I wish there were a way you could publish the predictors of failure in a way which wouldn't let people "game" the system unreasonably.
Actually, I think you have always published many of the predictors of failure. http://www.paulgraham.com/startupmistakes.html seems like a good start, but there are probably more specific indicators during an interview or during 3 months before Demo Day.
The ones which seem most relevant are all variants "not making something people want" -- either not making something effectively at all, or that which you make is a bad idea/market, or making something which is a good idea but crap implementation. Obviously several potential causes of each.
Actually, I think you have always published many of the predictors of failure. http://www.paulgraham.com/startupmistakes.html seems like a good start, but there are probably more specific indicators during an interview or during 3 months before Demo Day.
The ones which seem most relevant are all variants "not making something people want" -- either not making something effectively at all, or that which you make is a bad idea/market, or making something which is a good idea but crap implementation. Obviously several potential causes of each.