If you wanna know whether there is still time for significant polling shifts in Iowa, the answer is basically yes. On the equivalent date in 2016, Trump had 33% of the vote in the RCP average and finished with 24% (-9%), while Rubio had 12% and finished with 23% (+11%)
Also big shifts in 2004, where Kerry & Edwards surged and Gephardt & Dean collapsed late, and in 2012, where Santorum's surge came almost entirely in the final 5-6 days.
The possibility for late swings doesn't necessarily mean that there would be a *reversal* of current trends. They could also mean that Bernie winds up with 37% of the vote or something huge, etc.
