I noticed that the LTC difficulty estimate is jumping around. Are you by chance only basing it on the time for the last block rather than a rolling average?
This one shows -19.35%. A few minutes earlier it was +38%...
btc 239719 8.71G 2016 239904 185 90.82% 12153412.00 15428597.34 26.95% 1d 0hr 17m 16s
nmc 114171 1.13G 2016 114912 741 63.24% 2932635.36 2994885.92 2.12% 5d 0hr 55m 58s
ltc 364901 1.69G 2016 366912 2011 0.25% 604.60 487.58 -19.35% 4d 7hr 54m 6s
I do not use last block or a rolling average. I use the exact calcluation since last diff change. So if you notice that 0.25% in your LTC row that is saying that the measurement at that moment is only based off 0.25 * 2016 blocks. The estimate gets more accurate as more blocks are found. The result is that toward the end it should be more accurate than someone using a rolling block window,
Thanks for the feedback.