The National Snow and Ice Data Center in Boulder, Colorado has demonstrated how to construct an Arctic ice scam. The key feature of the scam is a graph like the one below, which necessarily starts in the Arctic record cold year of 1979, and appears to show a linear loss of sea ice headed towards zero.
What you would not want to do is include the 1990 IPCC/NOAA Arctic sea ice data which shows that 1979 was a peak, and that 1974 had about 1.8 million km² less ice than 1979. that would wreck your linear trend scam.
You would also not want to include this 1971 National Geographic map showing ice extent not terribly different from 2015.
The composite map below shows that ice extent in 1971 (brown) was much less than September, 1979 ice extent. It also shows that 1971 ice was much less than the NSIDC median (pink line.) Providing that information would completely destroy your linear trend scam and imaginary relationship with atmospheric CO2.
Another thing you most certainly not want to mention is that the low ice extent of 1971 was 1.5 million km² higher than earlier years. That would further wreck your scam.
If you were serious about destroying your multi-million dollar funding scam, you would mention that sea ice was no thicker in 1940 than it is now.
And you could bury your scam by showing that behavior of Arctic sea ice is cyclical rather than linear.
Keeping millions of dollars of climate scam money coming in requires lots of hard work cherry picking and ignoring reality.