Consistency in Starting Pitching Key to Repeated Playoff Success

I tweeted a lot of this, and then realized it would be better if I wrote a post on my blog.  :^)

Andy Baggarly tweeted:

Reds were NL's highest scoring team last year. Bumgarner facing their A lineup. And he's thrown 3 perfect innings thus far w/4 strikeouts.

Madison Bumgarner handling Reds easily is prime example of what I've been blogging about for years now, pitching is better than hitting in terms of consistency.

Ace Starting Pitchers' Consistency
That is, for ace level pitchers, consistency is something you can rely on relative to hitters. It goes back to the old metric, the Quality Start, and how many a pitcher gets. I've been studying PQS, Pure Quality Start, a new saber version of the Quality Start created by Baseball Forecaster, on my site, and if you look at starting pitchers, you see that the elite aces consistently (again, relative to hitters) have quality starts. 

Consistently a large percentage of starts, 50% minimum by my eyeball for the best starters, the best, elite Ace starters like Tim Lincecum in the 70%+ range, are quality starts for the best pitchers.  That is the closest you can get to consistency in baseball.  Hitters go on hot and cold streaks, but even the best hitters can have a poor week or two or three.  That is what dogged Barry Bonds reputation for years until 2002's playoffs.  And when a series is over in a week to 10 days, any hitter can be rendered impotent by the opposing team.

Consistency in the short term, however, is much more likely with ace-level starting pitchers like Lincecum, Matt Cain, and looking good for Bumgarner (and hopefully Zachary Wheeler once he develops fully).  When your team can count on you to throw a quality start every 2 out of 3 starts (or 67% dominant starts - DOM - per PQS terminology, only elites do that), even against good scoring teams, that is very good reliability and consistency.  If you have a rotation like that, you can count on a quality start in 3-4 out of 5 games, 4-5 out of 7 games.

And teams compile a great record in DOM games.  Baseball Forecaster compiled stats on DOM and found the ERA to be 2.39.  As you can see in my study (side bar) about pitching wins and losses, the Giants were 65-23 when they held the other team to 3 runs or less and in the NL, 908-278.  It is not like basketball or football, where the best teams can win 90% of the time, but that works out to 74% of the time (65/88) for the Giants, 77% of the time for NL teams, which is dominating for baseball.

Of course, that winning percentage assumes some average to bad pitchers in the mix for the other team, but at minimum, it sets a very high standard for the opposing team to match up with the Giants playoff pitching rotation.  And not many teams lineups will pass through our gauntlet of ace-level starting pitchers easily.

BP Research Confirms Starting Pitcher Dominance
And this is confirmed by Baseball Prospectus' study of Playoff Success in their Baseball Between the Numbers book (for some reason it is out of print already).  When they examined the correlation between having three good starters and winning in the playoffs, it was one of the most significant that they found, among the metrics they examined.  And this correlation was even stronger when the team has a good overall starting rotation:  the only metrics stronger was having a good closer (per their WRXL reliever metric) and yielding a low opposing team batting average (which is best accomplished by having a very high K/9).  Their study shows the competitive advantage of having a great rotation.

And the PQS DOM stats shows the mechanics of how that works out when a team has that advantage.  The top pitchers are much more consistent in throwing a DOM start (yes, we all know this intuitively, but the PQS DOM stats gives a number to it).  When you have a rotation of them, you have a great chance to win roughly two-thirds of those starts.  Again, it is not like football or basketball, but that is dominating for baseball.  Compared to, say, simply a good rotation, where you only get DOM starts in roughly 40% of the starts.

Having such a good rotation won't win you series every season you get into the playoffs.  But it surely improves your chances of advancing greatly when you consistently get DOM starts in 67% (or more) of the starts vs. 40% of the starts for a simply good rotation.

Not Just Great Starts, Avoiding Bad Starts
And it helps not only in terms of more DOM games, but it also helps in reducing DIS (or disaster) starts.  When a pitcher has a disaster start, in their study, they had a 11.19 ERA.  That pretty much guarantees a loss for your team.  Good pitchers still have disaster starts (DIS) sometimes.  Sanchez had 18% last season, Bumgarner had 22%.  Elite pitchers like Lincecum and Cain had  DIS% of 18% and 6%, respectively (Lincecum had off year, had DIS% of 6% in 2009, 0% in 2008).

Thus, by having a good to great starter with high DOM%, you increase your chances of a well pitched start.  But the flip side of that is that also means less starts where you can possibly have a DIS start.  Those who can keep their DIS starts at a below 20% rate are among the best in the majors, and those below 10% are the elite.

PQS analysis, both DOM% and DIS%, helps to explain how having so many good starters in your rotation gives your team a competitive advantage in the playoffs.  Great DOM% makes it easier to win any particular start of the pitcher, but great DIS% also keeps your team in the game by keeping the score close, and giving them the opportunity to win a tight game.  The more DIS starts you have, the more games you pretty much automatically lose.

The Ying-Yang of Dirty:  Why I Wanted to Keep Him
Jonathan Sanchez is an example of how inconsistency, particularly a poor DIS, hurts a starter's ERA.  Over the past three seasons, he has been 45%Dom/31%DIS in 2008 (5.01 ERA), 41%DOM/24%DIS in 2009 (4.24 ERA), and 48%DOM/18%DIS in 2010 (3.07 ERA).  Clearly, progress with reducing DIS starts has helped his ERA, even though his percentage of DOM starts have not really increased.

Here is why I have been a Sanchez supporter over the years when people want to trade him.  In 2008, first half, he was 53%DOM/21%DIS, which put him among the best starters in the majors, before tiring out in the second half.  He screwed up his mechanics early in 2009, but in the second half, when he was going good, he had a 60%DOM/13%DIS.  In 2010, he had no excuse for his poor first half (33%DOM/22%DIS) but he turned it on by walking his talk with a stellar 67%DOM/13%DIS in the second half.  As I noted, 70%+ is what the elite starters do.

If he can do that consistently over a whole season, you got yourself an elite starter to go with Lincecum and Cain.  If he didn't tire out during the playoffs, we might have won series in less total games played.  If he does do that consistently, he would fit right in between Lincecum and Cain in terms of DOM/DIS PQS proportions.

This is why I argued to keep him while people were asking me when we should trade a starting pitcher to get a hitter.  He could be an elite starter, and is for long stretches of the season, though not over an entire season yet.   That makes our rotation that much more powerful a gauntlet for the other team to get through in the playoffs.