How much research? It takes a lifetime. I've been reading history since I was a child, and all that reading contributes to what I do. However - when thinking about a new book I'll spend some months reading in a very concentrated way, though how long and how much depends on the book. I'm about to start a book set in the American Revolution and, though I know the period well, I still needed six or seven months of reading, though for, say, a Sharpe book, I'll only do about a month's dedicated research. But it's almost an unanswerable question, because my research began about fifty years ago and it's still going strong. As for facts versus story - my job is to tell a story. I'm not an historian. If you want to know what happened in The Peninsular War then read Professor Charles Esdaile's wonderful history, but my job (far more than his) is to entertain you. I do try and stick as closely as possible to the history, but where changing it would make for a better story, I change it. Thus, in Sharpe's Company, although no British soldier got through the breach of Badajoz, that was where the night's drama took place, and so Sharpe had to be in a breach, and if Sharpe attacks a breach, he gets through. So I changed the history - but, very important, I confessed my sins in the Historical Note. I don't think there is an accepted rule - Hollywood films change history outrageously - but my rule is to tell the reader what I changed.