Information powered Pirates to playoffs
When Travis Sawchik started covering the Pirates for the Pittsburgh TribuneReview in 2013, the franchise had a longer stretch of futility than even the Blue Jays. Not only had the Pirates missed the post-season for 20 years; they hadn’t even had a winning record in that stretch.
Then, with 90 per cent of the roster returned and one of the lowest payrolls in the majors, the Pirates won 15 more games than they had the year before to secure their first playoff berth in more than two decades. Sawchik, like many other observers of Major League Baseball, wondered: “How?”
His book, Big Data Baseball: Math, Miracles and the End of a 20-year Losing Streak, aims to answer that question.
Desperate to improve, but without the financial might of bigger markets, the 2013 Pirates had to be experimental and open-minded. Without the cash to spend on sluggers, they focused on run prevention and three strategies in particular: radically increasing their use of defensive shifts, signing the best freeagent pitch-framer (current Jays catcher Russell Martin), and teaching their pitchers to become ground-ball machines.
But Sawchik went beyond the numbercrunching and found a Pirates organization that cultivated a culture of collaboration and respect, where creative ideas could flourish between both the old and new schools.
Big Data Baseball reads like a sequel to Moneyball, Michael Lewis’s seminal book — later turned into a film starring Brad Pitt — about Billy Beane and the 2002 Oakland A’s, who challenged baseball orthodoxy in search of market inefficiencies. Sawchik didn’t have the same access as Lewis, but he similarly pulls back the curtain on the analytically inclined Pirates and the ingenuity with which they ended their 20-year postseason drought. But his book is also about how the flood of data in the last half-decade is not only changing how we think about the game, but also how it is played.
Sawchik, who still covers the Pirates for the Tribune-Review, spoke to the Star about the book, which comes out Tuesday. The interview has been edited for clarity and space.
Firstly, the people of Toronto would like to know what the Jays can steal from the Pirates so they too can end a 20-year playoff drought?
(Laughs). Well, every team, except maybe the Phillies, have at least one data analyst, if not a small army at this point. So every team has information similar to what the Pirates have. I think the Pirates’ advantage was in getting this information onto the field.
There’s still somewhat of a barrier between the traditional coach/player and the front-office analyst, and I think the Pirates and (manager) Clint Hurdle deserve a lot of credit for creating a culture where there was collaboration and respect between both the new school and old school that allowed for some of these ideas to be implemented and adopted in Pittsburgh.
This wasn’t just a top-down conversation. The players and coaches helped refine the data. I think that’s where the creativity came from. It led to better questions and better answers and better information.
The Pirates’ collaborative culture seems to have been as much a key to their success as the strategies themselves. Why haven’t other teams been able to create that kind of alliance between the front office and the field?
Not every player bought in. A.J. Burnett, for instance. He hated shifts for the whole season. He had a very public blow-up with shortstop Clint Barmes. But I think part of it was having the right players, like Barmes, who was open-minded. The coaches felt if they could get the middle infielders to buy in, they wouldn’t have a problem ratcheting up defensive shifts 500 per cent like they did.
I think it’s personalities. Hurdle and Dan Fox, the lead quantitative analyst . . . Hurdle didn’t trust this information at first and he didn’t trust Fox when he came aboard in 2011. But they interacted more, they talked about a lot of things outside baseball; they had similar interests in military history, they’re both pretty devout Christians. That allowed the trust to increase. Fox believes the more time you spend with someone the more trust you’ll have. They spent more and more time with each other and I think that allowed for the co-operation and collaboration to blossom.
I think that’s where personality comes into play. You can have the best data, but if you don’t have the right personalities and you don’t build a trust level it’s never going to see the field.
You write that Hurdle was initially resistant but it was a combination of “curiosity and desperation” that forced him to change. What about his personality led him down that path and not one where he dug in his heels and refused to change?
I think Hurdle, while he projects a sort of machismo baseball coach personality at times, I think he is also by nature a curious person. His office is always littered with different kinds of books. He’s a voracious reader. He really is a smart guy. I think he only had one B in high school and that was in driver’s ed.
So he’s a bright guy who’s curious and, I think, if he’s open to an idea that’s better than his, he’ll listen, he won’t just dictate. The great questions lead to the best answers and I think he was open to hearing questions that challenged his assertions. While maybe there was some initial resistance, that curiosity, that willingness to listen helped break down some of those traditional barriers.
Similarly, a lot of the players didn’t buy in until they saw the data visually. Was that a breakthrough in terms of earning the players’ respect and buy-in? Travis Sawchik’s new book looks at how the Pirates ended a 20-year streak of missing the post-season through a data-driven approach.
That’s something that kept coming up when interviewing the players and the analysts. In spring training Fox and some of the other analysts would try to make themselves available to players and through those interactions and conversations they learned the players understood things visually very quickly.
That makes sense when you think of a major-league hitter who has to react to a 95 m.p.h. fastball and has done all this visual learning in his career. It makes sense that maybe this guy isn’t interested in a spreadsheet of numbers, but if you show him a heat map that might sink in more quickly and it might be absorbed better.
That was key. The more you can democratize the data, the better chance you can make an impact in the field.
Russell Martin was specifically targeted by the Pirates for his pitchframing abilities and they went after him aggressively in free agency. What was his approach to the data?
Martin is an expert pitch framer, but he didn’t know his numbers. He didn’t know he was one of the best pitch framers; he just understood the process. It was taught to him, he had the physical gifts to maximize the lessons he was taught, and he’s just a good receiver. But he didn’t know he was worth 20 runs above average or whatever, he just knew he was good at it.
He was interested in data that made sense to him. He wanted to see what sort of hitters were panic hitters, which hitters would chase with two strikes, which hitters were more patient. He wasn’t concerned about every aspect of a player, but he was curious about some things he thought really mattered and he would run data on that.
He was still instinctual in terms of pitch-sequencing. He thinks that’s still important and I don’t think we’ve quantified that role as hobbyists or in the industry, but he thinks it matters.