On 20.1.2017 at 11:24 PM, ORCA1911 said:
Statistical theory relies on a sample being large, and how large depends on the skewness in the population we are trying to measure.
ouch, wow, dude this touches a thought process, especially now reading the threads.
in a game, where your goal is a “balanced unbalance” and achieve some kind of aesthetic by indirectly fumbling with the mechanics, usually statistics alone do not cut it anyway, without lateral creativity and respecting players as part of the artistic process. It’s like making music needs you to know the audience and auditorium as being part of it, otherwise the music will sound bad, and both will stop to enjoy the process, even if technically everything went allright - and in that thought now just imagine being the composer, and how limited you are in expressing yourself through this indirect barrier. Coz I think, thats where the dev sits.
But if your interpretation is wrong, sample size wont matter anymore. People who say its for the money, say its on purpose, dunno if I agree with that. It might be a mistake. Making mistakes is okay, as long as you don’t deny it, coz then you dont learn. We gotta endure, as fanbois.
Population changes over time of day, day of week, and generally. I dont think they are bad in statistics tho, it is probably even one of their strengths, as their maths in the game seems to work most of the time*, I just think, they rely too much on it, while the creation team is very creative, the balancing team is too analytic. I would confront both mindsets more internally, and I am sure, they will learn from this. But thats just my theory, as I assume that buff to be a mistake.
Looking at the individual problem, I dont think the hull size is the thargas main problem anyway, it’s rather the insane regeneration, while access to modules with multiple effects together with higher modularity just makes it a beast of a ship. In some modes, like survival, it might not be like that, all things you have to consider before interpretation.
And lets not think on relational effects between ship types and available modules, which creates an n:m problem, another limit for statistics afaik. Or the factor, that the players might range from monkey pressing buttons joyfully to an apm uberninja 180noscope roboaim nonstopplay tacticalcaesar minmaxer playing the game in 480p on a 4k monitor in the server room, and each players subjective tries to adapt yet again changing the result in each game.
I can totally understand (but wont condone), why some cry “play your own game” harshly, as I think, most of the time their buffs and nerfs are just sometimes a bit slow and late, especially when the house is on fire, but not so often wrong if I think back 3 years - while sometimes you get the feeling this or that could have been avoided if they used more probing for individual samples, to help them understand how much the numbers are okay or not. And of course, everybody, everybody knows better.
After all, if you forget to sample, a problem might not even show up, because you never measured the correct data.
We’ll get through this tho  Or do we.
*) actually, i am more a creative type, so statistics isnt my strength. i might be wrong, and bow to your understanding. but i do know about being at core development, not saying i know all.
1 hour ago, Oregyen said:
pff now ppl blame the new queue system.
imagine the old queues with the new ships. and that thought is imho quickly gone. actually the queue changes and the rank buffs were the single best change this game had in 2016. Srsly. That was better than Thargas and Destros. Ppl just need to open their minds to that. You know that xP
btw. should we tell them about adaptive? Or keep the belief, its a feature?
