Baseball analytics has become a staple of Major League Baseball over the last decade. The analytics movement can be seen on broadcasts, in newspaper and internet articles and, of course, at the Sabermetrics, Scouting, and the Science of Baseball conference. Jimmy Wulf gives us part one of his 2015 SaberSeminar Recap.
From the Sleeper Auditorium at Boston University, it’s the fifth annual “Sabermetrics, Scouting, and the Science of Baseball” conference, a.k.a. the SaberSeminar. I’m stocked with water, energy drinks and granola bars, and I’m ready for some SCIENCE delivered to me wrapped in a delicious baseball cover (like psychological bacon). For you poor souls without a ticket (or bacon), here’s what you missed from some of the key speakers at the two-day event.
Saturday 9 A.M. – Chuck Korb
The founder of the Sabermetrics, Scouting and the Science of Baseball seminar, Chuck Korb, batted leadoff (spoiler alert: he also closed out the day by reminding everyone to carry out their trash). He reminded us we’re really here about cancer as much as baseball – the awesome thing about the “Woodstock of Baseball” is that all ticket proceeds go to The Jimmy Fund every year.
Then he was off to the races with the thank-yous to BU, sponsors, speakers, students, attendees, you, me, half the local media corps and so on. Before Korb ceded the stage, we were given a last reminder that these are serious speakers, so “when you ask them questions, please make them mature questions.”
Saturday 9:30 A.M. – Tyler Tumminia
The Goldklang Group’s (owner of 4 mL teams) Senior Vice President of Marketing dropped in from Charlotte sans luggage and did her level best to steal the show away from the rest of the panel. Between stories of being raised by a 30-year veteran White Sox scout (“when I was playing, he often couldn’t come to my games. So he sent other scouts”) and teasing insider hints of the business of baseball, it was easy to see how she’d found so much success in the slightly more rough-and-tumble world of minor league marketing. Her influence led to a Baseball Scout Hall of Fame in each of Goldklang’s ballparks (including the Fort Myers Miracle, for the Sox-fan-snowbird with time on their hands). The most striking stories she told were how, even with a family connection, she couldn’t get a job in baseball, and the lengths she was forced to go to get in the door. They pay you peanuts and yet everyone seems to want to work there – baseball is, as Tumminia said “nutty, and you have to be nutty to work in it. We’re all crazy”. Also interesting was to hear how minor league teams have actually had their in-game marketing impacted by MLB’s pace-of-play experiments. Meaningless to the big clubs, but part of the DNA of the low-level operations.
Saturday 10:40 A.M. – Tom Tippett
Tom Tippett may be the real reason that Dave Dombrowski has to answer questions about being anti-numbers, and the House That Epstein Built is seen as the eastern capital of the Holy Moneyball Empire. Between 2001 and today, all 30 MLB teams have had to build an in-house computing analytics group out of whole cloth, starting with no institutional expertise. That’s incredibly hard, and there’s a tremendous GIGO effect in using that data to make decisions. I’m sure there are 30 interesting stories to be had about the paths teams traveled to get to the capabilities they have today. Detroit’s was obviously not as well-executed as most. With Dombrowski being late to the party in the mid-aughts certainly contributing to it, that was also a decade ago and the real story, I promise, has a lot more devils in the details than you could fit into a First Take segment. Either way, he now has the same resources that let the Red Sox take a shortcut to the cutting edge: he has Tom Tippett.
Tippett broke down the 2015 Red Sox season with an ironic take on the numbers full of self-skewering jokes which landed squarely with the crowd. Tidbits from that talk and the ensuing Q&A:
- He believes the loss of both Vazquez and in particular Hanigan, early in the season, were major factors in the team’s underperformance out of the gate.
- Moving Hanley to LF was based mostly on historical analysis of how middle infielders do when moving to the outfield. There’s no way to predict or model how a specific individual will do.
- He let drop that he’s been working on two software projects big enough to keep him out of most of last offseason’s free agency discussions. One of them has to be around the new Statcast data streams. Any guesses on the second?
- Important related points he made were around evaluating a roster in terms of uncertainty and risk factors rather than projected numbers, and the difficulty of breaking rookies into a win-now team. Incorporating more than two rookies into a lineup at any one time is definitely something teams think heavily about, even when the alternatives are expensive free agents. It’s not something you really ever see considered in fan lineup discussions.
