Summary: I'm on the ladder and got deflected 5 times in a row without my anti-deflect triggering at least once in 5 tries.
Me: Anti Deflect 175
Target: Deflect 175
Deflect 175 (0.25% per point, total of 43.75% activation chance)
Anti Deflect 175 (0.2% per point, total of 35% activation chance)
07:06 17/Oct/2014 Your attack on XXX was deflected.
06:05 17/Oct/2014 Your attack on XXX was deflected.
04:14 17/Oct/2014 Your attack on XXX was deflected.
02:55 17/Oct/2014 Your attack on XXX was deflected.
01:37 17/Oct/2014 Your attack on XXX was deflected.
Probability of Deflect activating 5x in a row:
0.4375 ^ 5 = 0.01602840423583984375 or 1.602840423583984375%
~ 1.6%
Probability of Anti-Deflect not activating at all in 5 tries:
0.65 ^ 5 = 0.1160290625 or 11.60290625%
~ 11.6%
As you can see above, the probability of either case occurring (especially the first one) is so low that it makes me wonder whether there are other variables involved that aren't being disclosed aside from those percentages written in the skills' description.
I would appreciate it if anyone (especially from HCS) can give any info on this.
Edit:
Please stop posting about 'nerfs' or 'changes'. There's no point in going thru these topics until HCS actually fixes the implementation of the skills.
And thanks to Kildek for the links. The posts made me realize just how often this problem occurs (which also spills over to inventing, resource extraction, item drops, etc). 12 Deflects in a row on the same target = 0.0049% chance of happening, its ridiculous. I've had similar streaks when extracting components and getting more than 8+ of the same component in a single batch. So this could also be an issue with the RNG used.
Edit Edit:
I checked the code and it looks to be correct
Assuming if this is true (a big IF), he should look into the RNG used. Streaks of 7+ should only happen very rarely.
Edited by Maj, 18 October 2014 - 09:19.