Almost from the start, the hottest issue in the contentious Fresno County district attorney's race has been the office's success rate of felony convictions.
Is incumbent Elizabeth Egan the prosecutorial juggernaut she says, with her office winning convictions in more than 90% of felony cases it filed last year?
Or is the office struggling under her leadership, successful in only slightly more than half the felony cases, as challenger Lisa Sondergaard Smittcamp says?
No issue in the race has been more pitched — or harder to pin down on a purely factual basis — than the conviction rate.
For voters, however, it may not be the best measuring stick for deciding between Egan and Smittcamp.
As multiple legal experts said, while each side's numbers are real, how they use them can be another matter.
"You can make statistics say anything you want them to say," said Anthony Capozzi, a longtime defense attorney and former federal prosecutor.
Still, in media advertising, in interviews and in debates, both sides have held fast to the felony conviction rate percentage, as they use it to buttress their respective positions: either that it shows Egan needs to go, as Smittcamp says, or she is running "a well-oiled machine," as Egan herself often says.
In debates or television commercials, neither candidate goes into detail about how they came up with their widely different percentages.
And those details, when explained by each candidate, aren't as clear-cut as it might seem to an average voter not versed in the law.
"The real risk for statistics is always that you are just cherry-picking," said John Sims, a constitutional law professor at the University of the Pacific's McGeorge School of Law in Sacramento. "You frame your question so you get the answer you want to get."
Egan's statistics are from her office's internal record-keeping system. They are based on a calendar year, and take convictions vs. total cases that reached conclusion in a particular year.
Last year, for example, 8,586 felony cases reached conclusion. Of those, 7,928 were convictions, and 658 were either dismissed or the defendant was acquitted of all charges. That's a 92.3% winning percentage, Egan says.
Her numbers are similar in recent years: 88% in 2009, 89% in 2010 and 2011 and 91% in 2012.
Smittcamp, by comparison, takes her numbers from the Judicial Council of California, which produces statistics on various court activities, including felony cases, by county. The state's data is based on a fiscal year.
In the 2012-13 fiscal year — the most recent data available — Fresno County had 4,878 felony convictions out of 9,287 case filings. That's a 52.53% winning percentage, Smittcamp says.
At this point things start to get complicated.
Egan, for instance, counts it as a win if a felony case ends with a conviction on a lesser misdemeanor charge. Smittcamp says that's bogus and misleads the public.
"You cannot call a misdemeanor conviction a felony conviction," she says.
If jurors cleared a defendant of a felony count, but convicted him of a lesser misdemeanor count on a case she was prosecuting, Smittcamp says she wouldn't count it as a win.
"For me, I would go home and cry in my Cheerios for three days," she says.
Egan defends it. If her office gets a conviction — any conviction — there is accountability, she says.
Her prosecutors, she says, can only argue a case and don't have the ultimate say over what a judge or jury may decide on finding guilt on a felony versus a misdemeanor, or how a judge may sentence someone.
Smittcamp's 52.53% felony conviction rate assertion, by comparison, only included felony convictions on felony cases — not misdemeanor convictions on felony cases.
The percentage also was reached by dividing that number into total filings, which is every case filed, and not into cases reaching conclusion.
It's not fair, Egan says.
"It's taking an apple and calling it an orange," she says.
Cases often are filed in one fiscal year, but not settled that same year. So it isn't right to look at "total filings" for a fiscal year, Egan said.
Smittcamp stands by her math. Where does a hung jury — a case where jurors deadlock on a case — fall? If the case isn't retried, the defendant goes free, but it isn't counted as a loss, she says.
Her campaign also used the state data to look at Egan's three-year average, saying that would account for cases that spill from one year to another.
The Fresno County District Attorney's Office felony conviction rate for the three most recent fiscal years ending June 30, 2012 — as a percentage of total case filings — is 54.04%, while the statewide average is 65.6%, Smittcamp's campaign says.
Beyond that, both women cry foul over the source of the other's information.
Smittcamp says Egan's numbers are from her own office.
"Her in-house stats can be easily manipulated," Smittcamp says. "Court stats cannot be manipulated."
But Egan says Smittcamp's court-generated statistics are supposed to determine the court's workload — not judge a district attorney's success.
Shelley Curran, a senior manager with the Judicial Council of California, agreed with Egan that the data collection was designed to determine the court's workload, and added the numbers shouldn't be used to measure the effectiveness of any prosecutor's office in California.
Curran said the data isn't complete. For instance, she said someone may be arrested for a felony violation, but also be on probation. That's also a probation violation. If justice may be served just as effectively by pursuing the probation violation and not the felony, it wouldn't be measured in the Judicial Council's statistics, she said.
And legal experts say that simply measuring the end game often leaves out good or bad decisions made long before a case ends.
What if, they ask, a district attorney's office decides to reject a case that should have been prosecuted? Or vice versa? Should that be counted as a failure? Instead, they said, it likely isn't even measured.
In addition, they say, there are so many chances for something to go wrong in a case, starting with the initial police work.
What's best, several experts said, is to have prosecutors who are free to make decisions that are in the best interest of all parties involved — from victim to defendant to their respective families and the public at large.
"What you want in a prosecutor is somebody who is not overly concerned about batting averages," said Sims, the McGeorge law school professor.
The reporter can be reached at firstname.lastname@example.org, (559) 441-6320, or johnellis24 on Twitter.