Yes. Please tell me how probabilities work differently in both polls and why I should be less skeptical of one vs. the other.
Okay - even though you asked sarcastically. One was a survey asking a sample of people what their opinions is to try to extrapolate the opinion of the overall population. A sample stating > 50% of the population holds one opinion is only wrong if 50% of that population doesn't hold that opinion. The forecast is an attempt to predict a future outcome. A probability > 50% doesn't mean the outcome is certain and the fact that Donald Trump won doesn't invalidate the forecast you posted.