More on the strikeout rate

Trevor May walked
three and struck out
four in four innings
Sunday.

The Monday print column grappled in large degree with the conflicting indicators for Mike Pelfrey and Kyle Gibson, who each have very good ERAs with very poor strikeout rates and strikeout/walk ratios.

I have for years referred to strikeouts per nine innings (K/9) and strikeout-walk ratio (K/BB) as the leading indicator stats for pitchers. Given a choice of pitchers, I'd rather have the guy strong in those ratios than in ERA, because they are better predictors of future ERA than ERA itself.

The Twins so far -- roughly one-fifth of the way through the season -- have used six starters. Gibson and Pelfrey have been in the rotation throughout, as has Phil Hughes. Trevor May joined the rotation in the second go-around, Ricky Nolasco spent some time on the disabled list and Tommy Milone, who opened the year in the rotation, is now at Triple A.

Of those six, only May and Hughes have good strikeout rates and strikeouts per walk. Culled from Baseball Reference in the early morning today and ranked by K/BB:

Hughes: 6.5 K/9. 8.00 K/BB
May: 6.9 K/9, 4.00 K/BB
Pelfrey: 4.2 K/9. 1.23 K/BB
Milone: 5.2 K/9, 1.18 K/BB
Nolasco: 4.8 K/9. 1.17 K/BB
Gibson: 2.7 K/9. 0,73 K/BB

The American League averages, for comparison: 7.4 K/9, 2.43 K/BB. So even May and Hughes, with easily the best strikeout rates in the rotation, are below league average in striking hitters out. (The league average includes relief pitchers, and their strikeout rates are higher, so May and Hughes aren't truly subpar in that category.)

I for years regarded the K/BB ratio as the more important of the leading indicator stats. I've become largely convinced, as strikeouts have mushroomed, that K/9 is now a bit more important. But either way, the bulk of the Twins rotation lags in these important predictors of future results.

Eye on 2014: Searching for strikeouts

Two key stats were at the root of the 2013 Minnesota Twins' "pitching" problem — pitching in quotation marks because one of the stats isn't about the guys on the mound so much as the men in the field.

• The Twins were far in arrears of all other teams at striking opponents out;
• The Twins were worst in the league at turning balls in play into outs.

It's a deadly combination, obviously; the pitching emphasis on letting the defense do the work is undermined when the defense doesn't make the plays.

The defensive issues get less attention than the "pitch to contact" approach, and it will in this post also. As the Twins search for upgrades to their staff, and specifically the starting rotation, the emphasis should be on strikeouts. They need to find some arms who can miss bats.

The average AL pitcher — the average — struck out 7.7 hitters per nine innings. This is the all-time high.

The average Twins pitcher struck out 6.1 men per nine. The Twins pitchers suppressed the league K rate by a point; take the Twins out, and the league rate is 7.8 K/9.

This is largely a starting pitcher problem. The Twins starters fanned 4.9 men per nine innings, the bullpen 7.9 — and having said that, the Twins relievers trailed the rest of the league's relievers as well. American League relief pitchers overall had an 8.5 K/9 rate. But the starters were at 7.2.

The prescription, then: Pursue strikeout rates.

Easier said, frankly, than done. This is Fangraphs' listing of free agent starting pitchers, sorted by K/9 rates 2011-13. There are 14 names on the list with K/9 rates about the 7.7 average rate. One of them, Tim Lincecum, is already off the market. Several others (Johan Santana, Josh Johnson and Roy Halladay among them) have significant health issues.

Earlier this week I suggested that Ubaldo Jimenez might encounter the same resistance as Kyle Lohse did last year in free agency. This list suggests otherwise. He's probably the safest bet among the free agent starters with high strikeout rates.

The pursuit of strikeouts, at least as it involves free agency, is going to require taking some significant risks. Almost all of the available high-K arms have issues. Which, I believe, is why Terry Ryan keeps repeating that there are no shortcuts. 

Scott Diamond and the missing strikeouts

Scott Diamond allowed just three hits through six innings
Thursday, but couldn't get through the seventh without help.

Scott Diamond on Tuesday had his first quality start (in the majors) since July 26. He got his first win (in the majors) since June 20.

It came against the White Sox, one of the few teams indisputably worse than the Twins, but it still had the left-hander feeling a good bit better about his pitching.

Me being me, I'm struck by what didn't happen in his outing: A strikeout.

Diamond got 19 outs, and none of them on strikes. For the season, Diamond is averaging just 3.5 strikeouts per nine innings. The American League average is 7.6, so Diamond's K rate is less than half the average.

I had to laugh when I came across this phrasing in the Pioneer Press gamer: "... Diamond failed to record a strikeout for just the third time this season ..." The implication is that it's common for starters to get a zero in that part of the linescore. 

