WTC Stats

Cryx is definitely strong, he's just trying to nuance the data to prevent people from jumping to conclusions like "Denny 1, Skarre 1, Coven, Wraith Engine, Revenant Crew and Dark Host all need to get nerfed!"

Also, to the people assuming that all players at the WTC are all of the same (extremely high) skill level, with all due respect, but that's not true. Some countries don't have a very large meta and thus get very little training, some might only play a few times a month, but are still "the best" of their country, because there is no one else. Some countries had 2 or even 3 teams, more often than not, those teams are not all of the same level. If you took all the best players in the world and then divided them into roughly equal teams, then this hypothesis would fly, but some countries just don't have the talent or pool of people to stand up to the best of other countries, and that's fine, because like at any tournament, someone needs to lose for the winners to win.

Personally, for instance, I think of myself as a middle of the pack player. I know and have played against most of the people of the winning German team. With any given list, I think I have maybe a 20% chance to win against them. Those guys are amazing, play a shit ton of games and know the game inside and out. I take the game seriously, but only get 1-2 games in per week. Most people, and thus most countries are like that. When we were doing the matchups, there were countries that were considered weak, some that were considered average to good and some that were considered strong. Not all teams are formed of the best players in the world, just regular people that like tournaments and want to join the most awesome warmachine experience in the world.

Okay, so:

H0: there is no relationship between playing Cryx in this format and individual wins**
H1: there is a relationship between playing Cryx in this format and individual wins.

Because of the results of the chi-squared, we reject H0.

So, there is a relationship between playing Cryx in this format and individual wins. There are a number of possible explanations for this. What I can think of are:

1. Cryx is, as of the WTC, very powerful or hard to answer
2. People who play Cryx are just better at Warmachine so they should win much more often
3. Cryx still has a surprise factor, and people aren't prepared to deal with it
4. The pairing process and format is giving Cryx a lot of favourable matchups, because team captains are either getting their Cryx player into a good matchup, or throwing someone under the Cryx bus to try to get good matchups in the other four games.

Explanation 2 just seems unlikely because we've seen this before with Mad Dogs, Una2, etc. Explanation 3 also seems unlikely, given that people have known about the power of GF and DH for months, and they would have had time, before and after lists were made public, to prepare.

4 and 1 are both possibilities. But even if it is an artifact of the pairing process, I think it is still useful information. After all, if the best strategy against Cryx is to dodge, that also has some implications about where they are at in the meta.

My personal opinion is that:
1. A couple Cryx casters need to be looked at. Not saying the sky is falling and they definitely, totally, must be nerfed, but they should be looked at sometime over the next few months.
2. There might be some implications about CID and power creep, faction lobbying, etc., when DH is crushing everything right out of CID
3. For Mk.IV, it would be nice to clean up the rules around models dying, because the whole three-stage death process and the interaction between RFP and field promotion and recursion is just plain weird and clunky when you actually get into it, and I feel like there has to be a simpler and more elegant way.
4. The recent theme changes changed the game so much that the WTC data is out of date already anyways, especially given the very binary nature of Ghost Fleet. It feels like unless you have some very specific tools, which aren't all that common in the wild, the matchup between two equally skilled players becomes very difficult for the non-GF player.. Those tools are now more common because of new themes and merc availability, so that is going to have a huge impact on the matchups, not to mention that Cryx got some new themes as well.

** of course, this is all out of date because of new theme forces, so it is kind of pointless to talk about. Really, to make any serious analysis, what we need is stats on the next big individual event post-theme.

Hey professorlust, I have the barest of statistical knowledge so I have to assume that your numbers are correct but I really don't understand how you are drawing your conclusions. You need to help me understand why you are doing what you are doing.

Also, asking whether players had experience with Ghost Fleet? That seems a little odd because the answer is "yes" and "a whole lot", they also still lost.

Again... Lacking statistical knowledge. What are those errors? What do you mean?

I've noticed that when you got annoyed you use jargon in an attempt to quell anyone talking back to you.

Im trying to get you to use basic standards of argumentation. You could have done that when I asked. You chose not to and expended energy undermining your argument instead.

Well, I've learned that there is something called a type 1 error and we should avoid making them. Apparently.

So the "Null hypothesis" is a conclusion that we don't know anything?