found out when i researched this book. i talked to about 25 people who were there that day. i could do a minute by minute breakdown of what he was doing that morning. he wasn't selling cigarettes. so that narrative is actually incorrect. >> what was he selling? >> he wasn't doing anything. he was feeling unwell. he came out of the bathroom. he broke up a fight and the police approached him because a cop had seen him on the way to work that day. but, you know, this is a classic example of why statistics-based policing doesn't work. i think when cops mix stats it vastly increases the likelihood that these incidents are going to happen. and new york is a city that has one of those sort of community policing regimes -- >> what's the breakdown? >> what's the breakdown? >> you said this shows the breakdown. >> in terms of stop and frisk in the city, every year of the program it was between 80 and 90% of the people stopped were black or hispanic and it's a city that's more than half white.