The discussion over the past several days has focused on two problems that together create a perplexing dilemma. On the one hand, many commenters are very concerned about “overuse” of the starters (Problem #1). On the other hand, a lot of commenters are very concerned that the bullpen is ineffective (Problem #2), which, of course, in the absence of any alternative, means that starters are going to be asked to go deeper into games than they would otherwise, and then would be optimal (allegedly) for their health and future effectiveness.

These discussions have taken place on the MLB Network, ESPN, and most D.C. media outlets that cover the Nationals.

Thanks to @Buster_ESPN as always! Talked Astros (CC @ac13alex), Nationals bullpen low usage & Bellinger (& @D_Ross3)https://t.co/WQilPphWzl

— Sarah Langs (@SlangsOnSports) May 23, 2017

The discussion raises a whole bunch of questions. Are the Nats starters’ high pitch counts responsible for their reduced effectiveness in recent weeks? Will it have an effect later in the season? Is Dusty Baker pushing them too hard to avoid using the bullpen? Will the bullpen lose more games if it’s used earlier in games? I think we all recognize that these questions just can’t be answered with any certainty, if they can be answered at all.

But one thing I thought we could try to analyze is what the impact would have been if Dusty had enforced a hard cap on pitches by a starter, which would be the solution to Problem #1. In other words, what might the cost have been of pulling starters sooner, and what might it be in the future?

In this analysis, I assumed that 100 is the number of pitches that pretty much any starter ought to be able to make without ill effect. Perhaps that’s not the case, but I had to draw the line of Problem No. 1 somewhere.

Personally, I also think that a 110 pitch limit is not significantly less protective of the pitcher. It’s also a cushion for a manager who thinks a pitcher can go another full inning and not pass 100 and, for quite understandable reasons (including Problem #2), wants the starter to complete the full 5th, 6th, or 7th inning even if that means going 105 or 108 pitches. So I also separately counted the games where a starter went over 110 pitches.

Finally, I tallied the games where a starter made more than 100 or 110 pitches that were won by 3 or fewer runs. These are the games where removing the starter prior to the last inning he pitched in those games could very well have had an impact on the outcome, in light of Problem #2.

So how many games might have been affected by these parameters? Take a look at the chart below:

The first thing to notice is that this is really an issue only with our Big 4 starters. All have pitched more than 100 pitches in well over half their starts, and more than 110 pitches two or three times already. We’ve had a pitcher go more than 100 pitches in 30 out of 44 games so far. Twenty-eight of those games were started by Max Scherzer, Stephen Strasburg, Gio Gonzalez or Tanner Roark. To limit those starters to 100 pitches or less would be a big change.

In addition, the effect on game outcomes of such a limit could be significant. In just 14 of our 44 games so far (32%) did our pitchers make less than 100 pitches. Eight of those games were started by our Big 4, and 6 by our other starters who filled the No. 5 starter spot (Joe Ross, Jacob Turner, A.J. Cole, and yes, Jeremy Guthrie.)

In 13 of the remaining 30 games, the Nats won by 3 runs or less. Those are games where giving the bullpen the ball an inning earlier might have been significant.

A hard 100 pitch limit could be accomplished in a few ways. Dusty could “hook” a starter, as Davey used to say, if he were at 85 or 90 pitches at the conclusion of an inning. That would allow a reliever to start the next inning with no men on base. Or he could remove a pitcher in the middle of an inning after an at bat in which the 97^{th} pitch is thrown. In either case, the bullpen would be used earlier than in 30 out of 44 games so far this year.

So how much of a risk are we willing to take to protect the starters’ arms and conserve their strength for later in the season? If just 5 of those 13 close games so far went the other way because of relievers entering the game one inning earlier, or in the middle of an inning, the Nats would have a record of 22-22 rather than 27-17. Would you have been OK with that record in order to save our most valuable arms?

Another approach would be to make 110 pitches rather than 100 the absolute limit. This would give Dusty a cushion to let a starter begin his final inning with 85-90 pitches already under his belt. It looks like such a rule would have a smaller effect on the team’s chances of victory, at least based on what we’ve seen this year. So far, only four wins might have been in jeopardy if 110 pitches were the absolute maximum for a starter.

One was a Scherzer start against the Braves where Max left with a 2-0 lead after 7 innings and 113 pitches, and the Nats ended up winning 3-1. If the bullpen had pitched the seventh, would the Nats have still won the game?

The others were three games started by Stras. In the first, he left after 111 pitches and 7 innings trailing the Phillies 2-0, but the Nats came back to win in extras 3-2. Would the bullpen have kept the game as close as Stras did? In the second, he completed only 6 innings on 119 pitches, but left with a 4-0 lead. The bullpen held off a Phillies rally to win 4-2. Would the Phillies’ comeback have been successful if they faced a reliever instead of Stras in the 6th?

Finally, against the Braves on Sunday, Stras left after 7.2 innings at 119 pitches. He made 25 pitches in the 8^{th} and gave up the only 2 runs the Braves scored in the game. Would a bullpen starting that inning with Stras leaving at 94 pitches have pitched a clean inning or given up more runs than Stras did? We’ll never know.

Based on what has happened so far this year, it seems to me that the marginal value of leaving a pitcher in to pitch more than 110 pitches is limited. The Nats’ pitchers have done it nine times, with a need in only four of those games to try to minimize use of the bullpen. Leaving Roark in for the 6^{th } inning against the D-backs when he already had made 100 pitches, and then letting him pitch the whole inning so he ended up with 125 pitches, is probably the worst decision Dusty has made this year. I think a hard limit of 110 pitchers would make some sense.

In contrast, it would be a big change, and a significant risk to the results of games, to impose a rule that a starter should come out after 100 or nearly 100 pitches.

What do you think?