MLB by the numbers: Walk isn’t as good as hit, but it’s better than half of one
BY JOHN GROCHOWSKI For Sun-Times Media April 22, 2013 10:17PM
No matter what your Little League coach told you, a walk is not as good as a hit, except in specific situations. A walk doesn’t advance other
runners. It can’t drive home a runner from second base. For that matter, it can’t drive in a run at all unless the bases are loaded.
So to evaluate a hitter fairly, sabermetric stats can’t give walks the same weight as they do singles. Regardless of whether the advanced metric being used is runs created (used by Bill James in calculating Win Shares), weighted on-base average (used by FanGraphs.com in its offensive WAR) or any other sophisticated model, walks are given less weight than hits.
None of those metrics, however, gives walks as little weight as OPS, which gives singles double the value of walks.
OPS is on-base percentage plus slugging percentage, and walks aren’t included in slugging percentage. If we were to boil the on-base and slugging components into one formula, the top half of the equation would be walks and hit by pitches + 2 x singles + 3 x doubles + 4 x triples + 5 x home runs. On the bottom half, all that would by divided by walks + hit by pitches + sacrifice flies + 2 x at-bats.
Hits are more valuable than walks, but they aren’t twice as valuable. That’s a big part of why base runs, runs created and weighted on-base average all correlate more closely to actual runs than OPS does. OPS underestimates the value of on-base percentage, which FanGraphs has calculated as 1.8 times as important as slugging percentage when it comes to producing runs.
We can see the differences in the weighting when we compare hitters with the same OPS but different on-base percentages. One 2012 example involves White Sox outfielder Alex Rios, who had the same .850 OPS as the Diamondbacks’ Paul Goldschmidt. But Goldschmidt had the higher on-base percentage (.359-.334). And by metrics more sophisticated than OPS, Goldschmidt comes out on top in runs created per 27 outs (6.48-5.97) and weighted on-base average (.363-.361).
Two Cubs, David DeJesus and Starlin Castro, tied in OPS at .753 last season. DeJesus had the higher on-base percentage (.350-.327), leading him to rank higher in runs created/27 outs (4.98-4.71) and weighted on-base average (.332-.323).
And to take two of the top hitters in baseball last season, the Twins’ Joe Mauer and the A’s Yoenis Cespedes each checked in with an .861 OPS. Cespedes had a strong .356 on-base percentage, but Mauer was spectacular at .416. Mauer topped Cespedes in runs created/
27 outs (6.66-6.52) and weighted on-base average (.376-.368).
If two players are even in OPS, the one with the higher on-base percentage usually will rank higher in other metrics.
Sabermetricians don’t value walks above all else, any fallout from ‘‘Moneyball’’ aside. And the real world is always more complex than any model of it. But the models that work best and most accurately reflect real-world results give bases on balls more weight than the half-a-hit value in OPS.