You’ve got a choice!

You can choose to use only the setting: the customs, habits, clothing, food, mores, religion, ethics, technology and social structures of your chosen period.  This is the bare minimum, otherwise you’re not writing an historical mystery.

Or you can use the setting, plus use real people from history as major characters.

Or you can use the setting, plus use real people from history, plus use real events that happened.

PC Doherty is a good example of using only the setting.  With that approach he’s written ancient mystery series in Ancient Egypt, Ancient Greece, Ancient Rome, Mediaeval England, and Victorian England.  Basically, he’s done one of everything going.

If using only the historical setting seems like the easy option, keep in mind that the Sherlock Holmes stories, the greatest detective stories ever told, are a setting-only series.  Sherlock Holmes wasn’t, strictly speaking, an historical series when they were written, but they certainly are now, and even when Conan Doyle was scribbling away he was harking back to a recent but cleaner age (at least in his own mind).

Lindsey Davis is a good example of using historical people but avoiding real events.  In that case you’re overwhelmingly likely to use the real historical person as the patron in the standard model.  Vespasian gets the gig for Davis and Simon Scarrow.  Titus often gets a go too, and Julius Caesar and, not unsurprisingly, Cicero.

You might also use real people as lesser characters.  It’s quite common to give real people cameos, and the attraction is obvious: it’s always a bit of a thrill to see someone famous strut their hour upon your pages and then be heard no more, especially if the cameo is Shakespeare (who appears one way or another in an astonishing number of Elizabethan mysteries).  People like to see historical characters they recognise, and that alone is a good reason for including them.

I use Pericles as the patron for Nicolaos, and roughly 70% of all my characters were real people.  I suspect that’s an unusually high percentage, but I have the advantage that people like Aristophanes and Plato used people they actually knew in their work, so I steal from them mercilessly.

Steven Saylor has a reputation for being not only a tremendous writer of historical mysteries—maybe the best ever–but also for being the most historically accurate.  The reason is he not only uses the setting, and peoples his books with many historically real people, but in addition every single novel is based on a real historical event, usually a very famous one.  His subject is the Late Roman Republic, and how a small group of incredibly talented, incredibly powerful, and incredibly vicious people fought it out for domination of the world.  It’s powerful stuff.  The one thing you can be sure of in a Saylor novel is this: that if Saylor says it happened, then either it happened, or else you can’t prove it didn’t happen.  (That last clause is very important.)

I did the same thing with my own series.  The first book deals with the for-real assassination of the man who started democracy in Athens.  It’s mentioned in Aristotle’s Athenian Constitution.  They never caught the men behind the plot.  The moment I saw it, I knew it was the perfect basis for an historical mystery.  The following two books are also based on real events, and I’ll stick with the format as long as I can.

There’s a fascination with integrating fiction into real events which is hard to describe.  It’s a technical tour-de-force (says Gary modestly) also fraught with danger because you could either (a) make an outright historical error, so much easier the more tightly you interweave, or (b) get so caught up in the technical history that you forget to write a fun story.  So it’s more dangerous, but also immensely rewarding.   What helps you is that if you’ve chosen wisely, then the period and the real events are inherently fascinating to the readers.  At least, that’s my experience as a reader.

The basic rules:

• Don't break history.
  
  
• If it doesn't break history, then anything goes.  Pick whatever is most dramatic for your story.  This is the essential difference between history and historical fiction.  Historians have to go with whatever's most probable.  Novelists go with whatever's most exciting, no matter how unlikely.
  
  
• Mores, customs and societies might change, but people never do.
  
  
• Let the times, the society and the people speak for themselves.  Never, never, never coerce your own modern views on the past.  (Easier said than done!)
  
  

I don’t know that any of the three integrating choices is better than the others when it comes to writing a good story that people want to read, but I’ll tell you that going the whole hog for setting + real people + real events calls for a lot of reading.  Huge amounts of reading.  Vast amounts of reading.

This isn’t necessarily a negative if you like reading!