Niederhoffer has some good insights into the markets, but that is only to be expected from an intelligent hard-working person who has traded for several decades.
Unfortunately he has a number of psychological inclinations which hinder his performance as a trader.
Firstly, Niederhoffer has relatively poor trading insticts (no genuine "gut feel"), which is mainly due to his neurotic obsession with counting/testing and his denigration of more "fuzzy" learning skills. This is probably due to his self-image as a scientific sceptic, critical thinker and debunker of "voodoo" (e.g. con artist technicians), and Randian/objectivist "rationalist" - everything must be black and white, and testable via hard concrete verifiable facts. But I find it strange that he data-mines so much, given his belief in the ever-changing nature of the markets.
Secondly, he apparently cannot keep away from excessive risk - he has a clear love of "action". I suspect that this is due to his childhood - as a boy he enjoyed the street games in New York, where reckless gambles would have been encouraged and even glorified.
He is also incapable of riding big trends, except upwards in the stockmarket, both of which are again purely down to his psychological makeup. He is philosophically biased towards entrepreneurial activity due to his Randian/libertarian beliefs, so of course stocks "must" go up in the long run. And he is a hard sceptic, so the notion that any other market (for example commodities or currencies) could have a persistent predictable price trend is anathema to his way of thinking. His belief is that inefficiencies are small and fleeting - so he will always take small profits and ride big losers, a surefire recipe for volatility if not outright bankruptcy.
Finally, he does not appear to have any real expertise in fundamental analysis - valuation appears to be an afterthought to him. The real nail in the coffin when it came to Niederhoffer was when I saw him in early 2000, recommending purchases of Nasdaq futures because the market had gone down a historically abnormal amount in a short period of time. He justified buying along the lines of "Testing shows that each time the nasdaq is down x % in y days, then it bounces z % on average over the next week". This was just before that one week where the nasdaq crashed over 25% - and he added several days in a row, presumably getting completely annihilated on a triple or quadruple size position as he averaged his loser. I couldn't help wondering why he hadn't considered that, on all the other occasions he had tested the data, the market had not just come out of a gigantic once in a lifetime bubble where tech stocks were valued about 5-10 times more expensively than they had been during previous selloffs. Such ignorance of key fundamental differences and their impact on the usefulness of historical data made me lose what little respect I still had for his trading ability.
Basically he has a whole host of psychological biases which impede what would otherwise be great market success. If he simply had an open mind and removed all his preconceptions about the markets, then within a few years he would probably a world class trader. Sadly, because of his ego and obsessive personality, he will never reach that position.
