StillRadioactive, on 22 December 2014 - 01:29 PM, said:
Russ says that only 2-3% of all matches are ‘ghost-drops’ - within the last 2-3 hours of the 24-hour window.
EDIT: Re-read the tweet, only talking about 2-3 hours before cease-fire... Re-worked numbers a bit to reflect this.
Let’s look at that for a moment.
First, know that the 2-3% statistic must be for ALL matches, averaged across ALL planets. It might be significantly higher for any single planet (such as, say the contested planets at the border with the largest population disparity between factions)…
Imagine a planet with a full queue on each side (60+). This means that a match is dropping there between every 1-10 minutes (1 min between when there is a 12-man ready-up on each side, 10 min when only 1 side is ready).
Each match takes between 5-30 minutes.
Between these two things, there is a huge range of variability (I’m not even going to attempt to calculate the ‘average time’ here), but suffice to say, teams are going to be constantly dropping in & out of matches when the population is full – the 10 minute wait is enough that only rarely will one side have enough more people to get a ghost drop. Maybe that is 1/100 (1%) overall matches, maybe it’s only 1/20 (5%). Either of those is totally believable, I think, and neither would make an insurmountable difference in the 11 (now 15) wins on a planet over the course of a day, or even a couple hours time...
BUT
Consider the 2-3 hour period he's talking about for that given planet. Assume that the queue has been relatively full for that time period– so there have been possibly more than 150 matches for that planet during that period (one per minute is the max possible under their programming, IIRC).
Now we get to the last drop possible before cease-fire. The populations are full for both sides. The queues are long (and because PGI capped what we see as “60+”, we can’t know HOW long). But if the number of dropping teams one side outnumbers the other, by, 2, or 3, or 5 12-mans, they could get that many ghost-drops right at the end. This runs up the score, and what percentage is 5/150? 3%.
So that 2-3% ghost-caps, if they come at the end (as at least a couple ALWAYS do), are in fact highly significant. In fact their value is completely out of proportion to the number itself. Because they change the score on the planet WHEN IT CAN’T BE REVERSED.
Edited by BenMartin, 23 December 2014 - 04:52 AM.