Vrain Matari
Mikramurka Shock Troop
Minmatar Republic
1942
   |
Posted - 2014.04.09 14:10:00 -
[1] - Quote
Heavenly Daughter wrote:Why are we getting all this extra SP at the moment, The million clone event a few weeks back, making over 3 million, the sidearm showdown making another 75K, and next week the Accelerated SP week at 5 times normal rate.
I seem to recall CCP saying at the start of this game it would take 7 years to spec into everything, So why are we getting given all this SP,
not complaining but, Thoughts anyone.
Does this really keep people interested, at this rate the SP accumulated over 7 years would be close to 1/4 Billion, for what.
I'm genuinely puzzled myself. Always though sp inflation of the existing playerbase was a bad thing, but obviously CCP doesn't share that view.
If this was beta i would say it was for testing reasons, e.g. CCP wanted to guage the adoption rate of more expensive gear as higher-tier skills were unlocked, and was trying to measure how isk-restricted mercs are.
But this isn't beta...unless it really still is. Smh, my lord, this situation is so fubared it makes me laugh/cry.
Alternative is they're trying to get the player numbers up. It appears to be working according to the graphs. Ofc we have to factor out the account-churners(activity here has prolly gone up due to recent publicity) and the afkers(how many peeps are on Oceania, anyways?).
If that's their motivation it's prolly related to wanting to have as big an active playerbase as possible when the fanfest build hits.
Ahhh CCP Shanghai, your greatest lasting contribution to the FPS genre so far is how many skeptics and cynics you've bred :/
I support SP rollover.
|
Vrain Matari
Mikramurka Shock Troop
Minmatar Republic
1942
|
Posted - 2014.04.09 15:23:00 -
[2] - Quote
DeeJay One wrote:KalOfTheRathi wrote:Until it is is released and functional there will be no matchmaking in Dust514.  Because matchmaking in a sandbox (with sandbox restricted to the notion of bringing/using any stuff you want) is a PITA to do, we'll probably never have a "proper" one. Number of skill points and war points don't matter, even your precious KDR doesn't matter as a metric to match against. Gear possesed? Nope. Maybe gear used in previous matches? More likely, but still won't stop protostomping (play enough matches with militia, get in a noobs only game, get your proto gear out - metric gamed ;)
The primary variable which saves us from all that theorizing and speculation is the end-of-match ranking for every match you've played. This ranking has to be weighted in each match by who you had with you and who you had against you, and updated with every match you play. We could incorporate win/loss info also, but that might be redundant if our primary variable was done right.
ELO rankings in chess are somewhat like this but my inclination would be to use a Bayseian ranking/prediction system, which is what Halo did, i believe. But we could definitely write a better system than Halo's.
For the matchmaker-writers i would say generate your list of candidate variables and their distributions, use maxent for unknown priors, and calculate mutual information against probability of a win. Rank the list by mutual information and you have your predictive variable candidates.
That's prolly good enough to get good matches, but if needed much more can be done to predict matches.
Minigame idea: When displaying the pre-match roster allow each player to predict the winning side(no wagering allowed) via a binary vote. Datamine these results for the better prophets in the playerbase as a way to vet your candidate variables .
I support SP rollover.
|