Kirb your enthusiasm!


"Pink isn't a color. It's a lifestyle." - Chumbalaya
"...generalship should be informing list building." - Sir Biscuit
"I buy models with my excess money" - Valkyrie whilst a waitress leans over him

Tuesday, February 28, 2012

3++ con statistics - Deeper look at Dark Eldar & Grey Knights


We begin some more detailed analysis of the 3++con results now that I've had some time to chill and done the initial digging. I'm still inputting the data in regards to some overall statistics (i.e. totals on winning conditions, who went first, etc.) but can start to pull out some numbers in relation to specific armies, etc. Unfortunately though, we again come across the issue of sample sizes. Very few armies had a lot of players so this is always going to be a consideration we need to have in the back of our minds. However, with six games each the sample size of games played is a bit larger though still dependent on the smaller sample size of players. This could give us some potentially more robust results but we're more likely to make inferential errors as single players still hold a huge sway over the total variance.

Enough number talk! Let's look at the numbers ^^. In this post we'll be focusing on the results of Dark Eldar and Grey Knights - specifically, we're looking at how Dark Eldar did against the 5th edition armies overall since there are very mixed feelings on Dark Eldar and how Grey Knights also did against the better armies/generals. It's a common belief that Grey Knights are overpowered yet this is generally countered by Grey Knights simply being an easy army to pick up and slay noobs with - against other good armies/generals they have no significant advantage. And of course, any match-ups between these armies will be looked at since Grey Knights are often thought of as one of the worst match-ups for Dark Eldar.

So first, the Win/Loss stats for both armies:

Dark Eldar (6 players) - 21-15 (58.33%)
Grey Knights (9 players) - 27-27 (50.00%)

So Dark Eldar did slightly better than average (and had two bracket wins) whilst Grey Knights did 'dead-on' average with a two bracket wins and the Runner-up position. What these final stats here represent is players from both groups had the skills and army tools to beat other good players of like skill levels on the 2nd day and take out their brackets.

So some pretty graphs now! Remember, we're looking for numbers outside the range of 40-60% - it's an artificial range I've created but it's what one would expect with relatively balanced lists going up against each other. The range accounts for error and minor imbalances - anything outside of this though and you're more than likely looking at some sort of discrepancy between the armies.

And a table of the results shown above:

Dark Eldar vs. Wins Losses
v .500+ 7 8
v .500 5 6
v .500- 10 0
Grey Knights vs. Wins Losses
v .500+ 3 14
v .500 12 7
v .500- 12 6
Dark Eldar Grey Knights
v .500+ 0.47 0.18
v .500 0.45 0.63
v .500- 1.00 0.67

Dark Eldar are pretty much as we would expect - roughly balanced against players winning as much as they are losing and against winning players and creaming up the crop against losing players. Their high-end numbers aren't positive however but are only a single win each away from the .500 mark which is something we want to see. Closer examination on how Dark Eldar do against 5th edition armies might provide more information. The undefeated record against losing players is particularly surprising since two Dark Eldar players themselves had losing records - one normally wouldn't expect this sort of bottom feeding from a fragile force like Dark Eldar but we can see a mild artifical inflation on their overall W/L record because of this. However, even without this bottom feeding, Dark Eldar are still close to the .500 mark (and this is against other winning players) which is what we are hoping to see.

On the other hand, Grey Knights have a staggeringly poor showing against other players with winning records though they did quite well (above that 60% mark) against players who won as much as they lost and lost more than they won. This certainly lends credit to Grey Knights being noob slayers but the very poor form against other winning players is very strange. This could be due to a number of factors - the majority of those losses could be from losing Grey Knights players and winning Grey Knight players matching up against each other regularly. Let's look at both of these to try and determine if plausible cause for this statistic.

Grey Knight Mirror Matches

There were three Mirror Matches with three Grey Knight players involved with winning records, two even record and one losing record. Only one of the winning Grey Knight players lost a game here so this isn't accounting for the terrible win percentage Grey Knights had against winning players at 3++con.

Losing & Winning Grey Knight players against Winning opponents

If we split the Grey Knight players into groups of winning and losing players and compare them to their opponents, do we get a different story? In theory we are expecting all 5th edition armies to do well against each other at the top of the heap where player and army are of equal levels and we expect Grey Knights to do quite well against losing players, even in the hands of losing players as they are quite easy to use - hence the many Internet cries of OP and cheese. Let's see what splitting them up does...

And the results in a table:

Winning Grey Knights Wins Losses
v .500+ 3 6
v .500 8 0
v .500- 6 0
Losing Grey Knights Wins Losses
v .500+ 0 8
v .500 4 7
v .500- 6 6
Win_GK Lose_GK
v .500+ 0.33 0.00
v .500 1.00 0.36
v .500- 1.00 0.50

A much better picture as we can see the losing Grey Knight players going 0-8 (0%) against winning players. However, our winning Grey Knight players still only went 3-6 (33%) which is out of the range we are looking for - this could be a factor of sample size or against the other top armies generals, Grey Knights are cutting it. I'd lean towards the former given past evidence and the population pool we are looking at though the 3-14 mark before might have indicated otherwise until we looked deeper.

