What recent history tells us about championship-caliber pitching

What a week. If it wasn't obvious already, the Yankees starting rotation has demonstrated in recent series that it is in desperate need of reinforcements. To make matters worse, the set of attractive trade candidates is shrinking by the day. Marcus Stroman was sent to the Mets yesterday in a head-scratching deal. And surges by presumed sellers like the Giants (who should sell) and Indians (who certainly should not) have removed further candidates like Madison Bumgarner and Trevor Bauer from the trading block.

Fans are rightly frustrated by the recent performance of the team's starters and the inaction of the front office to solve a glaring issue. But there is still time for Cashman to make a move and shore up a shaky pitching staff before we reach October. While of course everyone wants to bring in a top line starter like Noah Syndergaard or Max Scherzer, the Yankees deadline moves will likely be much more modest.

That got me thinking about what a World Series winning pitching staff really looks like. What are the characteristics of recent champions that the Yankees and Cashman can look to as a model for their own staff this fall?

In order to do this, I looked at the five most recent World Series champions to try to draw some lessons from the most successful teams of recent years. I chose five years for two reasons.

One, I think it roughly tracks the period in which pitching paradigms have shifted away from workhorse starters and towards stacked bullpens. Two, it also covers a period in which advanced metrics have come to dominate front offices, and thus the attitudes and decisions of baseball executives is most similar to today's game. And most importantly, it allows me to avoid analyzing two Red Sox championship teams, which I don't have the will to do after the events of recent days.

So with this in mind, let's dive into some of the key similarities between recent championship pitching staffs.

The playoff bump is a myth

It seems like every year, commentators love to talk about how much harder it is to score runs in the postseason. Among fans, I've also heard a narrative that, because every game is so important, managers are more willing to take risks and make unconventional moves with their bullpens that reduce overall runs scored. This might lead one to believe that playoff teams can outperform their regular season statistics in a short, must-win sequence of playoff games.

But there is next to no evidence to support this claim. While it may be true that manager usage of pitchers differs drastically in October, there is no marked shift in playoff performance among pitching staffs relative to the season as a whole. Each of the five most recent world champions were in the top 10 in the league in ERA during the regular season. And none of them managed to outperform their regular season ERA by greater than 0.75 runs in the playoffs.

As a lover of sabermetrics, I am far from suggesting that this is definitive proof that you can never markedly improve relative to your regular season stats. In fact, the small sample size of the postseason would suggest one of the next champions may well do just that. I simply use this as evidence that an expectation of sudden greatness is misguided.

World Series winning teams have great pitching staffs that demonstrate their competence over the course of an entire 162 game season. They then tend to carry that same quality in the playoffs. There is, in recent history, no suggestion that you can skate by with a mediocre rotation and work managerial magic in the postseason to glide to victory. This is by no means groundbreaking, but does suggest that there are no shortcuts to building a winning team.

Aces matter, but deep rotations matter more

There is also a prevalent belief that an ace is critical to winning a championship. I certainly won't dispute that notion out of hand, given that four of the five most recent champions had someone in their rotation that qualified as an ace. Madison Bumgarner, Kyle Hendricks, Justin Verlander, and Chris Sale were all indisputably aces in the years that their teams won the World Series.

But these pitchers averaged roughly 32 IP in the playoffs between them, even with Bumgarner's outsize role in the 2014 playoffs for the Giants. While each was pivotal to their team's success in their own right, it is hardly the norm to have an ace single handedly drag a team to a title (again, Bumgarner is the exception). Their impact is limited by feasibility constraints, since they can start at an absolute maximum 5 or 6 games all postseason. And in such a small sample, underperformance is a risk too. Both Sale and Hendricks were far from dominant in their limited appearances during their teams' championship runs.

Instead, the recipe for success is driven by having four highly competent starters who can keep the team competitive in every game. All five of the teams had the luxury of deep rotations filled with reliable arms, simultaneously keeping the bullpen fresh and avoiding non-competitive games. There is no precedent in recent years of a top heavy rotation compensating for the significant shortcomings of a disastrous back end of the staff. While it may allow a team to steal a series, or even two, the grueling postseason process weeds out the thinnest rotations and exposes them by the time they reach World Series.

