Those numbers might be skewed towards a certain selection bias. Per the article:
The CNN poll was conducted by SSRS from July 1 through 31 among a random national sample of 1,279 adults initially reached by mail. Surveys were either conducted online or by telephone with a live interviewer. Results for the full sample have a margin of sampling error of plus or minus 3.7 percentage points; it is larger for subgroups.
The fact that the people in the survey had to respond to initial correspondence by mail makes me think that the survey self selects for an older audience (age was mentioned nowhere.) That mixed with the sampling error could easily just represent a 50/50 split, which is pretty par for the course. Wouldn't call them tankies though as their ideological manipulation comes from the other side of the political spectrum.