We're also seeing the winning Grey Knight players dominating the even and losing opponents whilst losing Grey Knight players are holding their own against losing players. These indicate the tag of noob slayer for the Grey Knight army potentially holds a grain of truth, particularly when we combine with past results of Grey Knights not dominating against other winning players.

Let's look at Dark Eldar partitioned into winning and losing players for some comparison.

And the results in a table...

Winning Dark Eldar Wins Losses
v .500+ 6 1
v .500 3 1
v .500- 1 0
Losing Dark Eldar Wins Losses
v .500+ 1 3
v .500 0 4
v .500- 4 0
Win_DE Lose_DE
v .500+ 0.86 0.25
v .500 0.75 0.00
v .500- 1.00 1.00

We can see a marked difference between the winning Grey Knight players and the winning Dark Eldar players here - the Dark Eldar players are winning more often than not, regardless of opposition. The losing Dark Eldar players and are holding their own against other losing players but not so much against even or winning players which isn't unexpected.

What's important to notice here though is the winning Dark Eldar players (two) nearly make up the total number of games against winning players that the winning Grey Knight players (four) had despite having half their numbers - this is very indicative of the bracket wins the Dark Eldar players achieved compared to the Grey Knights (brackets 2 & 3 compared to 4 & 5).

Dark Eldar & Grey Knights versus 5th edition books

So what about both of these armies against 5th edition books? That's after all what they are and what they should generally be competiting against, particularly since we are calling 5th edition the most balanced we've seen. We'll leave in Tyranid and SoB results despite considering them the woopsadoodles of 5th edition.

Dark Eldar (6 players) - 14-12 (53.85%)
Grey Knights (9 players) - 19-18 (51.35%)

So again, we're seeing these armies hold their own against what is supposed to be their peers with perhaps a little inflation from wins against Tyranids and Sisters but these numbers are falling into the ranges we are looking for. Now let's look at their own match-up... We've done an army comparison before looking at strengths and weaknesses on both sides and it's one of the more polarising debates out there in relation to 5th edition balance and the prowess of Dark Eldar. There were six match-ups, three of them between winning players on both sides. Drum roll please...

Dark Eldar - 4 wins, 2 losses
Grey Knights - 2 wins, 4 losses

And between the winning players matched up against each other, Dark Eldar won all three (all against 4-2 Grey Knight players).

Conclusion & Closing Thoughts

The main thing to draw from this is when we dig deeper into the Dark Eldar statistics, we're not seeing an over-inflation of wins from the bottom shoring up a terrible showing up top. Whilst overall the Dark Eldar were bottom feeding at 3++con with a 10-0 record against losing players, they were still quite close to .500 marks against winning and even players and the winning Dark Eldar players were winning regardless of their opponents (with two losses between them). This is further indicating Dark Eldar are a capable army, balance well against other 5th edition armies and as evidenced by victories in Brackets 2 & 3, capable of beating their peers consistently when wielded by a good player.

On the flip side, digging further into the Grey Knight data does show extensive bottom feeding as expected (12-6 against losing players) (the title of noob slayer after all) but this unusually compensated for a terrible showing against winning players (3-14). Whilst a lot of this was losing Grey Knight players being shot out of the water by winning players (0-8), even the winning Grey Knight players didn't fair too well (3-6). I would imagine this is more of an anomaly and they would normally be closer to .500 against other winning players with their overall win percentage magnified by their bottom feed tendency (even amongst losing players).

Again, I will stress these results are all very tentative despite the increased level of digging I can do thanks to having a lot of the information at my fingertips. There are still a plethora of factors I cannot account for or factor into this analysis and we have the smaller sample size issue. These results do reinforce the belief that Dark Eldar can compete at the top level though. We've seen it at NOVA with the statistics to back it up and 3++con with some extra digging (which showed the two Dark Eldar players who won brackets had a very hard road to go through). Hopefully more data will follow this. With Grey Knights we've seen this in the past, particularly their bottom feeding tendencies which inflate their winning percentage, but such a poor showing against other winning players indicates they are not as powerful as many believe (not sweeping the top tables before them) which past results would agree with. However, such a poor showing is perhaps a particular inference of this tournament as we would expect them to at least be close to breaking even against other winning players, particularly with the runner-up position and two bracket wins in pocket. This is more likely an  indication of the opponent's played Day 1 and in the bottom brackets (less winning players) and more data may have shown a more even spread.

Phew - hope that satisfies everyone to not suggest I am claiming these results are now canon ^^. What armies would people like me to look at next or specifically, what match ups? Remember there is limited data so some armies we really can't look at too in-depth.

Follow us on Facebook!

Related Posts Plugin for WordPress, Blogger...