Ghogiel, on 09 October 2013 - 09:04 AM, said:
Based on your stats alone? The non complete stats missing W/L etc, to compare to the fact that statically at least ~81% of games that are lost result in death anyway? While ignoring video evidence and the fact that hit detection can be quite bad?
No sorry I'll pass.
Let me be clear that I don't argue that there's no hit detection or hit box issues - there is, to deny that would be foolish. What I'm trying to argue is that the Spider isn't more or less "broken" in this regard than any other 'mechs, and the stats I gave, while incomplete, at least suggest that I'm not dying less in a Spider than in a Commando - which at least I would think would be the case if the Spider was as broken as this thread and others argue.
Fierostetz, on 09 October 2013 - 09:07 AM, said:
You might just not be a good spider pilot - every mech doesn't work for every person. I wouldn't take your personal performance as anything more than a component of a whole. One person is far from a statistically valid sample size.
Sure that might be it, but as I argued in another post against the same claim - is it not fascinating that I've managed to apply 1000+ drops worth of experience in Commandos (0.42-0.53 D/M) to one Spider (5D - 0.42 D/M) but not the other (5K - 0.59 D/M), and that they also neatly bracket my medium D/M stats (0.42 - 0.59 D/M)?
Isn't it also very fascinating that for a 'mech that is supposedly so broken it needs to be taken out of the game it doesn't have a significantly lower deaths per match rating than other, non-broken 'mechs?
And on the statistically valid sample argument - no, of course I'm not representative of the population as a whole, but I've given deaths per match stats that cover 545 light matches and 390 medium matches spread over five light chassis and four medium chassis, and the spread for deaths per match is 0.42 - 0.59 - exactly the averages for my 5D and 5K respectively.