Even the 2015 Royals, who had the weakest and most inconsistent starting rotation of the set, allowed more than 5 runs in a game only four times throughout their playoff run. Edinson Volquez, Johnny Cueto, Yordano Ventura, and Jason Vargas each fought to keep games competitive through five or six innings. This allowed their bullpen to lock down tight leads and give their offense a chance to come back if they were trailing. While none were standouts, none were disasters either, reducing wear on bullpen arms and generating positive value for their team through their relative consistency in October.

Starting and relief pitching are substitutable, but only to a point

I leave this header intentionally vague, because I think it is true in a couple of ways. Most obviously, you can compensate from subpar performance in one facet of your pitching staff with excellent performance from the other. But if either the bullpen or starting rotation is consistently struggling, even perfect performance from the other group cannot save you.

The necessary balance is evidenced by the fact that, of the five sets of starters and relievers for these teams, only one ranked in the bottom half of the league in ERA. Again, this was the 2015 Royals, whose rotation was ranked 22nd but was just 0.20 behind the median team that year. And their bullpen was the best of the bunch, coming in with a ridiculous 2.72 ERA and multiple elite arms that made the difference for them in the playoffs.

The other interpretation of this header is that starters and relievers can be literally substituted to some extent in playoff games. Teams have experimented with openers, bullpen games, and the use of starters as relievers in recent years to varying degrees of success. The notion of openers and bullpening is relatively new, so it is too soon to judge the efficacy of these strategies through an entire postseason. But many of these championship teams have used starters out of the bullpen in close games to great effect.

Often the players employed in this role have dynamite stuff, but don't always demonstrate the consistency to be trusted with starts in playoff games. Think Lance McCullers and Nathan Eovaldi, two pitchers that have plus-plus pitches but have had careers plagued by an inability to successfully navigate the third time through the batting order. By using these pitchers in relief appearances, their managers were able to avoid lower level relievers in key situations and plug in elite arms to get key outs late in games.

But again, this strategy fails without the necessary depth in the rotation to sustain this usage pattern through an entire playoffs. While employing a starter this way might help seal a victory in one game, it may necessitate a less effective starter stepping in during a future game, which can have a trickle down effect through the whole pitching staff. Each of the teams that have employed this tactic have done so in moderation, and had the backstop of deep, high quality rotations to hedge against overuse of bullpen arms. Teams should always play to their strengths, but this shifting of burden can only go so far if one half of the pitching staff is an active liability on the mound.

What it means for the 2019 Yankees

So how should this information inform the Yankees approach to the trade deadline this year? I want to emphasize that this article is less scientific than most of the stuff I write, and relies more on general trends than hard causal relationships. But I still see important lessons the team can take in their approach to deadline moves this year.

Of the teams mentioned, I see the closest similarity to the 2015 Royals. They also had only a mediocre rotation but a truly excellent bullpen. In October, Boone will know he has the ability to shorten games with his relievers. But if games are consistently out of reach by the middle innings, as was the case in last year's ALDS, the team will be doomed to another year as also-rans.

Contrary to popular belief, I think the evidence outlined here suggests that multiple quality arms, not one ace, is the surer solution to the Yankees woes. They currently have just one consistent and reliable starter in Domingo German, who also may be facing an innings limit or risk of burnout given his youth. In an ideal world, Masahiro Tanaka and James Paxton would refind their form as solid middle rotation arms, and Luis Severino would return like a white knight in his ace form from last year's first half.

The front office should not and cannot take that gamble if they are serious about winning this year. Given the set of options available and the reality facing this team, the Yankees should bring in two starters who can ensure that the rotation can keep the team competitive throughout a World Series run. My picks would be Matthew Boyd and Robbie Ray, two young and controllable starters whose peripherals suggest a penchant for excellence.

This approach would have a variety of benefits. In a best case scenario, this would give Boone the ability to move some of his starters to a bullpen role in the playoffs. The best four starters would have reduced pressure, because the already strong bullpen could be shored up by the likes of a limited German or post-injury Severino who have experience in those roles already. In the most likely case, the team would have insurance for future injuries or continued inconsistency from guys they currently would have to rely on in October. And in a worst case scenario where the team does not win this year, these guys would continue to be young and cheap fixtures in the rotation for years to come.

Regardless, there is not scenario in which adding a bullpen arm or 2018 Lance Lynn is adequate for this team to win a World Series this year. They need to rectify the imbalance between the two halves of their pitching staff, and produce a set of four reliable arms at a minimum as they move into the postseason. I do not subscribe to the philosophy that only an ace can put this team over the top. But two legitimate MLB starters should be added by July 31 to put this team in the best position possible to win their 28th championship.