Pitcher Contact% and Strikeouts in 2009

Though it’s quite brief, I have to believe that Jeff Sullivan’s post on the correlation between pitcher contact rates and strikeouts from this past August has got to be one of the most exciting of the year. It makes so much sense — miss bats, get strikeouts — and yet Mr. Sullivan appears to be the only one on the interweb (besides our own Matthew Carruth) asking substantive questions about it*.

*If there are others, please don’t hesitate to mention them below.

You can see the correlation between the two in this image I’ve stolen directly, unforgivably from Lookout Landing:

Jeff's Contact Rate Image

That’s pretty striking. And it begs a question: If the correlation between Contact% and pitcher strikeouts is so strong, then couldn’t Contact% help us understand which pitchers might be poised to top the strikeout charts next year?

Yes. With some caveats.

In that same post to which I’ve linked above, Sullivan also notes that, beyond Contact% (league average 80.5%), pitchers also control their strikeout rates by other means, such as First-Pitch Strike% (league average 58.2%), Zone% (league average 49.3%), Called Strike%, etc. While none of these factors correlate as strongly as Contact% to pitcher strikeout rates, it does appear as though they’re not entirely negligible, either. To that end, I’ve included the first two of those below for comparison’s sake.

What happens when we run the 2009 numbers for Expected K%s (xK%)? That’s the question I asked myself — and which I answered, I think, by means of the kinda dumpy looking table you see below. Here are the Top 10 starting pitchers (50+ IP) by Contact% with expected and actual strikeout rates, plus K/9:

image003

So, what do we see here? Some likely suspects, for sure: Harden, Lincecum, Vazquez, de la Rosa. But also some surprises: Felipe Paulino (second!), Gio Gonzalez (third!), and Bud Norris. The first two of those guys I wrote about last week as tRA* surprises. Bud Norris I didn’t mention, but he actually finished 66th in tRA* out of the 183 pitchers with 150+ xOuts.

What conclusions can we draw from this? Well, we should probably be careful about that, as more research needs to be done in this area. Still, it’s probably reasonable — given his xK% and overall profile — to expect at least a small improvement in strikeouts from Felipe Paulino. Moreover, there’s reason to think — given his excellent Zone% — to think that Bud Norris will probably turn some of his walks into strikeouts.

How well do you think Norris will perform in 2010? Enter your Fan Projection here.




Print This Post



Carson Cistulli occasionally publishes spirited ejaculations at The New Enthusiast.


22 Responses to “Pitcher Contact% and Strikeouts in 2009”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Bobby Boden says:

    I thought about doing this same analysis, but the following thought crossed my mind: “Are people who deviate from the standard line really in store for more or less strikeouts the other year? Or is the reason they deviate from the standard line based on other skills?” The first and foremost of those skills in my mind, is the ability to slam the door with 2 strikes. Against many hitters, it’s a lot easier to get that #1 and #2 strike, but getting the #3 becomes more difficult, because they will be protecting more. If getting that last strike is part of the skill of striking somebody out (which it most likely is), then identifying outliers, is merely identifying people who are better at this skill. Getting a third strike on a batter, and your success at doing so, probably depends a lot on scouting the batters as well.

    Another skill that I included in my analysis was called strikes, this is another skill strikeout pitchers possesses (and seems like it is a different one, as some pitchers induce more called strikes then others). If you look among your top outliers, you’re going to see that many of them induce a lot of(or few) called strikes.

    Vote -1 Vote +1

  2. hamandcheese says:

    Missing the M in .com in the Bud Norris projection link.

    Vote -1 Vote +1

  3. Christian Seehausen says:

    Not surprised to see Norris and Paulino on this list. Those two pitchers are about the only thing an Astros fan can feel really good about improving their team for 2010.

    Don’t be surprised if one or both of them bring ace-like ERA pitching in front of what should also be a much-improved Houston defense.

    Speaking of Houston pitchers, any chance of another look at Wandy Rodriguez’s season at some point? Some prior posts sold him short, I think; he was a very good #2 starter, even if he wasn’t quite as good as his ERA made him look.

    Vote -1 Vote +1

  4. Joe says:

    Some of this analysis of baseball is starting to get ridiculous and out of hand. It’s getting to be too much.

