Vodrin Thales, on 13 February 2013 - 04:07 PM, said:
Your statement may not be hyperbole, but it is seriously flawed. I dropped in a 4 man ...
See that's the problem right here. Tolkien's analysis was made with PUGs, not pre-made (see http://mwomercs.com/...ost__p__1709146 ). Nobody will deny a well organised pre-made can counter the effects of ECM. The question is whether the skill level required to do so is set too high for most of the people.
There is an easy experiment to test if ECM is overpowered. Make ECM available to everyone (at a % of total tonnage cost for example) and see how many players do not take it.
I'm not sure why you're using quotation marks on tests, or statistics since I literally did tests then calculated the implied statistics based on the outcomes.
Spoiler
<P>In case you didn't know I actually had done these tests and want to look them
over, here's a link to them, and just to clarify yes it is at the better than
99.999% certainty that n+1 ECM is correlated with higher victory chance.</P>
<P></P>
<P>I'm not sure if you mistook my statement for hyperbole since you didn't know
I had actually done a study, but I assure you it is not.</P>
[bah lost original post, oh well] Before anything else - I need to live up to a mistake. I read your post as you saying having more ECM gave a 99.999 victory percentage, or having more ECM gave odds of 99.999:.001. When what you said was that there is 99.999% certainity that having more ECM meant a team was more likely to win. That is something I agree and have stated previously, though I believe it is happening less as things have changed since ECM was first released (especially during the first 6 weeks or so of ECM).
That said, I was indeed using quotes as I did not see your previous workings with ECM, though in most cases using quotes is correct since you're the first person I've seen who actually has made efforts at logging data (on either side of the argument). While I truly believe your data to be a good start and some good effort, I do find it to be incomplete and flawed. Regardles of agreeing or disagreeing with anything, I do hope you keep up the good work on it, and it really was some good work, kudos to you.
As to not crap up this thread, I'm posting further feedback in a spoiler for the ECM testing since I'm unable to respond to your linked thread
Spoiler
As you stated, you were not a true observer, things could be missed. Also, a larger data sample would be preferred, even more so now that there are observations of games with more snipers and ECM being less common. It's still common, just not as often to see 3-5 ECM per side in pugs and more 4-man groups are dropping without it that I have noticed. Better yet would be for new data after ECM can be powered down for X seconds from PPC shots since I'm thinking it will reduce an Atlas's ECM effectiveness with how easy they are to hit with PPCs. Also, even better data could be gained after ELO is working. Hopefully whoever the PGI employee who said they'd take a look at getting more data available is able to follow through (though, with a lot more then just "is there more ECM on the winning side").
Items which would make the data a lot better would be if it included:
List types of mechs and builds used as well as ECM.
List maps played
List afk/[obvious] suiciders and whether they had ECM
Larger sample size, greatly preferred after ELO or with the same players
being used as test subjects.
All that said, I'm not putting too much faith in your results as I've had
much different experiences. Following in Vodrin's footsteps - last night I
played about 15-17 matches, most were in 2-3 person drops with some solo,
the people I was with knew it was a build experimentation night so we weren't
exactly playing to the best we are capable of. Of those games, less than 5 times
I was on a team with ECM, which was almost always a light mech suiciding anyways
leaving us without ECM, and my team was always outnumbered in the ECM
count. I lost three matches last night, and overcoming ECM so regularly
(sometimes against a team with 5+ ECM to our 0) received some commentary as it
seemed really odd at the consistency that ECM was a non-factor last night,
except in annoying me when I tried using an LRM build.
Lastly - I took a quick look through your reportings and found a couple
of mistakes which threw off your 33 wins and 8 losses numbers. One of them you
did not count when you won with more ECM, which would have counted as another
"yes." Another you put as a "no" when it should have not counted. I had those
pulled, but when I lost my original post I decided to not go back and get them
again.
Other than the times when I mess up and get in a bad position/don't notice the splatcat, I typically will win a brawl with my Atlas. They are far from an "I win" button. They may be a bit of an overpowered one-trick-pony-for their weight, but there are plenty of counters.
Sure, mate.... I didnt see its that easy.....I just have to run atlas, like everyone else(who dont run spaltcat) and win. I though it was game with all sorts of mechs. But I have this feeling we are coming down to this anyway.....just like my last 2 maches, 4 atlases, 3 catapults and me alone in Hunchback.
Edited by Jerod Drekmor, 14 February 2013 - 04:19 PM.
Getting both side torsos and attached arms stripped off my lightly damaged hunchie in a single salvo (I actually laughed out loud) by a facehugging splatcat, it struck me that the convergence of those huge, cumbersome ears should be a lot worse than it currently is up close.
4
Vodrin Thales, on 13 February 2013 - 04:07 PM, said:
