Projections no guarantees
Metrics can provide insight, probabilities but not absolutes
Metrics- based projections deal in probabilities, something that can seem a little woolly to those who like to deal in absolutes.
So when the ZiPS projections at Fangraphs. com list the Cubs’ starting rotation of Jon Lester, Yu Darvish, Kyle Hendricks, Jose Quintana and Tyler Chatwood a total of 17.3 wins above replacement, it’s what the calculation sees as the most likely scenario out of a range of possibilities.
It tells us the Cubs not only have addressed the issues raised by the departures of Jake Arrieta and John Lackey, but they’ve strengthened a rotation that last season got 10.1 fWAR from Arrieta, Lackey, Hendricks, Lester and a half- season from Quintana, along with another 1.1 from MikeMontgomery’s time as a starter.
Modern front offices deal in probabilities in roster- building, balancing projected value against cost in trades and in signing free agents. Take Darvish and his sixyear, $ 126 million contract with the Cubs.
As soon as the deal was announced, ZiPS developer Dan Szymborski took to Twitter with projections for the life of the contract. The total fWAR was 17.9, meaning the Cubs were looking at about $ 7 million per win above replacement. Normal decline with age is built into the formula, but it still projects Darvish as a useful starter at 2.1 WAR in the sixth year.
That’s below market rates in the last several years. Free agents have been running closer to $ 9 million perWAR. This year’s free agents got less, and better projection tools are a possible factor, along with teams striving to stay below the luxury- tax limit and some saving money for the anticipated bumper crop of free agents next offseason.
Ballclub metrics aren’t precisely the same as those you see on Fangraphs, Baseball- Reference. com or BaseballProspectus.com. Scouts and analytics increasingly work together, and Statcast gives teams more detailed information than is available to fans.
FangraphsWAR emphasizes fielding- independent pitching, in effect assuming results on balls in play will cluster around league average. That might underestimate pitchers who can induce weak contact, something ballclubs might be able to take into account in developing their own metrics using Statcast data.
Nonetheless, the tools available to fans give us more insight into probabilities and the way front offices think than you would get by trying to forecast from victories, losses and ERAs.
For 2018, here’s the ZiPS breakdown for each of the five Cubs starters:
Quintana, 14- 7, 3.19 ERA, 4.9 WAR; Darvish, 13- 7, 3.27, 4.3; Lester, 13- 7, 3.54, 3.4; Hendricks, 10- 6, 3.48, 3.0; Chatwood, 12- 10, 4.24, 1.7.
Darvish’s projection is a step up from his 10- 12, 3.86, 3.5 for the Rangers and Dodgers in 2017, his first full season after Tommy John surgery. Quintana ( 11- 11, 4.15, 3.9 for the White Sox and Cubs) and Chatwood ( 8- 15, 4.69, 1.1 for the Rockies) also are projected for improvement.
None of that is set in stone, or there would be no need to play the season. They’re best estimates amid a range of probabilities. The fun is in seeing which probabilities play out.