- Twice, when referring to the bright future represented by Boston’s young prospect core breaking into the bigs now, he chose ‘2017’ instead of ‘2016’ to describe the starting point of the golden age window. Looking at how much uncertainty exists in the pitching staff next year, that seems reasonable, but will Red Sox Nation have the patience to endure another barren October, or will the new GM trade the nascent core for an established one? The juicy potential narratives abound.
Saturday 1:15 P.M. – Curt Schilling
It’s hard to reconcile the dueling mythologies of Curt Schilling, the person, without seeing him in this exact sort of setting – mic in hand, audience tuned in, firing from the hip with a leaky filter and three decades of insider knowledge. There he was, in camo pants and combat boots, leaving the crowd rolling in the aisles like an Original King of Comedy with line after line. It’s unlikely Schill’s favorables/unfavorables with the SaberSeminar demographic would make a pollster smile, but his ease with speaking and natural charisma remind you he was an alpha male in a sport of alpha males for a decade, and also put on a suit and got serious people to buy into a business vision. Some of the things he said would perhaps remind you of the turbulent clubhouses left in his wake, and the smoking crater his vision left in the Rhode Island landscape. Big Schill and the Big Lug, they’re both real and true; like everything else, it’s just a matter of timing and perspective.
As you’d expect at the ‘Sabermetrics Woodstock’, Schill dove into his own efforts as a player to be on the cutting edge, from a post-revolution perspective. Describing lugging around VHS tapes for video work in the Nasty Boys Phillies clubhouse, he quipped “They looked at me the way athletes look at you!” He noted there was a fantastic amount of data available now, but felt much of it might have been useless to him as a player. For instance, he couldn’t make his curveball bend any different just because PITCHf/x told him it might be more effective. Throughout the seminar, the challenge of breaking down complex data or conclusions into consumable, actionable instructions was a recurring theme from players, execs and analysts alike. With his typical tact, Schill may have put the finger on the problem most precisely, noting the simple fact that verbal or written communication of the numbers between the analytics staff and players has both language and status barriers. “Saying a guy ‘sucks’… it’s almost like, as a player, I can say it, but you can’t.” In a perfect world, should players be able to put that view aside for their own good?? Yeah, probably, but they don’t, and talking at them with all the logic in the world isn’t going to change that. It takes showing results, building trust, and a lot of time.
Schilling was also much more interested in numbers that told him about tendencies than numbers that explained outcomes. “I wanted to know when to throw a strike if the hitter was taking and a ball when he was swinging.” Things like a batter’s heat map with red and blue zones weren’t relevant to him specifically or how he approached a batter. There are a lot more formulas today that can take the data and put it into the kinds of terms he wanted. One of the biggest challenges he had was finding data on the strike zones of specific umpires, and he thinks that’s still an underrated element (also the 1-1 count, a shout-out to my personal favorite). With his more recent experiences, he also discussed how he brings his knowledge and data into the broadcast booth, showing an understanding of the need to balance ‘dry’ information with connecting to more casual fans, and explaining “I always thought people would want to hear us talk about the game the way we’d talk about it on the bench… not telling you Johnny Cueto’s basic numbers”.
Because Curt will always be Curt, there was no shortage of tales from playing days on display, and he pulled few punches. There were stories about finding out Bonds really could pick up his splitter before he let go of it, and being owned by both Todd Helton (“The one guy I never wanted to see at the plate”) and every 5’8” punch-and-judy middle infielder in the Show (“when I was a young pitcher, I threw a lot of high fastballs, and if you’re 5’8” in the majors the one thing you’d better be able to hit are high fastballs. Everything’s a high fastball to them, but when I was young, I was stubborn”), but the clubhouse tales were the biggest crowd-pleasers. Asked about team chemistry, Schill spoke at length about the importance of hidden clubhouse leaders, who may not be stars, but provided the day-to-day spirit of the team in every winning clubhouse he was a part of. Folks like Doug Mirabelli, Gabe Kapler and Todd Stottlemyre were mentioned, as well as 2004 Red Sox sparkplug Orlando Cabrera. Schill surfaced a moment down the stretch where Manny was “going through a phase” and asked out of a game. O-Cab went over to where Ramirez was changing at his locker in the trainer’s area, and when Manny said his leg was bothering him, Cabrera replied along the lines of “B******. You’re messing with my playoff paychecks now. You’re playing, and if your name isn’t on the card in 30 minutes, we’re going to have a fight.” Manny played, the Sox won that night and the rest is history.
