The probability that a project will succeed is different for different cost and time targets. There isn't a single right answer to "How long will it take?" or "How much will it cost?" There are a whole bunch of answers, each with its own probability of being right.
The conventional techniques for planning projects only give us one answer—and it's wrong; it's invariably optimistic. Optimism is a good thing as an attitude, but we don't want it in our estimates.
Worse, it usurps management authority to decide how much risk to accept in the target. Impossible goals and missed targets are bad for morale; poor morale leads to lower productivity and trouble retaining good staff. With better information, we can choose more realistic targets and hit them more often.
Effective planning and estimating tools should give us an accurate picture of the relationship between targets and the probabilities of meeting those targets. This is information we can use; if we know what shape the odds are, we can take action to improve the odds or hedge them. We can identify and address problems sooner.
Generally accepted techniques for project planning and estimating don't give us the information we need to do that. Calibrated estimators, good models, probability management and Monte Carlo simulation do.