The Least-Deserving All-Star Game Starting Pitchers

Much of the time, the pitcher who starts the All-Star Game is very deserving, and is very obvious. It may not be the most-deserving pitcher mind you, but the selection is generally someone good enough that the selection is not worth arguing. This year is such a year: Both Max Scherzer and Matt Harvey rank second in their respective league in WAR, and are close enough to the front runners (Felix Hernandez and Adam Wainwright) as to make the differences negligible. But that hasn’t always been the case. Let’s take a trip down memory lane and look at some of the starting pitchers who weren’t quite as deserving.

Using our first half splits, which we have all the way back to 1974, we can see that — including this season — 54 of the 80 pitchers who started the All-Star Game ranked in the top five of their respective league in pitcher WAR. There were some who were not as deserving, though, at least according to how they ranked against their peers:

A few things jump out right away:

— The manager who most frequently picked someone who wasn’t a top-five pitcher, in terms of WAR, was Tony La Russa. I’ll allow a few seconds for you to register your shock. Twice, he chose one of his own pitchers (in another, less egregious example, he also picked Chris Carpenter over Pedro Martinez. In 1990, he made the single most egregious choice, when he picked Bob Welch over Roger Clemens. Welch would go on to win the American League Cy Young Award that season, on the strength of his otherworldly, but hollow, 27-6 record. Welch starting the Midsummer Classic no doubt helped strengthen the myth of his greatness.

— Speaking of the myth of greatness, we see that sabermetric whipping boy Jack Morris pops up three times here as undeserving. Morris started the game three times, and on each occasion he was no better than the 11th-best pitcher in the AL. It wasn’t always a case of Morris even having better superficial stats either. In 1981, he was 9-3 with a 2.56 ERA. Ken Forsch was also 9-3, but he had a better WAR and also a better ERA, at 2.27. Len Barker, who led the AL in WAR in that first half, had a 2.08 ERA. In 1985, Morris was 10-6 with a 3.04 ERA. That hardly compared to some of his competitors who bettered him in WAR, such as Bret Saberhagen (10-4, 2.78), Ron Guidry (12-3, 2.58) and Dave Stieb (9-5, 1.87). The situation was similar in 1991 — Clemens was 11-5 with a 2.22 ERA, while Morris was 11-6 with a 3.65 ERA. In 1981 and 1091, the Tigers were barely above .500 at the break. Bottom line: Dudes just liked Jack Morris, statistics be damned.

— Clemens was frequently bypassed. He entered the break leading the AL in WAR in seven different seasons, though he only started the game in one of those seven years. He started on two other occasions, and ironically,he ranked fifth in those two years.

Of course, there’s another way to look at this: Just because a pitcher wasn’t in his top five in his respective league’s WAR doesn’t mean he wasn’t a good selection. For instance, Dan Haren ranked sixth in the AL in WAR in 2007, but he was still a mere .4 WAR behind Josh Beckett.

But what about the bigger WAR gaps between the starter and the WAR league-leader?

These were the biggest gaps. Twenty of the 80 were the actual league leaders in WAR. Of the remaining 60, 27 were within 1 WAR, and 34 were within 1.5 WAR. The other 26 you see above.

— Again, a stark picture is painted of La Russa’s failings. He committed three of the five most egregious offenses.

— We also see Clemens as the spurned ace who deserved to start. The three biggest gaps were between Clemens and another player. This was actually not all that uncommon. With a few different pitchers, the pattern seemed to be  they would get their All-Star Game start, and then be passed over in subsequent years. Whether it was  they didn’t want the honor, were injured or were simply bypassed to give someone new a turn, this phenomenon wasn’t isolated to Clemens. Steve Carlton started in 1979, and  watched four others start ahead of him, even though he led the NL in WAR heading into the break. Greg Maddux led the NL in WAR at the break from 1992 to 1997, but he only started the 1994 and 1997 games. He got another start in 1998, though he was just shy of first that season. Pedro Martinez started in 1999, when he led the league in WAR with an otherworldly 7.3. He then led his respective league in first-half WAR in 2000, 2001, 2002, 2003 and 2005 — but he didn’t start any of those contests. Clemens got the nod in 1986 when he led the AL, but despite leading the league again in 1988, 1990, 1991, 1992 and 1994, he didn’t get any of those starts. In fact, in ’94 he wasn’t even named to the All-Star team.

Looking closely at both lists, we see  most of the same names populate them. No matter which way you choose to look at it, several of these starters were simply unworthy. Those who end up on both lists are:

Eighteen pitchers are listed 21 times, with Morris being king of the undeserving and Welch being the most egregious error in terms of both rank and difference. What’s interesting is just how many stellar pitchers are on this list. Guys like David Price, Dwight Gooden, Jim Palmer and Don Sutton stick out, in particular.

When choosing a pitcher to start the game, there are often extenuating circumstances beyond who is the most deserving. The Sunday rule gives an out to many pitchers, including Wainwright this season. And, of course, evaluating by WAR isn’t the only way to decide the best pitcher was in the first half. There are plenty of other metrics from which to choose. But no matter what metric is chosen beyond simple wins and losses, there’s simply no way Welch or Kenny Rogers ends up as a deserving candidate to start an All-Star Game.

Starting an All-Star Game is an honor, and no doubt Scherzer and Harvey are excellent choices. In fact, the choices have been better in recent years: Counting 2013, eight of the last 10 starters were ranked in their respective league’s top five, and four of them ranked No. 1 . Hopefully things stay this way, because in the past they certainly weren’t, and the people who suffer the most are those in the game itself. The All-Star Game may be meaningless in terms of who wins the Fall Classic, but in October, the eyes of the nation are spread out among football, basketball, hockey and baseball. The Midsummer Classic is the time when baseball has the stage  to itself, and it must continue to put its best foot forward in the hopes of growing the game among casual fans.

Finally, because I love you, here’s the whole stats dump.

Other assorted thoughts:
— If you asked me to pick  one manager who correctly named his league’s respective WAR leader in both of his stints as manager, I definitely would choose Ron Washington. Neither guy was from his team. Impressive.
— Pedro in 1999. That’s an MVP-caliber season in half a season, and in one of the highest offense seasons of all-time. Just don’t tell George King or LeVelle Neal III.
— Jeremy Bonderman’s fall from grace was fast and furious.

Other assorted thoughts:
— Harvey’s season is actually 10th best among National League All-Star Game starters. Not too shabby. Maybe next year people will know what he looks like.
— Only Stieb started back-to-back games in this period for the American League; in the NL, there have been three repeat starters: Glavine, Johnson and Maddux.