    -6 Vote -1 Vote +1

    • Christian Seehausen says:

      If you don’t like in-depth baseball analysis, you’re probably on the wrong site. Just saying, you know.

      Vote -1 Vote +1

    • Scottwood says:

      When you logged on to FanGraphs, what did you expect? A discussion about how A Rod was only clutch this postseason b/c he was dating Kate Hudson? You can turn on ESPN to hear that crap.

      This analysis was really cool.

      Vote -1 Vote +1

    • Joe Morgan says:

      I couldn’t agree with you more.

      Seriously though I’ll remember this article come draft day and I pick up Bud Norris in the 20th round to a chorus of laughter.

      Vote -1 Vote +1

    • Mat Gonzales says:

      Easily one of the dumbest comments I’ve ever read on this site. Good work Joe. If your brain is overwhelmed go take some advil and lay down, the rest of us will continue doing your thinking for you.

      Vote -1 Vote +1

  5. SharksRog says:

    An analysis of strikeout likelihood should probably also focus on how frequently a pitcher gets two strikes on a hitter and the likelihood of either swing-throughs or called strikes on those pitchers.

    The pitcher I know best is Tim Lincecum. A high percentage of his strikeouts came on change ups, a pitch on which I believe he had a 43% swing-through rate. I don’t know how often his change is taken for a strike, but I don’t think it is a high percentage.

    Tim’s change up looks like a fastball, so very often the batter winds up swinging at a pitch that dips and/or fades outside the strike zone. His percentage of swinging strikes is high. I would venture that both his curve ball and slider also induce far more swinging strikeouts than called third strikes.

    Tim’s fastball is actually by far his easiest pitch to contact when swung at. It is likely also the pitch that is most often taken for a strike, although his curve ball probably ranks a close second.

    Tim is capable of blowing his fastball by a hitter with two strikes, but I think a fair percentage of his fastball strikeouts come when batters take a third strike while looking for the change up or a breaking pitch.

    Tim’s fastball is plenty good enough to keep hitters honest most of the time. But it is his secondary pitches — and in particular his change up — that achieve most of his strikeouts.

    In Tim’s case, he throws a higher percentage of secondary pitches with two strikes, which gives him a higher chance of a strikeout than merely looking at the swing-through and taken strikes percentage of his overall arsenal would indicate.

    I don’t know exactly what this means, but a high percentage of Tim’s strikeouts come with two outs. I don’t know if he develops more feel for his pitches as the inning goes on, if he takes special pleasure in ending an inning on a high note, if batters feel more pressure with two outs and thus are more vulnerable to his change up.

    I do know that Tim was nearly as likely as not to strike out the last batter of an inning. 104 of his 261 strikeouts came with two outs. His strikeouts with none and one out were close together, at 75 and 82, respectively. 40% of Tim’s strikeouts came with two outs.

    In 2008, Tim had 97 strikeouts with two outs, 82 with one out and 86 with no outs. As a rookie, there was little two-out bias., as Tim struck out 55 batters with two outs, 44 with one out and 51 with no outs.

    Tim has improved his strikeouts from over nine per nine innings in 2007 to over 10 per nine in 2008 and 2009. The difference has been in his added ability to strike out batters with two outs.

    I believe the most common strikeout combination for Tim is a swinging third strike on a change up with two outs. In fact, I don’t think there is any other combination of the three variables that comes overly close.

    Moral of the story: Don’t let Tim get two strikes on you — especially with two outs.

    Vote -1 Vote +1

  6. dcs says:

    Why didn’t you also look at O-Swing % ? It makes sense that hitters will miss a pitch more if it’s out of the strike zone.

    Vote -1 Vote +1

  7. Toffer Peak says:

    “Though it’s quite brief, I have to believe that Jeff Sullivan’s post on the correlation between pitcher contact rates and strikeouts from this past August has got to be one of the most exciting of the year…and yet Mr. Sullivan appears to be the only one on the interweb (besides our own Matthew Carruth) asking substantive questions about it*.

    *If there are others, please don’t hesitate to mention them below.”

    Looks like Jeff was beaten by a month and by a competing sister site nonetheless.

    http://www.draysbay.com/2009/7/21/956509/updated-expected-strikeouts-based

    Vote -1 Vote +1

    • Carson Cistulli says:

      That’s crazy. There’s no way I expected O-Contact% to be more important than Z-Contact%. I figure, if a pitcher can throw a ball in the strike zone and the batter can’t make contact — well, that’s a pretty good skill to have.

