tau Empire
15
6.20
Space Wolves
16
6.60

This table provides some interesting insight into the overall stats we've already explored. First, we can see that, as most people would expect, Imperial Knights tended toward the top half of the standings overall, and did well despite the stiffer competition rounds 5-8. We can also see that some of the armies that had some really terrible showings, like Dark Angels and Khorne Daemonkin had a high average bracket. This indicates that while they did well in early rounds, they suffered in the later rounds.

Similarly, two armies that did well, Renegades and Tyranids (and better than expected in the case of the latter), had middling bracket averages, meaning they won sometimes against the lower end of the tournament. Finally, the strength of schedule can show a really bad picture for Space Wolves and Tau, who not only trended toward the lower brackets, but didn't have much success once they got there.

So, what can we take away from the data we've looked at so far? First off is that the meta is extremely volatile - with the gap between winning armies and losing armies jumping dramatically about 30% wider than it had been last year. The second is that Eldar and Space Marines are extremely popular, but weighed down into the middle-region results as a consequence. I also see that despite my doubts, Knights and Renegades both have the ability to thrive in the current meta.

Finally, reading between the lines a little here, there seems to be a shift toward close-quarters to some degree, and this has punished traditionally static shooting armies (Guard and Tau) while also possibly stripping the advantage some armies previously had in the meta (KDK and Space Wolf/Dark Angel amalgamations).

What takeaways do you have from the data? What questions do you have as well - would you like me to address or highlight any particular stat I didn't cover? Thanks as always for checking out the post, and keep an eye out for part 3 of my NOVA Recap where I actually dig into my own games and experiences a bit. Also check out Part 1 of the Recap, which is a pic dump primarily from the 40k Hall and events.