Jon Lester and starting pitcher assessment

Occasionally, I enjoy asking questions I haven't received an acceptable answer for. Last season, at some point, I asked a rather basic one. For people who are very "anti" regarding the question: "Do pitcher wins and losses having any realistic value?", what should I post as an introduction for the starting pitchers for Dominican League games? It's not that noting a pitcher is 1-2 with a 4.17 ERA clarifies much, but posting something as a run-up to game time is traditional and at least marginally useful. This serves as a backdrop for a Jon Lester discussion I had last week.

Lester is a bit more divisive to discuss than you might think. Nobody is trying to diminish his value in the years around the World Series title. I doubt many would dispute his "bulldog mentality." He takes the ball when it's his turn, and battles. Last year, he had a statistic that was even more stark than I thought it would be. In his 31 starts in 2019, he recorded at least one seventh inning out just six times. That isn't to say he was terrible, or that he isn't likable. Only in about 20 percent of his starts did he let the bullpen have any realistic modicum of rest.

Again, this isn't to speak dismissively of Lester, the Cubs, or any group in particular. It calls into question my thought from earlier. For people for whom "10-8 with a 3.66 ERA (or whatever)" isn't descriptive enough, what would be better? The Quality Start is one such metric. It tells how many times a pitcher went six innings or more, and surrendered three earned runs or less. Not ideal, by any stretch, but it's a metric.

A while back, I noted I preferred granting a QS for a player who tosses six and is at three earned or less, but allows a fourth earned run (and possibly more) later. To a slight extent, I'm recanting my assessment. It's not that we have too many ways to assess starting pitching effectiveness. We have far too few.

Assessing the starting pitcher isn't about the elite (Jacob deGrom is good. We know that) or the statistically insignificant (a starter making his second MLB appearance will almost always be a difficult player to assess accurately). For the twenty-five percent of starters who top the list, and the quarter that might belong in Triple-A, the numbers are a bit easy to justify. How, though, should fans, and baseball scribes as a whole, assess the half of the league starters that are somewhat interchangeable?

For me, the first step is to try to back away from the constant barrage of the emotional side of the game. Assessing Lester versus a third or fourth starter on a middling rotation shouldn't be about clowning one or the other pitcher. It shouldn't be primarily on being dismissive of "someone you don't like." Is Lester better or worse now than, say, Gio Gonzalez, and how can we know?

With hitters, it's easy. Line up the OPS+ with a defensive ranking mechanism, and it's numbers-oriented. Fielding Independent Pitching is fine. Game score assessments are fine. But nothing gets even remotely close to in the day when you had a 17-6 with a 2.74 ERA against a 11-18 with a 4.16. It didn't swing the game that much, but the fans looking at the matchups had an assessment tool, however inaccurate.

What I'd like to see is a larger battery of pitching assessment options. Some starters aren't expected to go even three innings. Some are needed to go six, but rarely exceed five. Lester is much better at five or six than seven.

For "openers," how many of them allow no runs in their outing? Make a column. How many get their six outs in seven or less hitters? Make a column. How often does an opener allow more than one run in his outing? That would seem toxic.

For a traditional starter, an eight-inning start seems valuable, regardless of the runs surrendered. Make a column. I'd like an easy column on how often he recorded 21 outs, as well. How often did he allow four runs or less, regardless if earned or not. Add yours below. Get twelve columns for starters, and by the time they get to 20 starts, those dozen metrics will give a better read than an FIP. He might not be likely to toss a seven-inning shutout, but he is able to get an out in the seventh 47 percent of the time. That gets to be predictive, which is what statistics are supposed to be. Not methods for defending or criticizing a player based on pre-existing biases.