      Still, I’m not gonna argue with the numbers. I’ll probably just argue with my wife instead. And probably not about O- or Z-Contact Rates.

      Vote -1 Vote +1

      • Bronn says:

        That should tell you that a batter swinging at missing at pitches in the strike zone is largely a factor of luck, which makes sense.

        If the pitch is a strike and moves enough that a batter can’t hit it, it probably started out of the strike zone and therefore, he didn’t swing. Conversely, pitchers that batters are swinging and missing at usually start as strikes and break out of the strike zone.

        Called Strike% should be a better indicator, then, than Z-Contact%.

        Vote -1 Vote +1

  8. Simmy says:

    In reply to Bobby Boden:

    I think it bears mentioning that the original quoted article goes on to post a second graph which supports the notion that the difference between xK% and K% is skill and not luck. Namely, those with great differences between their xK% and K%s in one year seem to follow that up the very next year with a similar difference. If it was due to luck, then you’d expect the difference between xK% and K% to fluctuate wildly year to year.

    This kind of puts a damper on the whole thing, because it means you CAN’T really expect a Felipe Paulino or Bud Norris turnaround based merely on a large “xK%-K%”

    Vote -1 Vote +1

  9. You know, while it’s something of a no-brainer, as you stated, this actually surprises me quite a bit.

    These correlations between pitching numbers are far less frequent and dramatic than some might think.

    I did a bunch of stuff a few months ago on some of these stats (most of the stuff under “Plate Discipline,” velocity, and movement, looking at the Pitch Type Linear Weight data, and found that basically nothing correlated strongly.

    Who knew that fastball velocity has absolutely zero to do with fastball effectiveness? Not me, until I ran that.

    Anyway, this gives me some relief that SOMETHING in pitching influences something else; thank God we don’t have to take absolutely everything on a case-by-case basis.

    If anyone cares to read my findings on that, check out http://bleacherreport.com/articles/216816-heres-a-thought-examining-fastball-effectiveness.

    Anyways, great read as always, Carson! A lot of your pieces fascinate me.

    Vote -1 Vote +1

  10. Dave L. says:

    Obviously explains Tyler Clippard’s success.

    Among pitchers with at least 50 IP, Clippard was 8th lowest in all of the MLB with a contact rate of 68.7%

    His line drive % was also 5th lowest among all pitchers (with 50 IP) at 12.8%

    And of course… he sported the 4th lowest BABIP in the bigs (with 50 IP) at .207

    I know this article was about contact% but looking at all the stats I just posted, it seems to me like missing bats, and getting weak contact when you do have a pitch hit is actually a skill the pitcher controls.

    Am I interpreting this information wrong or does there seem to be a pattern that pitchers with low contact%’s are more likely to have some control on how hard a ball is actually hit? A skill, every pitcher would love to possess.

    Vote -1 Vote +1

  11. Detroit Michael says:

    “*If there are others, please don’t hesitate to mention them below.”

    Sorry that it’s available only to subscribers, but I wrote on article on July 1, 2009 on BaseballHQ.com on this topic too. Key findings were:
    - Of course K/9IP tends to persist from one year to the next, but adding Swinging Strike % (which is 1 – contact %) doesn’t appear to help. For the data set as a whole, the prior year’s expected Dom based on Swinging Strike % provides very little predictive information that wasn’t already captured in the prior year’s Dom.
    - However, there is predictive information when I looked at just the pitchers in the top 5 percentile. Out of 45 pitchers with expected Dom far below their actual Dom, 37 out of the 45 saw their actual Dom fall during the next year.

    My article cited the prior research on the topic that I found helpful. The first research I know of that said that swinging strike % or contact % wasn’t very predictive of future changes in K/9IP was a 2005 article by James Click.

    Vote -1 Vote +1

  12. Alex says:

    Has anybody studied whether any of this pitch data can be used to create a better predictor of future K% and BB% than K% and BB% themselves?

    Vote -1 Vote +1

  13. MDS says:

    I mainly look at O-Swing, O-Contact and F-Strike percentages when projecting k/9. Zone% and F-Strike% when sizing up the bb/9.

    Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Current ye@r *