Red Sox owner John Henry spent two months “under the hood” of the team’s analytics process late last season, writes Brian MacPherson of the Providence Journal. What he found led to the hiring of Dave Dombrowski who is well regarded as a talent evaluator. Now the Red Sox are poised to pivot away from their past role as Moneyball with money.
Per Henry, “A lot of our advantage was purely financial. We were never as far toward analytics as people thought we were.” The club has spent the second most on payroll since 2002 which helped to mask a multitude of mistakes. In particular, Henry noted an overemphasis on using past performance to project future ability. That could be why the Sox were more bullish about Hanley Ramirez and Pablo Sandoval than other clubs. While those are just two of the most glaring examples, Henry says the same evaluative techniques failed up and down the roster. While it’s popular to quip that Boston should have signed Jon Lester or traded for Cole Hamels prior to 2015, there were too many problems with the roster.
During the Theo Epstein era, the Red Sox had a blend of scouting and analytics, but that balance may have been lost upon his departure, writes Yahoo’s Jeff Passan. However, the shift in strategy could be an overcorrection. In particular, Dombrowski is known for using trades of minor leaguers to bolster the major league club. He’s already executed one high profile move in acquiring Craig Kimbrel. The previous front office seemed to emphasize developing talent, with players like Mookie Betts, Xander Bogaerts, and Blake Swihart representing the first wave of talent.
In an interesting reversal, the Phillies are now making strides to become more analytically proficient. If the Red Sox were the team most synonymous with sabermetrics, the Phillies were usually identified as the most traditionalist club. Now the two franchises are moving in opposite directions.
Director of baseball research and development Andy Galdi will drive the analytics process in Philadelphia, writes Todd Zolecki of MLB.com. The 30-year-old former Google and YouTube employee praised the Phillies’ recent efforts in building their PHIL computer system. Galdi also mentioned the important marriage between scouting and analytics, emphasizing that the two need to work hand and hand.
Galdi believes uncertainty is the missing piece in baseball analysis. Statistics shouldn’t be used to say a player will do X. Rather a range of possibilities should be identified with scouts helping to narrow the focus. It’s important to note that all statistics come with a margin of error. For what it’s worth, sabermetricians have emphasized the importance of error margins for years.
John Henry talks too much
I like giadi’s statement – I think too much trust is placed in analytics based on past performance, people seem to forget that just because a guy as a 4+ WAR it doesn’t mean that he is going to perform even half that well the following year. Analytics doesn’t measure clubhouse chemistry, coaching staff chemistry, marital problems, getting hit by pitches or any number of things that pretty commonly change what happens to a team’s season or post season –
Isn’t obvious that there needs to be a blend of sabermetric analytics and talent evaluation? Best of both worlds?
Sure, but the team that comes up with the best way to blend them, wins.
not to a lot of guys –
Theo always said it was
Pedro Sandoval’s BMI is 34, not 17, as John Henry would have us believe.
Actually, sabermetricians as a group have done a very poor job on uncertainty and the margin of error. Just look at the projections for team and players and show me where the margins of error are?. WAR for example uses one year of fielding data. No margin of error there. Run estimators, also used by WAR , with all their assumptions and estimations? No margin of error. Position adjustments whose source is entirely nebulous? No margin of error there. Reporting UZR or WAR to 1 decimal point provides those numbers with a statistical significance they don’t deserve
The problems with WAR mostly come from people who think it’s the be-all-end-all of statistics… when it isn’t, and no one who actually understands it thinks it is.
Maybe the Phillies have learned a lesson about long term expensive contracts for aging players like Howard, utley, Lee.