Lesser CLIFFORD Candidates

When I originally published my findings around CLIFFORD — my metric for predicting players that are at a higher risk of experiencing a collapse in their wOBA (defined as a drop of at least .30 points of wOBA) — I presented a limited number of players for 2013. The list only included six players that qualified under the criteria. As a reminder, players that experienced a significant decline in three out of four metrics (Z-Contact%, FA%, UBR, Spd) were tagged as CLIFFORD candidates. These players had 3.4 times the odds of collapse (53% versus 25% for non-CLIFFORD players).

The single largest driver of collapse was change in Z-Contact% — the percent of pitches in the strike zone that a batter swings and makes contact with. Hitters who saw their Z-Contact% decline by at least 1.4% had 1.68 times the odds of collapsing than those that did not experience such a decline. Since there were far more players that qualified with their Z-Contact% than the full CLIFFORD criteria I thought it would be helpful to share that data with everyone.

Behold!

Here are the 68 players who saw their Z-Contact% decline by at least 1.4% from 2011-2012 (sorted by the size of their decline):

One thing to note is that the relationship does not appear linear — meaning, as a hitter’s Z-Contact% declines his wOBA does not decline a lockstep.

Some interesting names outside of the original that I highlighted (i.e. Robinson Cano and Curtis Granderson) appear on the list. Josh Hamilton is probably the highest-profile, in general and among the big name free-agents of this past off-season. His plate discipline was just atrocious last year, but adding in the fact that he also had problems making contact in the zone is yet another reason for the Angels to be concerned about their investment. Besides Hamilton, we also see two-thirds of the Braves outfield appear on the list, as Jason Hewyard and B.J. Upton saw their Z-Contact% decline by 4.7% an 4.3% respectively.

Based on my original research, 34% of hitters who saw similar declines in Z-Contact% experienced a wOBA decline of at least .30 points the following year. To some extent then, it appears that Z-Contact% could be a decent signal of trouble in hitters.

For one thing, it’s incredibly stable. In fact, Z-Contact% is one of the most stable of hitting metrics on a year-to-year basis. I originally found it had a correlation of .80, and Matt Klaassen’s replication of that analysis was very much in line with that finding (.82).

Furthermore, when we look at how a hitter’s Z-Contact% changes as they age we see that it generally climbs by a few percent until age-25. After that, it is relatively stable. There is some fluctuation, to be sure, but it is nothing compared to, say, O-Contact%, which falls dramatically starting at age-29.

AgingCurves_Hitters_Discipline_All

This suggests that when we see a drastic change in a hitter’s ability to make contact with pitches in the zone it should, if nothing else, spur us to dig deeper to understand why. Is the decline simply due to an injury? Will an off-season of rehab not only heal the injury, but allow the hitter to go back to their normal performance? Is it a mechanical issue? Or, is it a sign of accelerated aging? If a hitter has an injury that is likely to linger, it could signal a general change in ability that needs to be taken into account when projecting future performance. Relatedly, if the decline is the result of a slowing bat — one slowing at a faster rate than we would expect given age — that, too, is a significant issue. One way to get underneath that is to isolate Z-Contact% to just fastballs, or at least fastballs above, say, 93 mph. (This is something I am likely to look into in the coming weeks.)

The answer to these questions will determine how alarmed we should be about these kinds of large drops in Z-Contact%. In either case, it’s something else to keep your eye on.




Print This Post



Bill works as a consultant by day. In his free time, he writes for The Hardball Times, speaks about baseball research and analytics, consults for a Major League Baseball team and appears on MLB Network's Clubhouse Confidential. Along with Jeff Zimmerman, he won the 2013 SABR Analytics Research Award for Contemporary Analysis. Follow him on Tumblr or Twitter @BillPetti.


10 Responses to “Lesser CLIFFORD Candidates”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Bodybuilder on Cam says:

    Clifford is what I call my left pectoral when it’s all oiled up.

    Vote -1 Vote +1

  2. rotowizard says:

    Is it just me or does that graph look like the hardest question on the SAT?

    Vote -1 Vote +1

  3. martin short says:

    “At least 1.4% had 1.68 times the odds of collapsing than those that did not experience such a decline.”

    Do the odds for those with greater than 1.4% increase exponentially? i.e. What do the odds go to for those with a 4% or higher decline

    Vote -1 Vote +1

    • adam w says:

      I am guessing that Mr. Petti used a binary logistic regression to calculate this, which accounts for buckets but does not account for degrees (hence why it’s called binary). So there’s no way to be sure without completely re-running the analysis.

      Vote -1 Vote +1

  4. CJ in Austin Tx says:

    Bill James used to identify collapse candidates by large declines in the BB:K ratio of hitters. BB:K rates tend to be very stable for most hitters. The frequency of collapse for hitters who met his threshold decline was startling, even though many of them continued to produce when the BB:K decline initially occurred. About 5 years ago, he identified Garrett Atkins as a collapse candidate…and, well, you know what happened. A fundamental change in plate discipline seems to foretell an erosion in baseball skill. Perhaps Z-contact% will be a more precise leading indicator.

    Vote -1 Vote +1

  5. Steve Staude. says:

    Very interesting, Bill. Looks like Z-Contact% is an underappreciated stat for both hitters and pitchers.

    I’m very much interested in seeing your breakdown of what’s behind this Z-Contact% effect.

    Vote -1 Vote +1

    • CJ in Austin TX says:

      A question I wonder about is whether the ZContact% is a uniform indicator for all types of hitters. For example, is it a bust indicator for patient hitters but no so much for free swingers?

      Vote -1 Vote +1

      • Steve Staude. says:

        Great point — definitely worth looking into.

        I think one other possible cause for a steep decline in ZContact% is that pitchers have suddenly discovered a hole in a batter’s swing (or less ability to make contact with a certain type of pitch). So I hope Bill also looks at whether batters are being pitched to differently before vs. during the drop.

        Vote -1 Vote +1

  6. Ruki Motomiya says:

    The big red dog?

    Vote -1 Vote +1

  7. Moonraker says:

    Interesting, but it seems some of these candidates actually improved markedly despite a sizable drop in Z-contact%, Encarnacion and Pierzynski being the most prominent examples. Perhaps the drop in Z-contact% just means they are swinging harder?

    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>