It isn't. Going K-less in three starts in a season is unusual. And not a good sign.

Earl Weaver and pitchers strikeouts

Earl Weaver during the 1979 World
Series. 

Frequent readers know I believe in the predictive power of two pitchers stats: strikeout-to-walk ratio and strikeouts per nine innings.

Pitchers who are subpar in either of those markers might be successful for a year or two, but they are extremely poor bets to maintain that success. There are certainly commenters on this blog who dispute that; to be blunt, they don't know what they're talking about.

And having said that, Earl Weaver — whose historical reputation is based in large part on his ability to get big seasons from his starting pitchers — always had below-average strikeout pitchers.

Before I talk about the specific numbers for Weaver's moundsmen, I should note the drastic difference between his era and today. In 1979 — when Weaver won his fourth and final American League pennant — AL pitchers averaged 3.3 walks and 4.5 strikeouts per nine innings (that last the lowest K rate since 1955). That's a K/BB ratio of 1.36, which was also worse than a typical season, but not markedly out of line ( it was 1.39 in 1978, 1.44 in 1980).

In 2012 — no doubt because teams are increasingly focused on pitchers who can miss bats — the K/9 rate was 7.4 — a full half-strikeout higher than in any previous season. The past seven seasons have have the seven highest strikeout rates in AL history. And the walk rate is at historically low (post-deadball era) levels as well, 3.0 BB/9. The combination leads to a 2.45 K/BB ratio, the highest in league history.

(A recently as a decade ago it was safe to assume that a pitcher who struck out twice as many as he walked was going to be successful. Today he's well behind the field.)

But even accounting for the lower strikeout rates of the period, Weaver's pitchers didn't strike guys out. Weaver had 20-game winners almost every season, but he never had a pitcher rack up 200 strikeouts in a season. Even the great Jim Palmer, who had eight 20-win seasons and four times topped 300 innings, never fanned 200. Palmer's K rate for his career was a relatively mediocre 5.0 — and that's above the norm for Weaver's staffs. (Mike Cuellar, who averaged 18 wins in seven seasons pitching for Weaver, was at 4.5 K/9 in that period. Dave McNally, who also averaged 18 wins a year for Weaver, was at 4.4.)

As Chris Jaffe notes in this fine Hardball Times piece on Weaver, only the Royals and Brewers struck out fewer opponents in 1969-82, the core of Weaver's career.

How did he get away with it? Defense. Jaffe points to some defensive metrics that suggest the 1973 Orioles were the greatest defensive team ever, and the '69 Birds were the next best. Regardless of how accurate the metric is, Weaver certainly put a premium on glove work — men like Mark Belanger and Rich Dauer lasted a long time with him despite a lack of offensive production.

It's also worth noting that the low strikeout rates of the Orioles didn't stand out as much then as they would now. Combine those same strikeout rates with less than stellar defense in today's era, and you get the 2012 Twins pitching staff.

Problem spot: Carl Pavano

(This was originally set up to publish early Wednesday morning.  But then Francisco Liriano threw his no-hitter, and I pushed to back a day. Then the Twins won again on Wednesday, and I decided,  hey, let's not post "problem spot" pieces the day after a Twins win. So here it is, the day after an off day. Hey, it had to go sometime.)

Carl Pavano after his bat-swinging attack on a dugout trash can on Sunday:

"The (bat) wouldn't break. I couldn't break a bat in the dugout and I couldn't break any out there (on the mound). It was embarrassing."

An illuminating comment, that: It raises the notion, which I should have thought of earlier, that breaking bats is a way those who "pitch to contact" gauge their effectiveness.

Pavano's strikeout rate after five starts is 4.1 K/9, which is down some from last season's K rate and down sharply from his 7.2 K/9 with the Twins in 2009. The 2009 rate was a bit of fluke; Pavano's career rate is 5.7. (Stats line here.) Last season's decline is why I would have preferred to see the Twins let Pavano sign elsewhere during the winter.

Pavano was, as we know, effective last season (3.75 ERA in 221 innings) and ineffective so far this season (5.84 ERA). That's a two-runs per nine innings swing, which is pretty significant.

Without question, part of that swing is explained by the deteriorating Minnesota infield defense. But I suspect that part of it is that Pavano isn't breaking bats — which means batters are hitting more line drives against him, and fewer pop-ups and weak grounders.

Pavano is never going to be a high strikeout guy. He's got to keep the ball off the sweet spot of the bat. This is something I'm going to start watching for with Pavano — and, for that matter, Nick Blackburn and Brian Duensing, two other Twins starters whose strikeout rates are either subpar (Blackburn) or just barely acceptable (Duensing).