Saturday 1:45p – Ben Cherington
John Farrell and Ben Cherington were both scheduled to speak on Saturday, timing made downright insane by the events of the past week. Schilling stepped in as a guest to replace Farrell as the Sox manager fights with his own cancer, but Cherington’s troubles are of a more existential grade. The former GM’s desire to support this also-fighting-cancer event meant he swallowed his troubles and kept his date to stand for an hour of questioning from the masses. Ben clearly wore the last week with a drained demeanor and a gallows edge to his humor – after being described in a question as a ‘successful GM’, he replied “that’s an interesting term”. The sharpest edge he showed was in defense of Hanley Ramirez the player, observing that some of the criticism aimed at him this season seemed to carry an unnecessarily personal tone, and rejecting the idea that lack of effort was in any way part of the issue. “There’s no player in baseball who doesn’t want to go out there and do well. (Hanley) came here wanting to be a Red Sox and wanting to play LF, accepting the challenge. He knows it hasn’t gone well”.
Both in the context of Ramirez and in the overall state of analysis in the game, Cherington underscored the uncertainty of the projections and conclusions they are dealing with. Quite predictably, questions were asked about “saber-oriented” versus “scouting-oriented” front offices, and just as predictably he rejected the binary premise. Cherington believes “There are rarely people in the front office, especially at the highest levels, who are purely one or the other. All of them are a combination of things that defy simple reputations. Everyone’s trying to manage data and make decisions. One team may weigh things more heavily than another, but in my experience no GMs fall into one or another bucket.” He expanded on that in a question about whether stats investigate the “whys” enough, reiterating the need to include multiple perspectives in a conversation. “The more questions asked on both sides of the line will produce better analysis.” He stressed that the further away you get from the MLB level, the less hard data becomes available, and even within a single level there is a tremendous amount of environmental variance, so context and observational data will always play a critical role there.
The crowd seemed dissatisfied with the responses from both Ben and Tom Tippett on Ramirez that they were essentially making a bet based on historical data about middle infield-to-outfield transitions, and it being impossible to make an accurate guess about a specific player, being a unique thing to them that’s never happened before in history. Whether they end up being right or wrong, we on the outside want to believe there’s a more quantitative, definitive conclusion teams work from when they make decisions, more than just ‘making a bet’. The more we learn, the more we seem to realize there isn’t. Cherington demonstrated this again with Pablo Sandoval – there were no secrets to be found in park factors analysis to give the deal a thumbs-up or down; they were just looking to fill a black hole and out of the players in the right general age and production range, he’s the one they got. The Sox front office inquired about Josh Donaldson and were told he wasn’t available, only to then later learn about him moving to Toronto. For those who might want the Sox to act more boldly in response, Cherington’s belief is that the times they were most likely to make mistakes are “when we got in a rush to ‘do things’.”
The burden to make these bets a little smarter is on everyone who cares about the game. For this audience, maybe the most important thing said at the whole conference was by Cherington as his opening statement, and I’ll quote him as best I can here:
“One thing that I want to say to this group, we have an incredible opportunity to really make a difference. As everyone knows, there’s so much more and new types of data coming quicker and quicker. It’s exciting, a lot to learn, but because it’s so new, this community has an enormous task ahead, and it has to be done well. It used to be you could research years of data to come to a conclusion, now we’re asking for new analysis in a day, and that’s dangerous. It’s so hard to deliver good analysis really quickly, and we just have to be careful and work harder.”
And later, in response to a question about proprietary data:
“I think a lot of the data is available to everyone. Then it depends on who is doing the analysis and building on the data. You can’t assume every data stream is precise, you need to do filtering on it. A lot of us have access to the same stuff.”
I believe the gauntlet has been thrown.