Frankfurt Major Day 1 + 2 Stats

uberxD

412 Posts

November 15 2015 00:10 GMT

Here you have the same table than yesterday, but adding day 2.

[image loading]

I added 5 more heroes, so we can see top 15. It's clear which are the top heroes during this event, that's why I hope when playoffs start we will se some fancy new strats from the teams.

Some quick notes:

- Doom managed to win some matches today, which makes his WR goes up but not even close to be "the most op hero in the patch". Have teams prepared for doom?

- Look at that Undying WR, 1st place during group stage. Is Undying the answer to doom? (bad death-wording joke).

- Another hero having a very bad event is Dazzle... poor guy, it helped me win so much matches in pubs.

- QoP and Undying are leading the ban increase when you compare it to the patch.

Well, hope you enjoy it!

the bear jew

United States3674 Posts

November 15 2015 15:45 GMT

Really surprised by juggernaut's low win rate, especially when it's good for the patch. But I see my old love undying still going strong.

aboxcar

United States447 Posts

November 15 2015 16:59 GMT

what would be really interesting is if there were a way to control for the teams actually at the tournament

b/c comparisons like the above are made with stats from all teams across all sorts of matches, and compared with a limited sampling of games from a limited pool of teams. thus, the interpretable information is liable to be limited or even misleading. superficially, one can mostly only comment on the "meta" of the tournament against the overall meta.

one would more deeply like to be able to answer questions such as:

1) of the teams present at this tournament, were they always better at, or worse against DS? or is there some difference at this tournament?

2) of specific teams that were picking and winning with dazzle, has something changed at this tournament?

and so on

uberxD

412 Posts

November 15 2015 17:59 GMT

On November 16 2015 01:59 aboxcar wrote:
what would be really interesting is if there were a way to control for the teams actually at the tournament

b/c comparisons like the above are made with stats from all teams across all sorts of matches, and compared with a limited sampling of games from a limited pool of teams. thus, the interpretable information is liable to be limited or even misleading. superficially, one can mostly only comment on the "meta" of the tournament against the overall meta.

one would more deeply like to be able to answer questions such as:

1) of the teams present at this tournament, were they always better at, or worse against DS? or is there some difference at this tournament?

2) of specific teams that were picking and winning with dazzle, has something changed at this tournament?

and so on

In short I'm agree that tournaments meta is extremely hard to analyze as a whole. There's too many variables but too small sample size. It's very interesting to watch, anyways.

The most interesting, and I think conclusive, data you can get is the very low WR Doom has. For his 'condition' a minimum of around 50% for like 6 patches played would be OK for an OP hero.

aboxcar

United States447 Posts

November 15 2015 18:43 GMT

But _why_ is doom's winrate low? My point is that superficial stats alone don't really tell you a lot. We want to know how, why, and what is significant.

For example, if you observed Secret in past tournaments, you would notice that after losing games with Envy on "big carries", instead Puppey put him on doom, or sacrificed him in the offlane with dual-lane ember, etc.

But this tournament, again Puppey tried to put Envy on a big carry like anti-mage, and they won.

Things like this are interesting, but you cannot really glean them from basic stats. It's still nice to see the 'meta' of a tournament, but it leaves us hungrier for more depth

NeThZOR

South Africa7387 Posts

November 17 2015 07:38 GMT

What is dead may never die. Even being "killed" by a patch or two.

Please or register to reply.

Frankfurt Major Day 1 + 2 Stats

Completed

Ongoing

Upcoming